Attention: Here be dragons
This is the latest
(unstable) version of this documentation, which may document features
not available in or compatible with released stable versions of Godot.
Checking the stable version of the documentation...
ProjectSettings
繼承： Object
儲存全域變數。
說明
儲存可以從任何地方存取的變數。請使用 get_setting()、set_setting()、has_setting() 存取。儲存在 project.godot 中的變數也會被載入到 ProjectSettings 中，因此這個物件常用於讀取自訂遊戲配置選項。
指定“專案設定”的屬性時，請使用設定的完整路徑，包括類別。例如專案名稱應使用 "application/config/name"。類別和屬性名稱可以在“專案設定”對話框中查看。
功能標籤：可以使用特性標籤來針對特定的平臺和配置（除錯、發行……）做專案設定的覆蓋。
覆蓋：在專案的根目錄下建立名為 override.cfg 的檔案，就可以對任意專案設定進行覆蓋。對於已匯出的專案，把這個檔放在與專案二進位檔案相同的目錄下，也可以達到覆蓋的目的。覆蓋時仍會考慮基礎專案設定的功能標籤。因此，如果你想讓它們在所有平臺和配置上覆蓋基礎專案設定，請確保也用所需的功能標籤覆蓋該設定。
教學
屬性
方法
void |
add_property_info(hint: Dictionary) |
check_changed_settings_in_group(setting_prefix: String) const |
|
void |
|
get_changed_settings() const |
|
get_setting(name: String, default_value: Variant = null) const |
|
get_setting_with_override(name: StringName) const |
|
get_setting_with_override_and_custom_features(name: StringName, features: PackedStringArray) const |
|
globalize_path(path: String) const |
|
has_setting(name: String) const |
|
load_resource_pack(pack: String, replace_files: bool = true, offset: int = 0) |
|
localize_path(path: String) const |
|
save() |
|
save_custom(file: String) |
|
void |
set_as_basic(name: String, basic: bool) |
void |
set_as_internal(name: String, internal: bool) |
void |
set_initial_value(name: String, value: Variant) |
void |
|
void |
set_restart_if_changed(name: String, restart: bool) |
void |
set_setting(name: String, value: Variant) |
訊號
settings_changed() 🔗
Emitted when any setting is changed, up to once per process frame.
屬性說明
String accessibility/general/accessibility_driver = "accesskit" 🔗
Accessibility driver:
-accesskit (default): AccessKit driver.
-dummy: Dummy driver, screen reader support is disabled.
int accessibility/general/accessibility_support = 0 🔗
Accessibility support mode:
Auto (
0): Accessibility support is enabled, but updates to the accessibility information are processed only if an assistive app (such as a screen reader or a Braille display) is active (default).Always Active (
1): Accessibility support is enabled, and updates to the accessibility information are always processed, regardless of the status of assistive apps.Disabled (
2): Accessibility support is fully disabled.
Note: Accessibility debugging tools, such as Accessibility Insights for Windows, Accessibility Inspector (macOS), or AT-SPI Browser (Linux/BSD), do not count as assistive apps. To test your project with these tools, use Always Active.
int accessibility/general/updates_per_second = 60 🔗
The number of accessibility information updates per second.
bool animation/compatibility/default_parent_skeleton_in_mesh_instance_3d = false 🔗
If true, MeshInstance3D.skeleton will point to the parent node (..) by default, which was the behavior before Godot 4.6. It's recommended to keep this setting disabled unless the old behavior is needed for compatibility.
Note: If you disable this option in an existing project, it's strongly recommended to use the Project > Tools > Upgrade Project Files... option to ensure existing scenes do not break.
bool animation/warnings/check_angle_interpolation_type_conflicting = true 🔗
If true, AnimationMixer prints the warning of interpolation being forced to choose the shortest rotation path due to multiple angle interpolation types being mixed in the AnimationMixer cache.
bool animation/warnings/check_invalid_skeleton_modifier_node_paths = true 🔗
If true, SkeletonModifier3D prints a warning if there's no matching object for the track path in the scene when assigning.
bool animation/warnings/check_invalid_track_paths = true 🔗
If true, AnimationMixer prints the warning of no matching object of the track path in the scene.
Color application/boot_splash/bg_color = Color(0.14, 0.14, 0.14, 1) 🔗
啟動介面的背景色。
String application/boot_splash/image = "" 🔗
Path to an image used as the boot splash. If left empty, the default Godot Engine splash will be displayed instead.
Note: Only effective if application/boot_splash/show_image is true.
Note: The only supported format is PNG. Using another image format will result in an error.
Note: The image will also show when opening the project in the editor. If you want to display the default splash image in the editor, add an empty override for editor_hint feature.
int application/boot_splash/minimum_display_time = 0 🔗
啟動畫面的最小顯示時間（單位為毫秒）。不建議設定為過高的值。
bool application/boot_splash/show_image = true 🔗
如果為 true，將在引擎啟動時顯示 application/boot_splash/image 所指定的圖像。為 false 時，僅顯示 application/boot_splash/bg_color 所指定的純色。
int application/boot_splash/stretch_mode = 1 🔗
Specifies how the splash image will be stretched. For the original size without stretching, set to disabled. See SplashStretchMode constants for more information.
bool application/boot_splash/use_filter = true 🔗
如果為 true，則在縮放圖像時會套用線性篩選（推薦用於高解析度圖稿）。如果為 false，則使用最近鄰插值（推薦用於圖元畫）。
bool application/config/auto_accept_quit = true 🔗
如果為 true，則該套用程式會自動接受退出請求。
String application/config/custom_user_dir_name = "" 🔗
This user directory is used for storing persistent data (user:// filesystem). If a custom directory name is defined, this name will be appended to the system-specific user data directory (same parent folder as the Godot configuration folder documented in OS.get_user_data_dir()).
The application/config/use_custom_user_dir setting must be enabled for this to take effect.
Note: If application/config/custom_user_dir_name contains trailing periods, they will be stripped as folder names ending with a period are not allowed on Windows.
String application/config/description = "" 🔗
專案的描述，在專案管理器中懸停時顯示為工具提示。
bool application/config/disable_project_settings_override = false 🔗
If true, disables loading of project settings overrides (file defined in application/config/project_settings_override and res://override.cfg) and related CLI arguments.
String application/config/icon = "" 🔗
專案所使用的圖示，在專案載入時設定。匯出器也會在必要時使用此圖示作為退回。
String application/config/macos_native_icon = "" 🔗
設定 macOS 上遊戲的圖示，使用 .icns 圖示集格式。啟動時會通過調用 DisplayServer.set_native_icon() 自動完成。
String application/config/name = "" 🔗
專案名稱。會在專案管理器和匯出器中使用。可以通過翻譯當地語系化檔中的值來翻譯項目名稱。視窗標題將設定為在啟動時自動配對專案名稱。
注意：如果 application/config/use_custom_user_dir 為 false，更改此值也會更改使用者資料檔案夾的路徑。重命名專案後，你將無法再存取 user:// 中的現有資料，除非你重命名舊資料夾以配對新項目名稱。有關更多資訊，請參閱文件中的 《資料路徑》。
Dictionary application/config/name_localized = {} 🔗
Translations of the project's name. This setting is used by OS tools to translate application name on Android, iOS and macOS.
Note: When left empty, the application name is translated using the project translations.
String application/config/project_settings_override = "" 🔗
指定一個檔來覆蓋專案設定。例如：user://custom_settings.cfg。參見 ProjectSettings 頂部類描述中的“Overriding”以獲取更多資訊。
注意：不管這個設定的值如何，res://override.cfg 仍然會被讀取來覆蓋專案設定。
bool application/config/quit_on_go_back = true 🔗
如果為 true，則該套用程式會在導覽返回時自動退出（例如在 Android 上使用系統“返回”鍵）。
bool application/config/use_custom_user_dir = false 🔗
如果為 true，專案會將使用者資料保存到它自己的使用者目錄中。如果 application/config/custom_user_dir_name 為空，將使用 <作業系統使用者資料目錄>/<專案名稱> 目錄。如果為 false，專案會將使用者資料保存到 <作業系統使用者資料目錄>/Godot/app_userdata/<專案名稱>。
另見 Godot 專案中的檔路徑。該設定僅在桌面平臺上有效。
如果為 true，該專案會使用隱藏目錄（.godot）來儲存項目特定的資料（中繼資料、著色器快取等）。
如果為 false，將改用非隱藏目錄（godot）。
注意：請在更改此設定後重新開機套用程式。
注意：修改這個值可能有助於在禁止使用隱藏目錄模式的平臺或協力廠商工具上使用。但是，僅在你確定你的環境需要更改此設定時才進行修改，因為某些外部工具或插件可能使用預設的 .godot 資料夾，因此更改預設設定可能會影響它們的使用。
String application/config/version = "" 🔗
專案的人類可讀版本識別碼。如果版本識別碼沒有在那裡被覆寫，則匯出器將使用它。如果application/config/version 是空字串並且版本標識符在匯出器中未被覆蓋，匯出器將使用1.0.0 作為版本標識符。
String application/config/windows_native_icon = "" 🔗
設定 Windows 上遊戲的圖示，使用 .ico 圖示集格式。啟動時會通過調用 DisplayServer.set_native_icon() 自動完成。
bool application/run/delta_smoothing = true 🔗
影格差異量的時間樣本可能會受到平臺帶來的隨機影響，即使垂直同步能夠讓影格按照固定的間隔顯示也無濟於事。因此可能導致卡頓。差異量平滑通常能夠帶來更好的效果，會通過篩選輸入差異量來修改更新率帶來的微小波動。
注意：只會在 display/window/vsync/vsync_mode 為 enabled 時嘗試進行差異量平滑，因為沒有垂直同步時無法很好地工作。
最開始啟動平滑前可能會需要有若干秒的穩定影格率。只會在性能足以讓算繪影格與更新率配對的機器上啟動。
bool application/run/disable_stderr = false 🔗
If true, disables printing to standard error. If true, this also hides error and warning messages printed by @GlobalScope.push_error() and @GlobalScope.push_warning(). See also application/run/disable_stdout.
Changes to this setting will only be applied upon restarting the application. To control this at runtime, use Engine.print_error_messages.
bool application/run/disable_stdout = false 🔗
If true, disables printing to standard output. This is equivalent to starting the editor or project with the --quiet command line argument. See also application/run/disable_stderr.
Changes to this setting will only be applied upon restarting the application. To control this at runtime, use Engine.print_to_stdout.
If true, allows the Alt + Space keys to display the window menu. This menu allows the user to perform various window management operations such as moving, resizing, or minimizing the window.
Note: When the menu is displayed, project execution will pause until the menu is fully closed due to Windows behavior. Consider this when enabling this setting in a networked multiplayer game. The menu is only considered fully closed when an option is selected, when the user clicks outside, or when Escape is pressed after bringing up the window menu and another key is pressed afterwards.
Note: This setting is implemented only on Windows.
bool application/run/flush_stdout_on_print = false 🔗
如果為 true，則每次列印一行時更新標準輸出流。這會影響終端紀錄記錄和檔紀錄記錄。
運作專案時，如果希望由 systemd/journalctl 等服務管理器收集紀錄，則必須啟用此設定。預設情況下，在發行版本中禁用此設定，因為如果快速連續列印大量行，則在每個打印行上更新都會對性能產生負面影響。此外，如果啟用此設定，如果套用程式當機或以其他方式被使用者殺死（不會“正常”關閉），則仍會成功寫入紀錄檔。
注意：無論此設定如何，標準錯誤流 (stderr) 在列印一行時總是被更新。
對此設定的更改只會在重新開機套用程式時套用。
bool application/run/flush_stdout_on_print.debug = true 🔗
除錯建構覆蓋application/run/flush_stdout_on_print，因為在除錯過程中性能不那麼重要。
僅在重新開機套用程式時才會套用此設定的更改。
int application/run/frame_delay_msec = 0 🔗
Forces a constant delay between frames in the main loop (in milliseconds). In most situations, application/run/max_fps should be preferred as an FPS limiter as it's more precise.
This setting can be overridden using the --frame-delay <ms;> command line argument.
bool application/run/load_shell_environment = false 🔗
If true, loads the default shell and copies environment variables set by the shell startup scripts to the app environment.
Note: This setting is implemented on macOS for non-sandboxed applications only.
bool application/run/low_processor_mode = false 🔗
If true, enables low-processor usage mode. When enabled, the engine takes longer to redraw, but only redraws the screen if necessary. This may lower power consumption, and is intended for editors or mobile applications. For most games, because the screen needs to be redrawn every frame, it is recommended to keep this setting disabled.
int application/run/low_processor_mode_sleep_usec = 6900 🔗
啟用低處理器使用模式時影格間的睡眠量（以微秒計）。值越高，CPU佔用率越低。
String application/run/main_loop_type = "SceneTree" 🔗
實作引擎主迴圈的型別名稱。
String application/run/main_scene = "" 🔗
專案運作時將載入的主場景檔的路徑。
int application/run/max_fps = 0 🔗
Maximum number of frames per second allowed. A value of 0 means "no limit". The actual number of frames per second may still be below this value if the CPU or GPU cannot keep up with the project logic and rendering.
Limiting the FPS can be useful to reduce system power consumption, which reduces heat and noise emissions (and improves battery life on mobile devices).
If display/window/vsync/vsync_mode is set to Enabled or Adaptive, it takes precedence and the forced FPS number cannot exceed the monitor's refresh rate. See also DisplayServer.screen_get_refresh_rate().
If display/window/vsync/vsync_mode is Enabled, on monitors with variable refresh rate enabled (G-Sync/FreeSync), using an FPS limit slightly lower than the monitor's refresh rate will reduce input lag while avoiding tearing. At higher refresh rates, the difference between the FPS limit and the monitor refresh rate should be increased to ensure frames to account for timing inaccuracies. The optimal formula for the FPS limit value in this scenario is r - (r * r) / 3600.0, where r is the monitor's refresh rate.
If display/window/vsync/vsync_mode is Disabled, limiting the FPS to a high value that can be consistently reached on the system can reduce input lag compared to an uncapped framerate. Since this works by ensuring the GPU load is lower than 100%, this latency reduction is only effective in GPU-bottlenecked scenarios, not CPU-bottlenecked scenarios.
See also physics/common/physics_ticks_per_second.
This setting can be overridden using the --max-fps <fps> command line argument (including with a value of 0 for unlimited framerate).
Note: This property is only read when the project starts. To change the rendering FPS cap at runtime, set Engine.max_fps instead.
bool application/run/print_header = true 🔗
If true, the engine header is printed in the console on startup. This header describes the current version of the engine, as well as the renderer being used. This behavior can also be disabled on the command line with the --no-header option.
float audio/buses/channel_disable_threshold_db = -60.0 🔗
當聲音在給定的時間內低於給定的 dB 閾值時，音訊匯流排將自動關閉。這可以節省 CPU，因為分配給該匯流排的效果將不再做任何處理。
float audio/buses/channel_disable_time = 2.0 🔗
當聲音在給定的時間內低於給定的 dB 閾值時，音訊匯流排將自動關閉。這可以節省 CPU，因為分配給該匯流排的效果將不再做任何處理。
String audio/buses/default_bus_layout = "res://default_bus_layout.tres" 🔗
專案中使用的預設 AudioBusLayout 資源檔，除非被場景覆蓋。
Specifies the audio driver to use. This setting is platform-dependent as each platform supports different audio drivers. If left empty, the default audio driver will be used.
The Dummy audio driver disables all audio playback and recording, which is useful for non-game applications as it reduces CPU usage. It also prevents the engine from appearing as an application playing audio in the OS' audio mixer.
To query the value that is being used at run-time (which may be overridden by command-line arguments or headless mode), use AudioServer.get_driver_name().
Note: The driver in use can be overridden at runtime via the --audio-driver command line argument.
bool audio/driver/enable_input = false 🔗
If true, microphone input will be allowed. This requires appropriate permissions to be set when exporting to Android or iOS.
Note: If the operating system blocks access to audio input devices (due to the user's privacy settings), audio capture will only return silence. On Windows, make sure that apps are allowed to access the microphone in the OS' privacy settings.
int audio/driver/mix_rate = 44100 🔗
Target mixing rate used for audio (in Hz). In general, it's better to not touch this and leave it to the host operating system.
Note: On iOS and macOS, mixing rate is determined by audio driver, this value is ignored.
Note: Input and output mixing rates might be different. Use AudioServer.get_mix_rate() and AudioServer.get_input_mix_rate() to get actual values.
int audio/driver/mix_rate.web = 0 🔗
audio/driver/mix_rate 在 Web 平臺上更安全的覆蓋項。這裡的 0 表示“讓流覽器選擇”（因為有些流覽器不喜歡強制混合率）。
int audio/driver/output_latency = 15 🔗
Specifies the preferred output latency in milliseconds for audio. Lower values will result in lower audio latency at the cost of increased CPU usage. Low values may result in audible crackling on slower hardware.
Audio output latency may be constrained by the host operating system and audio hardware drivers. If the host can not provide the specified audio output latency then Godot will attempt to use the nearest latency allowed by the host. As such you should always use AudioServer.get_output_latency() to determine the actual audio output latency.
Audio output latency can be overridden using the --audio-output-latency <ms> command line argument.
Note: This setting is ignored on Android.
int audio/driver/output_latency.web = 50 🔗
audio/driver/output_latency 在 Web 平臺上更安全的覆蓋項，能夠避免一些音訊問題，尤其是在移動裝置上。
float audio/general/2d_panning_strength = 0.5 🔗
所有 AudioStreamPlayer2D 節點的聲像效果的基本強度。可以使用 AudioStreamPlayer2D.panning_strength 在每個節點上進一步縮放聲像強度。0.0 的值會完全禁用身歷聲聲像，只保留音量衰減。如果聲音恰好位於聽者的左側（或右側），則 1.0 的值會使其中一個通道完全靜音。
0.5 的預設值是針對耳機進行調諧的。當使用揚聲器時，可能會發現較低的值效果更好，因為與耳機相比，揚聲器的身歷聲分離度較低。
float audio/general/3d_panning_strength = 0.5 🔗
The base strength of the panning effect for all AudioStreamPlayer3D nodes. The panning strength can be further scaled on each Node using AudioStreamPlayer3D.panning_strength. A value of 0.0 disables stereo panning entirely, leaving only volume attenuation in place. A value of 1.0 completely mutes one of the channels if the sound is located exactly to the left (or right) of the listener.
The default value of 0.5 is tuned for headphones which means that the opposite side channel goes no lower than 50% of the volume of the nearside channel. You may find that you can set this value higher for speakers to have the same effect since both ears can hear from each speaker.
int audio/general/default_playback_type = 0 🔗
實驗性： 此屬性可能在未來版本中變更或移除。
Specifies the default playback type of the platform.
The default value is set to Stream, as most platforms have no issues mixing streams.
int audio/general/default_playback_type.web = 1 🔗
實驗性： 此屬性可能在未來版本中變更或移除。
Specifies the default playback type of the Web platform.
The default value is set to Sample as the Web platform is not suited to mix audio streams outside of the Web Audio API, especially when exporting a single-threaded game. Sample allows for lower latency on the web platform at the cost of flexibility (AudioEffects are not supported).
Warning: Forcing Stream on the Web platform may cause high audio latency and crackling, especially when exporting a multi-threaded game.
bool audio/general/ios/mix_with_others = false 🔗
Sets the mixWithOthers option for the AVAudioSession on iOS. This will override the mix behavior, if the category is set to Play and Record, Playback, or Multi Route.
Ambient always has this set per default.
int audio/general/ios/session_category = 0 🔗
Sets the AVAudioSessionCategory on iOS. Use the Playback category to get sound output, even if the phone is in silent mode.
bool audio/general/text_to_speech = false 🔗
If true, text-to-speech support is enabled on startup, otherwise it is enabled the first time any TTS method is used. See also DisplayServer.tts_get_voices() and DisplayServer.tts_speak().
Note: Enabling TTS can cause additional idle CPU usage and interfere with the sleep mode, so consider disabling it if TTS is not used.
int audio/video/video_delay_compensation_ms = 0 🔗
Setting to hardcode audio delay when playing video. Best to leave this unchanged unless you know what you are doing.
bool collada/use_ambient = false 🔗
如果為 true，則該線將是垂直的。如果 false，該線將是水平的。
int compression/formats/gzip/compression_level = -1 🔗
gzip 的預設壓縮層級。影響壓縮的場景和資源。較高的級別會以壓縮速度為代價導致文件變小。解壓縮速度大多不受壓縮層級的影響。-1 使用預設的 gzip 壓縮級別，該級別與 6 相同，但由於底層 zlib 更新，未來可能會發生變化。
int compression/formats/zlib/compression_level = -1 🔗
Zlib 的預設壓縮層級。影響壓縮的場景和資源。較高的級別會以壓縮速度為代價導致文件變小。解壓縮速度大多不受壓縮層級的影響。-1 使用預設的 gzip 壓縮級別，該級別與 6 相同，但由於底層 zlib 更新，未來可能會發生變化。
int compression/formats/zstd/compression_level = 3 🔗
Zstandard 的預設壓縮層級。影響壓縮的場景和資源。較高的級別會以壓縮速度為代價導致檔案變小。解壓縮速度大多不受壓縮層級的影響。
bool compression/formats/zstd/long_distance_matching = false 🔗
啟用 Zstandard 的長距離配對。
int compression/formats/zstd/window_log_size = 27 🔗
使用 Zstandard 的長距離配對進行壓縮時，允許的最大大小限制（2 的冪）。更高的值可以產生更好的壓縮，但是在壓縮和解壓縮時需要更多的記憶體。
Color debug/canvas_items/debug_redraw_color = Color(1, 0.2, 0.2, 0.5) 🔗
If canvas item redraw debugging is active, this color will be flashed on canvas items when they redraw.
float debug/canvas_items/debug_redraw_time = 1.0 🔗
If canvas item redraw debugging is active, this will be the time the flash will last each time they redraw.
bool debug/file_logging/enable_file_logging = false 🔗
If true, logs all output and error messages to files. See also debug/file_logging/log_path, debug/file_logging/max_log_files, and application/run/flush_stdout_on_print.
bool debug/file_logging/enable_file_logging.pc = true 🔗
debug/file_logging/enable_file_logging 在桌面平臺的覆蓋項，因為在移動/Web 平臺上不容易存取紀錄檔。
String debug/file_logging/log_path = "user://logs/godot.log" 🔗
Path at which to store log files for the project. Using a path under user:// is recommended.
This can be specified manually on the command line using the --log-file <file> command line argument. If this command line argument is specified, log rotation is automatically disabled (see debug/file_logging/max_log_files).
int debug/file_logging/max_log_files = 5 🔗
Specifies the maximum number of log files allowed (used for rotation). Set to 1 to disable log file rotation.
If the --log-file <file> command line argument is used, log rotation is always disabled.
int debug/gdscript/warnings/assert_always_false = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an assert call always evaluates to false.
int debug/gdscript/warnings/assert_always_true = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an assert call always evaluates to true.
int debug/gdscript/warnings/confusable_capture_reassignment = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a local variable captured by a lambda is reassigned, since this does not modify the outer local variable.
int debug/gdscript/warnings/confusable_identifier = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an identifier contains characters that can be confused with something else, like when mixing different alphabets.
int debug/gdscript/warnings/confusable_local_declaration = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an identifier declared in the nested block has the same name as an identifier declared below in the parent block.
int debug/gdscript/warnings/confusable_local_usage = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an identifier that will be shadowed below in the block is used.
int debug/gdscript/warnings/confusable_temporary_modification = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a built-in property of type Packed*Array is modified using a complex assignment chain or a non-const method call. In this case, you are only modifying a temporary value, and the property's value remains unchanged.
int debug/gdscript/warnings/deprecated_keyword = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when deprecated keywords are used.
Note: There are currently no deprecated keywords, so this warning is never produced.
Dictionary debug/gdscript/warnings/directory_rules = { "res://addons": 0 } 🔗
The rules for including or excluding scripts when generating warnings, as a dictionary. Each rule is an entry consisting of a directory path (key) and a decision (value). When trying to generate a warning, the GDScript parser chooses the most specific rule, i.e. the most nested directory containing the script. If the decision is Exclude, warnings are not generated for this script. If the decision is Include or the script doesn't satisfy any of the rules, the warning configuration specified in the Project Settings is applied.
It is recommended to include your own addons/libraries, either project-specific or actively being developed at the moment. Third-party or project-agnostic addons/libraries should be excluded, as they may be incompatible with the project's warning configuration.
Note: It is not recommended to remove or change the rule for "res://addons" as the project's warning configuration may break third-party addons. Instead, consider including individual addons, if necessary.
Note: The editor does not check whether the specified paths are existing directories. It also does not automatically update these paths when directories are moved.
int debug/gdscript/warnings/empty_file = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an empty file is parsed.
bool debug/gdscript/warnings/enable = true 🔗
如果為 true，則啟用特定的 GDScript 警告（請參閱 debug/gdscript/warnings/* 設定）。如果為 false，則禁用所有 GDScript 警告。
int debug/gdscript/warnings/enum_variable_without_default = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a variable has an enum type but no explicit default value, but only if the enum does not contain 0 as a valid value.
int debug/gdscript/warnings/get_node_default_without_onready = 2 🔗
When set to Warn or Error, produces a warning or an error respectively when Node.get_node() (or the shorthand $) is used as default value of a class variable without the @onready annotation.
int debug/gdscript/warnings/incompatible_ternary = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a ternary operator may emit values with incompatible types.
int debug/gdscript/warnings/inference_on_variant = 2 🔗
When set to Warn or Error, produces a warning or an error respectively when a static inferred type uses a Variant as initial value, which makes the static type to also be Variant.
int debug/gdscript/warnings/inferred_declaration = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when a variable, constant, or parameter has an implicitly inferred static type. In GDScript, type inference is performed by declaring a variable with := instead of = and leaving out the type specifier. For example, var x := 1 will infer the int type, while var x: int = 1 explicitly declares the variable as int.
Note: This warning is recommended in addition to debug/gdscript/warnings/untyped_declaration if you want to always specify the type explicitly. Having INFERRED_DECLARATION warning level higher than UNTYPED_DECLARATION warning level makes little sense and is not recommended.
int debug/gdscript/warnings/int_as_enum_without_cast = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when trying to use an integer as an enum without an explicit cast.
int debug/gdscript/warnings/int_as_enum_without_match = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when trying to use an integer as an enum when there is no matching enum member for that numeric value.
int debug/gdscript/warnings/integer_division = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when dividing an integer by another integer (the decimal part will be discarded).
int debug/gdscript/warnings/missing_await = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when calling a coroutine without await.
int debug/gdscript/warnings/missing_tool = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when the base class script has the @tool annotation, but the current class script does not have it.
int debug/gdscript/warnings/narrowing_conversion = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when passing a floating-point value to a function that expects an integer (it will be converted and lose precision).
int debug/gdscript/warnings/native_method_override = 2 🔗
When set to Warn or Error, produces a warning or an error respectively when a method in the script overrides a native method, because it may not behave as expected.
int debug/gdscript/warnings/onready_with_export = 2 🔗
When set to Warn or Error, produces a warning or an error respectively when the @onready annotation is used together with the @export annotation, since it may not behave as expected.
int debug/gdscript/warnings/redundant_await = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a function that is not a coroutine is called with await.
int debug/gdscript/warnings/redundant_static_unload = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when the @static_unload annotation is used in a script without any static variables.
bool debug/gdscript/warnings/renamed_in_godot_4_hint = true 🔗
啟用後，使用自 Godot 3 以來重命名的屬性、列舉或函式，將在發生錯誤時產生一個提示。
int debug/gdscript/warnings/return_value_discarded = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when calling a function without using its return value (by assigning it to a variable or using it as a function argument). These return values are sometimes used to indicate possible errors using the Error enum.
int debug/gdscript/warnings/shadowed_global_identifier = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when defining a local or member variable, signal, or enum that would have the same name as a built-in function or global class name, thus shadowing it.
int debug/gdscript/warnings/shadowed_variable = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a local variable or local constant shadows a member declared in the current class.
int debug/gdscript/warnings/shadowed_variable_base_class = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a local variable or local constant shadows a member declared in a base class.
int debug/gdscript/warnings/standalone_expression = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when calling an expression that may have no effect on the surrounding code, such as writing 2 + 2 as a statement.
int debug/gdscript/warnings/standalone_ternary = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when calling a ternary expression that may have no effect on the surrounding code, such as writing 42 if active else 0 as a statement.
int debug/gdscript/warnings/static_called_on_instance = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when calling a static method from an instance of a class instead of from the class directly.
int debug/gdscript/warnings/unassigned_variable = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when using a variable that wasn't previously assigned.
int debug/gdscript/warnings/unassigned_variable_op_assign = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when assigning a variable using an assignment operator like += if the variable wasn't previously assigned.
int debug/gdscript/warnings/unreachable_code = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when unreachable code is detected (such as after a return statement that will always be executed).
int debug/gdscript/warnings/unreachable_pattern = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when an unreachable match pattern is detected.
int debug/gdscript/warnings/unsafe_call_argument = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when using an expression whose type may not be compatible with the function parameter expected.
int debug/gdscript/warnings/unsafe_cast = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when a Variant value is cast to a non-Variant.
int debug/gdscript/warnings/unsafe_method_access = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when calling a method whose presence is not guaranteed at compile-time in the class.
int debug/gdscript/warnings/unsafe_property_access = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when accessing a property whose presence is not guaranteed at compile-time in the class.
int debug/gdscript/warnings/unsafe_void_return = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when returning a call from a void function when such call cannot be guaranteed to be also void.
int debug/gdscript/warnings/untyped_declaration = 0 🔗
When set to Warn or Error, produces a warning or an error respectively when a variable or parameter has no static type, or if a function has no static return type.
Note: This warning is recommended together with EditorSettings.text_editor/completion/add_type_hints to help achieve type safety.
int debug/gdscript/warnings/unused_local_constant = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a local constant is never used.
int debug/gdscript/warnings/unused_parameter = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a function parameter is never used.
int debug/gdscript/warnings/unused_private_class_variable = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a private member variable is never used.
int debug/gdscript/warnings/unused_signal = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a signal is declared but never explicitly used in the class.
int debug/gdscript/warnings/unused_variable = 1 🔗
When set to Warn or Error, produces a warning or an error respectively when a local variable is unused.
String debug/settings/crash_handler/message = "Please include this when reporting the bug to the project developer." 🔗
引擎當機時，在回溯之前顯示的消息。預設情況下，由於僅限編輯器的覆蓋會套用於該設定，故該消息僅用於匯出的專案中。
String debug/settings/crash_handler/message.editor = "Please include this when reporting the bug on: https://github.com/godotengine/godot/issues" 🔗
debug/settings/crash_handler/message 的僅限編輯器的覆蓋。不會影響以除錯或發行模式匯出的專案。
bool debug/settings/gdscript/always_track_call_stacks = false 🔗
Whether GDScript call stacks will be tracked in release builds, thus allowing Engine.capture_script_backtraces() to function.
Note: This setting has no effect on editor builds or debug builds, where GDScript call stacks are tracked regardless.
bool debug/settings/gdscript/always_track_local_variables = false 🔗
Whether GDScript local variables will be tracked in all builds, including export builds, thus allowing Engine.capture_script_backtraces() to capture them when enabling its include_variables parameter.
Enabling this comes at the cost of roughly 50 bytes of memory per local variable, for every compiled class in the entire project, so can be several MiB in larger projects.
Note: This setting has no effect when running the game from the editor, where GDScript local variables are tracked regardless.
int debug/settings/gdscript/max_call_stack = 1024 🔗
除錯 GDScript 時允許的最大呼叫堆疊。
bool debug/settings/physics_interpolation/enable_warnings = true 🔗
If true, enables warnings which can help pinpoint where nodes are being incorrectly updated, which will result in incorrect interpolation and visual glitches.
When a node is being interpolated, it is essential that the transform is set during Node._physics_process() (during a physics tick) rather than Node._process() (during a frame).
int debug/settings/profiler/max_functions = 16384 🔗
分析時單影格允許的最大函式數量。
int debug/settings/profiler/max_timestamp_query_elements = 256 🔗
Maximum number of timestamp query elements allowed per frame for visual profiling.
bool debug/settings/stdout/print_fps = false 🔗
每秒將影格率列印到標準輸出。
bool debug/settings/stdout/print_gpu_profile = false 🔗
每秒將 GPU 設定檔資訊列印到標準輸出。這包括平均每影格需要 GPU 算繪多長時間，細分為算繪管道的不同步驟，例如 CanvasItems、陰影、輝光等。
bool debug/settings/stdout/verbose_stdout = false 🔗
運作時將更多資訊列印到標準輸出。它顯示諸如記憶體洩漏、正在載入哪些場景和資源等信息。這也可以使用 --verbose 或 -v 命令列參數來啟用，即使在導出的專案中也是如此。另見 OS.is_stdout_verbose() and @GlobalScope.print_verbose()。
bool debug/shader_language/warnings/device_limit_exceeded = true 🔗
設為 true 時，當著色器超出裝置極限時會產生警告。目前唯一偵測的設備極限是 Uniform 緩衝的大小。未來會加入更多裝置極限。
bool debug/shader_language/warnings/enable = true 🔗
如果為 true，則啟用特定的著色器警告（請參閱 debug/shader_language/warnings/* 設定）。如果為 false，則禁用所有著色器警告。
bool debug/shader_language/warnings/float_comparison = true 🔗
When set to true, produces a warning when two floating-point numbers are compared directly with the == operator or the != operator.
bool debug/shader_language/warnings/formatting_error = true 🔗
設為 true 時，當遇到格式錯誤時會產生警告。目前唯一偵測的是空語句。未來可能加入更多格式錯誤。
bool debug/shader_language/warnings/magic_position_write = true 🔗
When set to true, produces a warning when the shader contains POSITION = vec4(vertex, as this was very common code written in Godot 4.2 and earlier that was paired with a QuadMesh to produce a full screen post processes pass. With the switch to reversed z in 4.3, this trick no longer works, as it implicitly relied on the VERTEX.z being 0.
bool debug/shader_language/warnings/treat_warnings_as_errors = false 🔗
設為 true 時，會將警告作為錯誤對待。
bool debug/shader_language/warnings/unused_constant = true 🔗
設為 true 時，當從未使用某個常數時會產生警告。
bool debug/shader_language/warnings/unused_function = true 🔗
設為 true 時，當從未使用某個函式時會產生警告。
bool debug/shader_language/warnings/unused_local_variable = true 🔗
設為 true 時，當從未使用某個區域變數時會產生警告。
bool debug/shader_language/warnings/unused_struct = true 🔗
設為 true 時，當從未使用某個結構體時會產生警告。
bool debug/shader_language/warnings/unused_uniform = true 🔗
設為 true 時，當從未使用某個 uniform 時會產生警告。
bool debug/shader_language/warnings/unused_varying = true 🔗
設為 true 時，當從未使用某個 varying 時會產生警告。
Color debug/shapes/avoidance/2d/agents_radius_color = Color(1, 1, 0, 0.25) 🔗
避障代理半徑的顏色，在“除錯”功能表中啟用“顯示避障”時可見。
bool debug/shapes/avoidance/2d/enable_agents_radius = true 🔗
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示避障”時顯示避障代理的半徑。
bool debug/shapes/avoidance/2d/enable_obstacles_radius = true 🔗
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示避障”時顯示避障障礙物的半徑。
bool debug/shapes/avoidance/2d/enable_obstacles_static = true 🔗
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示避障”時顯示靜態避障障礙物。
Color debug/shapes/avoidance/2d/obstacles_radius_color = Color(1, 0.5, 0, 0.25) 🔗
避障障礙物半徑的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/2d/obstacles_static_edge_pushin_color = Color(1, 0, 0, 1) 🔗
靜態障礙物的頂點纏繞順序會將代理推入時，障礙物邊的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/2d/obstacles_static_edge_pushout_color = Color(1, 1, 0, 1) 🔗
靜態障礙物的頂點纏繞順序會將代理推出時，障礙物邊的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/2d/obstacles_static_face_pushin_color = Color(1, 0, 0, 0) 🔗
靜態障礙物的頂點纏繞順序會將代理推入時，障礙物面的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/2d/obstacles_static_face_pushout_color = Color(1, 1, 0, 0.5) 🔗
靜態障礙物的頂點纏繞順序會將代理推出時，障礙物面的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/3d/agents_radius_color = Color(1, 1, 0, 0.25) 🔗
避障代理半徑的顏色，在“除錯”功能表中啟用“顯示避障”時可見。
bool debug/shapes/avoidance/3d/enable_agents_radius = true 🔗
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示避障”時顯示避障代理的半徑。
bool debug/shapes/avoidance/3d/enable_obstacles_radius = true 🔗
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示避障”時顯示避障障礙物的半徑。
bool debug/shapes/avoidance/3d/enable_obstacles_static = true 🔗
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示避障”時顯示靜態避障障礙物。
Color debug/shapes/avoidance/3d/obstacles_radius_color = Color(1, 0.5, 0, 0.25) 🔗
避障障礙物半徑的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/3d/obstacles_static_edge_pushin_color = Color(1, 0, 0, 1) 🔗
靜態障礙物的頂點纏繞順序會將代理推入時，障礙物邊的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/3d/obstacles_static_edge_pushout_color = Color(1, 1, 0, 1) 🔗
靜態障礙物的頂點纏繞順序會將代理推出時，障礙物邊的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/3d/obstacles_static_face_pushin_color = Color(1, 0, 0, 0) 🔗
靜態障礙物的頂點纏繞順序會將代理推入時，障礙物面的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/avoidance/3d/obstacles_static_face_pushout_color = Color(1, 1, 0, 0.5) 🔗
靜態障礙物的頂點纏繞順序會將代理推出時，障礙物面的顏色，在除錯功能表中啟用“顯示避障”時可見。
Color debug/shapes/collision/contact_color = Color(1, 0.2, 0.1, 0.8) 🔗
碰撞形狀之間接觸點的顏色，在除錯功能表中啟用“顯示碰撞形狀”時可見。
bool debug/shapes/collision/draw_2d_outlines = true 🔗
設定當除錯功能表中的啟用“顯示碰撞形狀”時，2D 物理是否會在遊戲中顯示碰撞輪廓。
int debug/shapes/collision/max_contacts_displayed = 10000 🔗
當在除錯功能表中啟用“顯示碰撞形狀”時，碰撞形狀之間顯示的最大接觸點數。
Color debug/shapes/collision/shape_color = Color(0, 0.6, 0.7, 0.42) 🔗
碰撞形狀的顏色，當在除錯功能表中啟用“顯示碰撞形狀”時可見。
用於顯示已啟用的導覽代理路徑的顏色，代理啟用除錯時生效。
柵格大小（圖元），如果導覽代理啟用了除錯，則會用於算繪該代理的路徑點。
用於顯示導覽區域之間的邊緣連接的顏色，在“除錯”選單中，啟用“可見導覽”時可見。
如果處於啟用狀態，會在導覽代理啟用除錯時顯示其路徑。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時顯示導覽地區之間的邊緣連接。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時顯示導覽網格多邊形的邊緣。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時為每個導覽網格多邊形面使用隨機顏色著色。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時顯示導覽連結的連接。
用於顯示已啟用導覽網格多邊形邊緣的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
用於顯示已禁用導覽網格多邊形邊緣的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
用於顯示已啟用導覽網格多邊形面的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
用於顯示已禁用導覽網格多邊形面的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
導覽連結連接的顏色，在除錯功能表中啟用“顯示導覽”時可見。
被禁用的導覽連結連接的顏色，在除錯功能表中啟用“顯示導覽”時可見。
用於顯示已啟用的導覽代理路徑的顏色，代理啟用除錯時生效。
柵格大小（圖元），如果導覽代理啟用了除錯，則會用於算繪該代理的路徑點。
用於顯示導覽區域之間的邊緣連接的顏色，在“除錯”選單中，啟用“可見導覽”時可見。
如果處於啟用狀態，會在導覽代理啟用除錯時顯示其路徑。
如果處於啟用狀態，會在導覽代理啟用除錯時透過幾何體顯示其路徑。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時顯示導覽地區之間的邊緣連接。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時透過幾何體顯示導覽地區之間的邊緣連接。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時顯示導覽網格多邊形的邊緣。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時透過幾何體顯示導覽網格多邊形的邊緣。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時為每個導覽網格多邊形面使用隨機顏色著色。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時顯示導覽連結的連接。
如果處於啟用狀態，會在啟用“除錯”功能表中的“顯示導覽”時透過幾何體顯示導覽連結的連接。
用於顯示已啟用導覽網格多邊形邊緣的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
用於顯示已禁用導覽網格多邊形邊緣的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
用於顯示已啟用導覽網格多邊形面的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
用於顯示已禁用導覽網格多邊形面的顏色，在“除錯”功能表中啟用“顯示導覽”時可見。
導覽連結連接的顏色，在除錯功能表中啟用“顯示導覽”時可見。
被禁用的導覽連結連接的顏色，在除錯功能表中啟用“顯示導覽”時可見。
Color debug/shapes/paths/geometry_color = Color(0.1, 1, 0.7, 0.4) 🔗
曲線路徑幾何圖形的顏色，在“除錯”功能表中啟用“可見路徑”時可見。
float debug/shapes/paths/geometry_width = 2.0 🔗
曲線路徑幾何圖形的線寬，在“除錯”功能表中啟用“可見路徑”時可見。
String display/display_server/driver 🔗
Sets the driver to be used by the display server. This property can not be edited directly, instead, set the driver using the platform-specific overrides.
String display/display_server/driver.android 🔗
Android override for display/display_server/driver.
String display/display_server/driver.ios 🔗
iOS override for display/display_server/driver.
String display/display_server/driver.linuxbsd 🔗
LinuxBSD override for display/display_server/driver.
String display/display_server/driver.macos 🔗
MacOS override for display/display_server/driver.
String display/display_server/driver.visionos 🔗
visionOS override for display/display_server/driver.
String display/display_server/driver.windows 🔗
Windows override for display/display_server/driver.
String display/mouse_cursor/custom_image = "" 🔗
滑鼠游標的自訂圖像（最大 256×256）。
Vector2 display/mouse_cursor/custom_image_hotspot = Vector2(0, 0) 🔗
自訂滑鼠游標圖像的熱點。
Vector2 display/mouse_cursor/tooltip_position_offset = Vector2(10, 10) 🔗
工具提示相對於滑鼠游標熱點的位置偏移量。
bool display/window/dpi/allow_hidpi = true 🔗
如果為 true ，則允許在 Windows、macOS、Android、iOS 和 Web 上使用 HiDPI 顯示器。如果為 false ，則在 HiDPI 顯示器上將使用該平臺的低 DPI 退回，這樣會導致視窗模糊或圖元化的方式顯示（並且可能導致各種視窗管理問題）。因此，建議你讓專案支援多個解析度，而不是禁用此設定。
注意：這個設定在 Linux 上無效，因為 Linux 不支援 DPI 感知退回。
bool display/window/energy_saving/keep_screen_on = true 🔗
如果為 true，則保持螢幕打開（即使在不活動的情況下），因此螢幕保護程式不會接管。適用於桌面和移動平臺。
bool display/window/frame_pacing/android/enable_frame_pacing = true 🔗
Enable Swappy for stable frame pacing on Android. Highly recommended.
Note: This option will be forced off when using OpenXR.
int display/window/frame_pacing/android/swappy_mode = 2 🔗
Swappy mode to use. The options are:
pipeline_forced_on: Try to honor Engine.max_fps. Pipelining is always on. This is the same behavior as a desktop PC.auto_fps_pipeline_forced_on: Calculate the max FPS automatically. The actual max FPS will be between0and Engine.max_fps. While this sounds convenient, beware that Swappy will often downgrade the max FPS until it finds a value that can be maintained. That means, if your game runs between 40fps and 60fps on a 60hz screen, after some time Swappy will downgrade the max FPS so that the game renders at a perfect 30fps.auto_fps_auto_pipeline: Same asauto_fps_pipeline_forced_on, but if Swappy detects that rendering is very fast (for example it takes less than 8ms to render on a 60hz screen), Swappy will disable pipelining to minimize input latency. This is the default.
Note: If Engine.max_fps is 0, the actual max FPS will be considered to be the screen's refresh rate (often 60hz, 90hz, or 120hz, depending on device model and OS settings).
int display/window/handheld/orientation = 0 🔗
在移動裝置上使用的預設螢幕朝向。可能的取值見 ScreenOrientation。
注意：設為豎屏時，該專案設定不會自動翻轉專案解析度的寬度和高度。你必須設定與之相對應的 display/window/size/viewport_width 和 display/window/size/viewport_height。
bool display/window/hdr/request_hdr_output = false 🔗
If true, HDR output is requested for the main window and the editor. The main window and editor will automatically switch between HDR and SDR if it is moved between screens, screen capabilities change, or system settings are modified. This will internally force Viewport.use_hdr_2d to be enabled on the main Viewport. All other SubViewport of the Window must have their Viewport.use_hdr_2d property enabled to produce HDR output.
Note: This property is only read when the project starts. To change this property at runtime, set Window.hdr_output_requested.
bool display/window/ios/allow_high_refresh_rate = true 🔗
如果為 true，則允許支援高更新率“ProMotion”的 iOS 裝置以每秒 120 影格的速度算繪。
bool display/window/ios/hide_home_indicator = true 🔗
如果為 true，home 指示器將自動隱藏。這只會影響沒有物理 home 鍵的 iOS 裝置。
bool display/window/ios/hide_status_bar = true 🔗
如果為 true，則狀態列將在套用程式運作時隱藏。
bool display/window/ios/suppress_ui_gesture = true 🔗
如果為 true，則需要兩次滑動才能存取使用手勢的 iOS UI。
注意：如果 hide_home_indicator 為 true，則該設定對 home 指示器沒有影響。
bool display/window/per_pixel_transparency/allowed = false 🔗
如果為 true，則允許視窗背景的逐圖元透明度。這樣會影響性能，因此除非需要，否則請將其保留為 false。另見 display/window/size/transparent 和 rendering/viewport/transparent_background。
bool display/window/size/always_on_top = false 🔗
強制主視窗始終置頂。
注意：該設定在 iOS、Android 和 Web 上被忽略。
bool display/window/size/borderless = false 🔗
強制主視窗無邊框。
注意：該設定在 iOS、Android 和 Web 上被忽略。
bool display/window/size/extend_to_title = false 🔗
主視窗的內容會擴充到整個視窗的大小。與無邊框視窗不同的是，視窗的框架仍是完整的，可以用來調整視窗的大小，標題列是透明的，但有最小/最大/關閉按鈕。
注意：該設定只在 macOS 上實作。
Vector2i display/window/size/initial_position = Vector2i(0, 0) 🔗
Main window initial position (in virtual desktop coordinates), this setting is used only if display/window/size/initial_position_type is set to "Absolute" (0).
Note: This setting only affects the exported project, or when the project is run from the command line. In the editor, the value of EditorSettings.run/window_placement/rect_custom_position is used instead.
int display/window/size/initial_position_type = 1 🔗
Main window initial position.
0 - "Absolute", display/window/size/initial_position is used to set window position.
1 - "Primary Screen Center".
3 - "Other Screen Center", display/window/size/initial_screen is used to set the screen.
4 - "Center of Screen With Mouse Pointer".
5 - "Center of Screen With Keyboard Focus".
Note: This setting only affects the exported project, or when the project is run from the command line. In the editor, the value of EditorSettings.run/window_placement/rect is used instead.
int display/window/size/initial_screen = 0 🔗
Main window initial screen, this setting is used only if display/window/size/initial_position_type is set to "Other Screen Center" (2).
Note: This setting only affects the exported project, or when the project is run from the command line. In the editor, the value of EditorSettings.run/window_placement/screen is used instead.
bool display/window/size/maximize_disabled = false 🔗
If true, the main window's maximize button is disabled.
bool display/window/size/minimize_disabled = false 🔗
If true, the main window's minimize button is disabled.
int display/window/size/mode = 0 🔗
Main window mode. See WindowMode for possible values and how each mode behaves.
Note: Game embedding is available only in the "Windowed" mode.
bool display/window/size/no_focus = false 🔗
主視窗無法成為焦點。無焦點的視窗將忽略除滑鼠點擊之外的所有輸入。
bool display/window/size/resizable = true 🔗
If true, allows the window to be resizable by default.
Note: This property is only read when the project starts. To change whether the window is resizable at runtime, set Window.unresizable instead on the root Window, which can be retrieved using get_viewport().get_window(). Window.unresizable takes the opposite value of this setting.
Note: Certain window managers can be configured to ignore the non-resizable status of a window. Do not rely on this setting as a guarantee that the window will never be resizable.
Note: This setting is ignored on iOS.
bool display/window/size/sharp_corners = false 🔗
If true, the main window uses sharp corners by default.
Note: This property is implemented only on Windows (11).
bool display/window/size/transparent = false 🔗
If true, enables a window manager hint that the main window background can be transparent. This does not make the background actually transparent. For the background to be transparent, the root viewport must also be made transparent by enabling rendering/viewport/transparent_background.
Note: To use a transparent splash screen, set application/boot_splash/bg_color to Color(0, 0, 0, 0).
Note: This setting has no effect if display/window/per_pixel_transparency/allowed is set to false.
Note: This setting has no effect on Android as transparency is controlled only via display/window/per_pixel_transparency/allowed.
int display/window/size/viewport_height = 648 🔗
設定遊戲的主視口高度。在桌面平臺上，這也是初始視窗高度，在 2D 編輯器中使用靛藍色的矩形表示。使用 canvas_items 和 viewport 拉伸模式時也會以此作為參考。另見 display/window/size/viewport_width、display/window/size/window_width_override 和 display/window/size/window_height_override。
int display/window/size/viewport_width = 1152 🔗
設定遊戲的主視口寬度。在桌面平臺上，這也是初始視窗寬度，在 2D 編輯器中使用靛藍色的矩形表示。使用 canvas_items 和 viewport 拉伸模式時也會以此作為參考。另見 display/window/size/viewport_height、display/window/size/window_width_override 和 display/window/size/window_height_override。
int display/window/size/window_height_override = 0 🔗
在桌面平臺上，覆蓋遊戲的初始視窗高度。另見 display/window/size/window_width_override、display/window/size/viewport_width 和 display/window/size/viewport_height。
注意：預設情況下，或者當設定為0 時，初始視窗高度為 display/window/size/viewport_height。在 iOS、Android 和 Web 上會忽略這個設定。
int display/window/size/window_width_override = 0 🔗
On desktop platforms, overrides the game's initial window width. See also display/window/size/window_height_override, display/window/size/viewport_width and display/window/size/viewport_height.
Note: By default, or when set to 0, the initial window width is the display/window/size/viewport_width. This setting is ignored on iOS, Android, and Web.
String display/window/stretch/aspect = "keep" 🔗
Defines how the aspect ratio of the base size is preserved when stretching to fit the resolution of the window or screen.
"ignore": Ignore the aspect ratio when stretching the screen. This means that the original resolution will be stretched to exactly fill the screen, even if it's wider or narrower. This may result in non-uniform stretching: things looking wider or taller than designed.
"keep": Keep aspect ratio when stretching the screen. This means that the viewport retains its original size regardless of the screen resolution, and black bars will be added to the top/bottom of the screen ("letterboxing") or the sides ("pillarboxing").
"keep_width": Keep aspect ratio when stretching the screen. If the screen is wider than the base size, black bars are added at the left and right (pillarboxing). But if the screen is taller than the base resolution, the viewport will be grown in the vertical direction (and more content will be visible at the bottom). You can also think of this as "Expand Vertically".
"keep_height": Keep aspect ratio when stretching the screen. If the screen is taller than the base size, black bars are added at the top and bottom (letterboxing). But if the screen is wider than the base resolution, the viewport will be grown in the horizontal direction (and more content will be visible to the right). You can also think of this as "Expand Horizontally".
"expand": Keep aspect ratio when stretching the screen, but keep neither the base width nor height. Depending on the screen aspect ratio, the viewport will either be larger in the horizontal direction (if the screen is wider than the base size) or in the vertical direction (if the screen is taller than the original size). This is the default for projects created starting in Godot 4.7.
String display/window/stretch/mode = "disabled" 🔗
Defines how the base size is stretched to fit the resolution of the window or screen.
"disabled": No stretching happens. One unit in the scene corresponds to one pixel on the screen. In this mode, display/window/stretch/aspect has no effect. Recommended for non-game applications.
"canvas_items": The base size specified in width and height in the project settings is stretched to cover the whole screen (taking display/window/stretch/aspect into account). This means that everything is rendered directly at the target resolution. 3D is unaffected, while in 2D, there is no longer a 1:1 correspondence between sprite pixels and screen pixels, which may result in scaling artifacts. Recommended for most games that don't use a pixel art aesthetic, although it is possible to use this stretch mode for pixel art games too (especially in 3D). This is the default for projects created starting in Godot 4.7.
"viewport": The size of the root Viewport is set precisely to the base size specified in the Project Settings' Display section. The scene is rendered to this viewport first. Finally, this viewport is scaled to fit the screen (taking display/window/stretch/aspect into account). Recommended for games that use a pixel art aesthetic.
float display/window/stretch/scale = 1.0 🔗
用於2D 元素的比例因子乘數。這將乘以由display/window/stretch/mode 確定的最終比例因子。如果使用Disabled 拉伸模式，此比例因子按原樣應用。可以對其進行調整以使UI 在某些顯示器上更易於閱讀。
String display/window/stretch/scale_mode = "fractional" 🔗
The policy to use to determine the final scale factor for 2D elements. This affects how display/window/stretch/scale is applied, in addition to the automatic scale factor determined by display/window/stretch/mode.
"fractional": The scale factor will not be modified.
"integer": The scale factor will be floored to an integer value, which means that the screen size will always be an integer multiple of the base viewport size. This provides a crisp pixel art appearance.
Note: When using integer scaling with a stretch mode, resizing the window to be smaller than the base viewport size will clip the contents. Consider preventing that by setting Window.min_size to the same value as the base viewport size defined in display/window/size/viewport_width and display/window/size/viewport_height.
bool display/window/subwindows/embed_subwindows = true 🔗
If true, subwindows are embedded in the main window (this is also called single-window mode). Single-window mode can be faster as it does not need to create a separate window for every popup and tooltip, which can be a slow operation depending on the operating system and rendering method in use.
If false, subwindows are created as separate windows (this is also called multi-window mode). This allows them to be moved outside the main window and use native operating system window decorations.
This is equivalent to EditorSettings.interface/editor/display/single_window_mode in the editor.
int display/window/vsync/vsync_mode = 1 🔗
Sets the V-Sync mode for the main game window. The editor's own V-Sync mode can be set using EditorSettings.interface/editor/display/vsync_mode.
See VSyncMode for possible values and how they affect the behavior of your application.
Depending on the platform and rendering method, the engine will fall back to Enabled if the desired mode is not supported.
V-Sync can be disabled on the command line using the --disable-vsync command line argument.
Note: The Adaptive and Mailbox V-Sync modes are only supported in the Forward+ and Mobile rendering methods, not Compatibility.
Note: This property is only read when the project starts. To change the V-Sync mode at runtime, call DisplayServer.window_set_vsync_mode() instead.
String dotnet/project/assembly_name = "" 🔗
.NET 程式集的名稱。這個名稱會被用作 .csproj 和 .sln 檔的名稱。預設情況下，它被設定為專案的名稱（application/config/name），這樣將來進行修改時就不會影響 .NET 程式集。
int dotnet/project/assembly_reload_attempts = 3 🔗
重新建構 .NET 程式集前嘗試重新載入程式集的次數。實際也是等待腳本程式集解除安裝完成的超時秒數。
String dotnet/project/solution_directory = "" 🔗
包含 .sln 檔的目錄。預設情況下，.sln 檔在專案目錄的根部，和 project.godot 和 .csproj 檔案在同一個目錄。
改變這個值可以設定包含多個 .csproj 的多專案方案。請記住，Godot 專案被認為是工作空間中的 C# 專案之一，根目錄應該包含 project.godot 和.csproj。
bool editor/export/convert_text_resources_to_binary = true 🔗
If true, text resource (tres) and text scene (tscn) files are converted to their corresponding binary format on export. This decreases file sizes and speeds up loading slightly.
Note: Because a resource's file extension may change in an exported project, it is heavily recommended to use @GDScript.load() or ResourceLoader instead of FileAccess to load resources dynamically.
Note: The project settings file (project.godot) will always be converted to binary on export, regardless of this setting.
int editor/import/atlas_max_width = 2048 🔗
The maximum width to use when importing textures as an atlas. The value will be rounded to the nearest power of two when used. Use this to prevent imported textures from growing too large in the other direction.
bool editor/import/reimport_missing_imported_files = true 🔗
There is currently no description for this property. Please help us by contributing one!
bool editor/import/use_multiple_threads = true 🔗
如果為 true，則會多執行緒執行資源的匯入。
int editor/movie_writer/audio_bit_depth = 16 🔗
Number of bits per audio sample written to the .avi file. Only 16 and 32-bit are supported.
bool editor/movie_writer/disable_vsync = false 🔗
如果為 true，則在寫入電影時會請求禁用垂直同步（類似於將 display/window/vsync/vsync_mode 設定為 Disabled）。如果硬體的算繪速度足夠快，那麼就可以加速影片的寫入，以高於顯示器更新率的畫面播放速率算繪、編碼和保存影片。
注意：如果作業系統或圖形驅動程式強制垂直同步，則套用程式無法禁用，editor/movie_writer/disable_vsync 無效。
int editor/movie_writer/fps = 60 🔗
輸出電影時，在影片中每秒記錄的影格數。模擬速度將調整為始終與指定的影格數相配對，這意味著引擎在較高的 editor/movie_writer/fps 值下會顯得運作較慢。部分 FPS 值將需要你調整 editor/movie_writer/mix_rate，防止音訊隨著時間的推移而出現不同步。
可以在命令列中使用 --fixed-fps <fps> 命令列參數手動指定。
int editor/movie_writer/mix_rate = 48000 🔗
寫入影片時，錄製的音訊所使用的音訊混合取樣速率（單位為 Hz）。可以和 audio/driver/mix_rate 不同，但這個值必須能夠被 editor/movie_writer/fps 整除，從而防止音訊可能逐漸不同步的問題。
String editor/movie_writer/movie_file = "" 🔗
The output path for the movie. The file extension determines the MovieWriter that will be used.
Godot has 3 built-in MovieWriters:
OGV container with Theora for video and Vorbis for audio (
.ogvfile extension). Lossy compression, medium file sizes, fast encoding. The lossy compression quality can be adjusted by changing editor/movie_writer/video_quality and editor/movie_writer/ogv/audio_quality. The resulting file can be viewed in Godot with VideoStreamPlayer and most video players, but not web browsers as they don't support Theora.AVI container with MJPEG for video and uncompressed audio (
.avifile extension). Lossy compression, medium file sizes, fast encoding. The lossy compression quality can be adjusted by changing editor/movie_writer/video_quality. The resulting file can be viewed in most video players, but it must be converted to another format for viewing on the web or by Godot with VideoStreamPlayer. MJPEG does not support transparency. AVI output is currently limited to a file of 4 GB in size at most.PNG image sequence for video and WAV for audio (
.pngfile extension). Lossless compression, large file sizes, slow encoding. Designed to be encoded to a video file with another tool such as FFmpeg after recording. Transparency is currently not supported, even if the root viewport is set to be transparent.
If you need to encode to a different format or pipe a stream through third-party software, you can extend this MovieWriter class to create your own movie writers.
When using PNG output, the frame number will be appended at the end of the file name. It starts from 0 and is padded with 8 digits to ensure correct sorting and easier processing. For example, if the output path is /tmp/hello.png, the first two frames will be /tmp/hello00000000.png and /tmp/hello00000001.png. The audio will be saved at /tmp/hello.wav.
float editor/movie_writer/ogv/audio_quality = 0.5 🔗
The audio encoding quality to use when writing Vorbis audio to a file, between -0.1 and 1.0 (inclusive). Higher quality values result in better-sounding output at the cost of larger file sizes. Even at quality 1.0, compression remains lossy.
Note: This does not affect video quality, which is controlled by editor/movie_writer/video_quality instead.
int editor/movie_writer/ogv/encoding_speed = 4 🔗
The tradeoff between encoding speed and compression efficiency. Speed 1 is the slowest but provides the best compression. Speed 4 is the fastest but provides the worst compression. Video quality is generally not affected significantly by this setting.
int editor/movie_writer/ogv/keyframe_interval = 64 🔗
Forces keyframes at the specified interval (in frame count). Higher values can improve compression up to a certain level at the expense of higher latency when seeking.
int editor/movie_writer/speaker_mode = 0 🔗
保存電影時，錄製的音訊中所使用的揚聲器模式。可能的值見 SpeakerMode。
float editor/movie_writer/video_quality = 0.75 🔗
The video encoding quality to use when writing a Theora or AVI (MJPEG) video to a file, between 0.0 and 1.0 (inclusive). Higher quality values result in better-looking output at the cost of larger file sizes. Recommended quality values are between 0.75 and 0.9. Even at quality 1.0, compression remains lossy.
String editor/naming/default_signal_callback_name = "_on_{node_name}_{signal_name}" 🔗
訊號回呼函式名稱的預設格式（在訊號連接對話方塊中使用）。可以使用以下替換：{NodeName}、{nodeName}、{node_name}、{SignalName}、{signalName}、{signal_name}。
String editor/naming/default_signal_callback_to_self_name = "_on_{signal_name}" 🔗
訊號回呼函式名稱的預設格式，用於連接到與發出訊號相同節點的情況（在訊號連接對話方塊中使用）。可以使用以下替換：{NodeName}、{nodeName}、{node_name}、{SignalName}、{signalName}、{signal_name}。
int editor/naming/node_name_casing = 0 🔗
When creating node names automatically, set the type of casing to use in this project. This is mostly an editor setting.
int editor/naming/node_name_num_separator = 0 🔗
用什麼來分隔節點名稱和編號。這主要是一個編輯器的設定。
int editor/naming/scene_name_casing = 2 🔗
When generating scene file names from scene root node, set the type of casing to use in this project. This is mostly an editor setting.
int editor/naming/script_name_casing = 0 🔗
When generating script file names from the selected node, set the type of casing to use in this project. This is mostly an editor setting.
String editor/run/main_run_args = "" 🔗
The command-line arguments to append to Godot's own command line when running the project. This doesn't affect the editor itself.
It is possible to make another executable run Godot by using the %command% placeholder. The placeholder will be replaced with Godot's own command line. Program-specific arguments should be placed before the placeholder, whereas Godot-specific arguments should be placed after the placeholder.
For example, this can be used to force the project to run on the dedicated GPU in an NVIDIA Optimus system on Linux:
prime-run %command%
PackedStringArray editor/script/search_in_file_extensions 🔗
腳本編輯器的“在檔中搜尋”功能中包含的基於文字的檔副檔名。你可以新增例如 tscn，如果你也想解析你的場景檔，特別是如果你使用的是在場景檔中序列化的內建腳本。
Note: The returned array is copied and any changes to it will not update the original property value. See PackedStringArray for more details.
String editor/script/templates_search_path = "res://script_templates" 🔗
對於專案特定的腳本範本的搜索路徑。Godot 將在編輯器的特定路徑和此專案的路徑中搜索。
bool editor/version_control/autoload_on_startup = false 🔗
There is currently no description for this property. Please help us by contributing one!
String editor/version_control/plugin_name = "" 🔗
There is currently no description for this property. Please help us by contributing one!
bool filesystem/import/blender/enabled = true 🔗
If true, Blender 3D scene files with the .blend extension will be imported by converting them to glTF 2.0.
This requires configuring a path to a Blender executable in the EditorSettings.filesystem/import/blender/blender_path setting. Blender 3.0 or later is required.
bool filesystem/import/blender/enabled.android = false 🔗
filesystem/import/blender/enabled 在 Android 上的覆蓋項，Godot 無法輕易存取到 Blender。
bool filesystem/import/blender/enabled.web = false 🔗
filesystem/import/blender/enabled 在 Web 上的覆蓋項，Godot 無法輕易存取到 Blender。
bool filesystem/import/fbx2gltf/enabled = true 🔗
If true, Autodesk FBX 3D scene files with the .fbx extension will be imported by converting them to glTF 2.0.
This requires configuring a path to an FBX2glTF executable in the editor settings at EditorSettings.filesystem/import/fbx/fbx2gltf_path.
bool filesystem/import/fbx2gltf/enabled.android = false 🔗
Override for filesystem/import/fbx2gltf/enabled on Android where FBX2glTF can't easily be accessed from Godot.
bool filesystem/import/fbx2gltf/enabled.web = false 🔗
Override for filesystem/import/fbx2gltf/enabled on the Web where FBX2glTF can't easily be accessed from Godot.
int gui/common/default_scroll_deadzone = 0 🔗
ScrollContainer.scroll_deadzone，它將用於所有 ScrollContainer，除非fug。
int gui/common/drag_threshold = 10 🔗
The minimum distance the mouse cursor must move while pressed before a drag operation begins in the default viewport. For custom viewports see Viewport.gui_drag_threshold.
int gui/common/show_focus_state_on_pointer_event = 1 🔗
Determines whether a Control should visually indicate focus when that focus is gained using a mouse or touch input.
Never (
0) show the focused state for mouse/touch input.Text Input Controls (
1) show the focused state even if that focus was gained via mouse/touch input (similar to browser behavior).Always (
2) show the focused state, even if that focus was gained via mouse/touch input.
bool gui/common/snap_controls_to_pixels = true 🔗
如果為 true，則將 Control 節點的頂點吸附到最近的圖元，確保即便相機發生移動或縮放也能夠保持銳利。
int gui/common/swap_cancel_ok = 0 🔗
How to position the Cancel and OK buttons in the project's AcceptDialog windows. Different platforms have different conventions for this, which can be overridden through this setting.
Auto follows the platform convention: OK first on Windows, KDE, and LXQt; Cancel first on macOS and other Linux desktop environments.
Cancel First forces the Cancel/OK ordering.
OK First forces the OK/Cancel ordering.
To check if these buttons are swapped at runtime, use DisplayServer.get_swap_cancel_ok().
Note: This doesn't affect native dialogs such as the ones spawned by DisplayServer.dialog_show().
int gui/common/text_edit_undo_stack_max_size = 1024 🔗
TextEdit 欄位的最大撤銷/重做歷史大小。
bool gui/fonts/dynamic_fonts/use_oversampling = true 🔗
If set to true and display/window/stretch/mode is set to "canvas_items", font and DPITexture oversampling is enabled in the main window. Use Viewport.oversampling to control oversampling in other viewports and windows.
String gui/theme/custom = "" 🔗
Theme 資源檔的路徑，用於自訂專案主題（副檔名為 .theme 或通用的 .tres/.res）。
String gui/theme/custom_font = "" 🔗
自訂 Font 資源的路徑，用作專案中所有 GUI 元素的預設字形。
int gui/theme/default_font_antialiasing = 1 🔗
預設專案字形的字形抗鋸齒模式。請參閱FontFile.antialiasing。
注意： 此設定不會影響專案中使用的自訂Font。請使用 Import 停靠列（請參閱ResourceImporterDynamicFont.antialiasing）。
bool gui/theme/default_font_generate_mipmaps = false 🔗
如果設定為 true，則預設字形將生成 mipmap。這樣可以防止文字在 Control 被按比例縮小或從遠距離查看 Label3D 時看起來有顆粒感（如果 Label3D.texture_filter 設定為顯示 mipmap 的模式）。
啟用 gui/theme/default_font_generate_mipmaps 會增加字形生成時間和內存使用量。請只在你確實需要時才啟用此設定。
注意：此設定不會影響專案中使用的自訂 Font。
int gui/theme/default_font_hinting = 1 🔗
預設專案字型的字型提示模式。請參閱FontFile.hinting。
注意： 此設定不會影響專案中使用的自訂Font。請使用 Import 停靠列（請參閱ResourceImporterDynamicFont.hinting）。
bool gui/theme/default_font_multichannel_signed_distance_field = false 🔗
如果設定為 true，預設字形將使用多通道帶符號距離場（MSDF），任何尺寸都能夠進行清晰的算繪。由於這種方法不需要在每次字形大小更改時都對字形進行光柵化，因此可以即時調整字形大小，不會造成任何性能損失。對於按比例縮小的 Control（或從遠距離查看的 Label3D），文字也不會看起來有顆粒感。
MSDF 字形算繪可以與 gui/theme/default_font_generate_mipmaps 結合使用，從而進一步提高縮小時的字形算繪品質。
注意：此設定不會影響專案中使用的自訂 Font。
int gui/theme/default_font_subpixel_positioning = 1 🔗
預設專案字型的字型字字形子像素定位模式。請參閱FontFile.subpixel_positioning。
注意： 此設定不會影響專案中使用的自訂Font。請使用 Import 停靠列（請參閱ResourceImporterDynamicFont.subpixel_positioning）。
float gui/theme/default_theme_scale = 1.0 🔗
The default scale factor for Controls, when not overridden by a Theme.
Note: This property is only read when the project starts. To change the default theme scale at runtime, set ThemeDB.fallback_base_scale instead. However, to adjust the scale of all 2D elements at runtime, it's preferable to use Window.content_scale_factor on the root Window node instead (as this also affects overridden Themes). See Multiple resolutions in the documentation for details.
int gui/theme/lcd_subpixel_layout = 1 🔗
LCD 次圖元佈局，用於字形抗鋸齒。見 FontLCDSubpixelLayout。
float gui/timers/button_shortcut_feedback_highlight_time = 0.2 🔗
When BaseButton.shortcut_feedback is enabled, this is the time the BaseButton will remain highlighted after a shortcut. This duration is not affected by Engine.time_scale.
int gui/timers/incremental_search_max_interval_msec = 2000 🔗
在 Tree、ItemList 等控制項中為差異量搜索設定計時器（單位為毫秒）。
float gui/timers/text_edit_idle_detect_sec = 3 🔗
偵測 TextEdit 空閒的計時器（單位為秒）。
float gui/timers/tooltip_delay_sec = 0.5 🔗
工具提示的預設延遲（單位為秒）。
float gui/timers/tooltip_delay_sec.editor_hint = 0.5 🔗
Delay for tooltips in the editor.
Dictionary input/ui_accept 🔗
預設 InputEventAction，用於確認焦點按鈕、功能表或列表項，或用於驗證輸入。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_accessibility_drag_and_drop 🔗
Default InputEventAction to start or end a drag-and-drop operation without using mouse.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_cancel 🔗
預設 InputEventAction，用於放棄模態或掛起的輸入。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_close_dialog 🔗
Default InputEventAction to close a dialog window.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_close_dialog.macos 🔗
macOS specific override for the shortcut to close a dialog window.
Dictionary input/ui_colorpicker_delete_preset 🔗
Default InputEventAction to delete a color preset in a ColorPicker.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_copy 🔗
預設 InputEventAction，用於將選中內容複製到剪貼板。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_cut 🔗
預設 InputEventAction，用於將選中內容剪切到剪貼板。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_down 🔗
預設 InputEventAction，用於在 UI 中向下移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_end 🔗
預設 InputEventAction，用於前往 Control 的末尾（例如 ItemList 和 Tree 中的最後一個專案），與常見桌面 UI 系統中 @GlobalScope.KEY_END 的 行為一致。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_filedialog_delete 🔗
Default InputEventAction to delete the selected file in a FileDialog.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_filedialog_find 🔗
Default InputEventAction to open file filter in a FileDialog.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_filedialog_focus_path 🔗
Default InputEventAction to focus path edit field in a FileDialog.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_filedialog_focus_path.macos 🔗
macOS specific override for the shortcut to focus path edit field in FileDialog.
Dictionary input/ui_filedialog_refresh 🔗
預設 InputEventAction，用於更新 FileDialog 中目前的目錄的內容。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
預設 InputEventAction，用於開關 FileDialog 中隱藏檔和目錄的顯示。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_filedialog_up_one_level 🔗
預設 InputEventAction，用於在 FileDialog 前往上一級目錄。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_focus_mode 🔗
Default InputEventAction to switch TextEdit input/ui_text_indent between moving keyboard focus to the next Control in the scene and inputting a Tab character.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_focus_next 🔗
預設 InputEventAction，用於聚焦場景中的下一個 Control。聚焦行為可以通過 Control.focus_next 配置。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_focus_prev 🔗
預設 InputEventAction，用於聚焦場景中的上一個 Control。聚焦行為可以通過 Control.focus_previous 配置。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_graph_delete 🔗
預設 InputEventAction，用於刪除 GraphEdit 中的某個 GraphNode。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_graph_duplicate 🔗
預設 InputEventAction，用於製作 GraphEdit 中某個 GraphNode 的副本。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_graph_follow_left 🔗
Default InputEventAction to follow a GraphNode input port connection.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_graph_follow_left.macos 🔗
macOS specific override for the shortcut to follow a GraphNode input port connection.
Dictionary input/ui_graph_follow_right 🔗
Default InputEventAction to follow a GraphNode output port connection.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_graph_follow_right.macos 🔗
macOS specific override for the shortcut to follow a GraphNode output port connection.
Dictionary input/ui_home 🔗
預設 InputEventAction，用於前往 Control 的開頭（例如 ItemList 和 Tree 中的第一個專案），與常見桌面 UI 系統中 @GlobalScope.KEY_HOME 的 行為一致。
注意：預設的 ui_* 動作不能被刪除，因為它們是幾個 Control 的內部邏輯所必需的。然而，分配給動作的事件可以被修改。
Dictionary input/ui_left 🔗
預設 InputEventAction，用於在 UI 中向左移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
預設 InputEventAction，用於在文字欄位中打開本文功能表。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_page_down 🔗
預設 InputEventAction，用於在 Control（例如 ItemList 和 Tree）中向下翻頁，與常見桌面 UI 系統中 @GlobalScope.KEY_PAGEDOWN 的 行為一致。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_page_up 🔗
預設 InputEventAction，用於在 Control（例如 ItemList 和 Tree）中向上翻頁，與常見桌面 UI 系統中 @GlobalScope.KEY_PAGEUP 的 行為一致。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_paste 🔗
預設 InputEventAction，用於從剪貼板貼上內容。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_redo 🔗
預設 InputEventAction，用於重做被撤銷的動作。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_right 🔗
預設 InputEventAction，用於在 UI 中向右移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_select 🔗
預設 InputEventAction，用於在 Control（例如 ItemList 和 Tree）中選中專案。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_swap_input_direction 🔗
Default InputEventAction to swap input direction, i.e. change between left-to-right to right-to-left modes. Affects text-editing controls (LineEdit, TextEdit).
Dictionary input/ui_text_add_selection_for_next_occurrence 🔗
如果文字欄位的最近一個游標處存在選中的文字，則會搜索所選內容下一次出現的位元置，在那個位置新增一個游標，然後選中該處的內容。
如果文字欄位的最近一個游標處沒有選中文字，則會選中目前游標下的單詞。
該動作可以連續執行，選中最近一個游標所選文字的所有出現位置，所有現存游標均適用。
視口會根據最近新新增的文字游標進行調整。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_backspace 🔗
預設 InputEventAction，用於刪除文字游標前的字元。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_backspace_all_to_left 🔗
預設 InputEventAction，用於刪除文字游標前的所有文字。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_backspace_all_to_left.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應刪除文字游標前所有文字的快捷鍵。
Dictionary input/ui_text_backspace_word 🔗
預設 InputEventAction，用於刪除游標前、空白或標點字元後的所有字元。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_backspace_word.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應刪除單詞的快捷鍵。
Dictionary input/ui_text_caret_add_above 🔗
預設 InputEventAction，用於在文字中各個游標上方再新增一個游標。
Dictionary input/ui_text_caret_add_above.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應在各個游標上方再新增一個游標的快捷鍵。
Dictionary input/ui_text_caret_add_below 🔗
預設 InputEventAction，用於在文字中各個游標下方再新增一個游標。
Dictionary input/ui_text_caret_add_below.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應在各個游標下方再新增一個游標的快捷鍵。
Dictionary input/ui_text_caret_document_end 🔗
預設 InputEventAction，用於將文字游標移動到文字的末尾。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_document_end.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應將文字游標移動到文字末尾的快捷鍵。
Dictionary input/ui_text_caret_document_start 🔗
預設 InputEventAction，用於將文字游標移動到文字的開頭。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_document_start.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應將文字游標移動到文字開頭的快捷鍵。
Dictionary input/ui_text_caret_down 🔗
預設 InputEventAction，用於將文字游標向下移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_left 🔗
預設 InputEventAction，用於將文字游標向左移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_line_end 🔗
預設 InputEventAction，用於將文字游標移動到該行的末尾。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_line_end.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應將文字游標移動到改行末尾的快捷鍵。
Dictionary input/ui_text_caret_line_start 🔗
預設 InputEventAction，用於將文字游標移動到該行的開頭。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_line_start.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應將文字游標移動到改行開頭的快捷鍵。
Dictionary input/ui_text_caret_page_down 🔗
預設 InputEventAction，用於將文字游標向下移動一頁。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_page_up 🔗
預設 InputEventAction，用於將文字游標向上移動一頁。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_right 🔗
預設 InputEventAction，用於將文字游標向右移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_up 🔗
預設 InputEventAction，用於將文字游標向上移動。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_word_left 🔗
預設 InputEventAction，用於將文字游標向左移動到最近的空白或標點字元。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_word_left.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應將文字游標向後移動一個單詞的快捷鍵。
Dictionary input/ui_text_caret_word_right 🔗
預設 InputEventAction，用於將文字游標向右移動到最近的空白或標點字元。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_caret_word_right.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應將文字游標向前移動一個單詞的快捷鍵。
Dictionary input/ui_text_clear_carets_and_selection 🔗
如果處於活動狀態並且有選區的游標只有一個，則清除該選區。
如果有多個游標處於活動狀態，請移除次要游標並清除其選區。
注意：預設的 ui_* 動作無法移除，因為它們對於部分 Control 的內部邏輯是必需的。但是，可以修改分配給該動作的事件。
Dictionary input/ui_text_completion_accept 🔗
Default InputEventAction to accept an autocompletion hint.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_text_completion_query 🔗
Default InputEventAction to request autocompletion.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_text_completion_replace 🔗
Default InputEventAction to accept an autocompletion hint, replacing existing text.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_text_dedent 🔗
預設 InputEventAction，用於取消文字縮進。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_delete 🔗
預設 InputEventAction，用於刪除文字游標後的字元。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_delete_all_to_right 🔗
預設 InputEventAction，用於刪除文字游標後的所有文字。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_delete_all_to_right.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應刪除文字游標後所有文字的快捷鍵。
Dictionary input/ui_text_delete_word 🔗
預設 InputEventAction，用於刪除游標後、空白或標點字元前的所有字元。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_delete_word.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應刪除文字游標前的單詞的快捷鍵。
Dictionary input/ui_text_indent 🔗
預設 InputEventAction，用於縮進目前行。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_newline 🔗
預設 InputEventAction，用於在文字游標處插入換行。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_newline_above 🔗
預設 InputEventAction，用於在目前行之前插入一個空行。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_newline_blank 🔗
預設 InputEventAction，用於在目前行之後插入一個空行。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_scroll_down 🔗
預設 InputEventAction，用於向下滾動一行文字。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_scroll_down.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應向下滾動一行的快捷鍵。
Dictionary input/ui_text_scroll_up 🔗
預設 InputEventAction，用於向上滾動一行文字。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_scroll_up.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應向上滾動一行的快捷鍵。
Dictionary input/ui_text_select_all 🔗
預設 InputEventAction，用於全選文字。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_select_word_under_caret 🔗
如果文字方塊中目前沒有選中文字，則選中文字游標下的單詞。如果有目前選中的文字，則取消選中。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_select_word_under_caret.macos 🔗
針對 macOS 的快捷鍵覆蓋項，對應選中游標處單詞的快捷鍵。
Dictionary input/ui_text_skip_selection_for_next_occurrence 🔗
If no selection is currently active with the last caret in text fields, searches for the next occurrence of the word currently under the caret and moves the caret to the next occurrence. The action can be performed sequentially for other occurrences of the word under the last caret.
If a selection is currently active with the last caret in text fields, searches for the next occurrence of the selection, adds a caret, selects the next occurrence then deselects the previous selection and its associated caret. The action can be performed sequentially for other occurrences of the selection of the last caret.
The viewport is adjusted to the latest newly added caret.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_text_submit 🔗
預設 InputEventAction，用於提交文字欄位。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_text_toggle_insert_mode 🔗
預設 InputEventAction，用於開關文字欄位的插入模式。在插入模式中，插入新的文字會覆蓋游標後的字元，除非下一個字元是分行符號。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_undo 🔗
預設 InputEventAction，用於撤銷最近的動作。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
Dictionary input/ui_unicode_start 🔗
Default InputEventAction to start Unicode character hexadecimal code input in a text field.
Note: Default ui_* actions cannot be removed as they are necessary for the internal logic of several Controls. The events assigned to the action can however be modified.
Dictionary input/ui_up 🔗
預設在 UI 中向上移動InputEventAction。
注意：預設的 ui_* 動作是部分 Control 的內部邏輯所必需的，無法刪除。但是可以修改分配給該動作的事件。
bool input_devices/buffering/agile_event_flushing = false 🔗
如果為 true，按鍵/觸摸/操縱桿事件將在每個空閒影格和物理影格之前刷新。
如果為 false，該類事件將在引擎反覆運算之間的每個程序影格中僅更新一次。
啟用該功能可以極大地提高對輸入的回應能力，特別是在每個可見（程序）影格需要運作多個物理影格的裝置中，因為它們無法以目標畫面播放速率運作。
注意：目前只在 Android 上實作。
bool input_devices/compatibility/legacy_just_pressed_behavior = false 🔗
如果為 true，則 Input.is_action_just_pressed() 和 Input.is_action_just_released() 只有在動作仍然處於相應狀態時返回 true，也就是說，會忽略在同一影格按下並釋放的動作。
如果為 false，則不會丟失任何輸入。
注意：幾乎所有情況下都應該優先設定為 false。舊有行為的用途是支援老專案，這樣不需要修改腳本就可以使用以前的邏輯。
bool input_devices/joypads/ignore_joypad_on_unfocused_application = false 🔗
If true, joypad input (including motion sensors) and LED light changes will be ignored and joypad vibration will be stopped when the application is not focused.
String input_devices/pen_tablet/driver 🔗
Specifies the tablet driver to use. If left empty, the default driver will be used.
Note: The driver in use can be overridden at runtime via the --tablet-driver command line argument.
Note: Use DisplayServer.tablet_set_current_driver() to switch tablet driver in runtime.
String input_devices/pen_tablet/driver.windows 🔗
Override for input_devices/pen_tablet/driver on Windows. Supported values are:
auto(default), useswintabif Windows Ink is disabled in the Wacom Tablet Properties or system settings,wininkotherwise.winink, uses Windows native "Windows Ink" driver.wintab, uses Wacom "WinTab" driver.dummy, tablet input is disabled.
bool input_devices/pointing/android/disable_scroll_deadzone = false 🔗
If true, disables the scroll deadzone on Android, allowing even very small scroll movements to be registered. This may increase scroll sensitivity but can also lead to unintended scrolling from slight finger movements.
bool input_devices/pointing/android/enable_long_press_as_right_click = false 🔗
If true, long press events on an Android touchscreen are transformed into right click events.
bool input_devices/pointing/android/enable_pan_and_scale_gestures = false 🔗
If true, multi-touch pan and scale gestures are enabled on Android devices.
bool input_devices/pointing/android/override_volume_buttons = false 🔗
If true, system volume changes are disabled when the buttons are used within the app.
int input_devices/pointing/android/rotary_input_scroll_axis = 1 🔗
On Wear OS devices, defines which axis of the mouse wheel rotary input is mapped to. This rotary input is usually performed by rotating the physical or virtual (touch-based) bezel on a smartwatch.
bool input_devices/pointing/emulate_mouse_from_touch = true 🔗
如果為 true，則在點擊或滑動觸控式螢幕時發送滑鼠輸入事件。
bool input_devices/pointing/emulate_touch_from_mouse = false 🔗
如果為 true，則在點擊或拖動滑鼠時發送觸摸輸入事件。
bool input_devices/sensors/enable_accelerometer = false 🔗
If true, the accelerometer sensor is enabled and Input.get_accelerometer() returns valid data.
bool input_devices/sensors/enable_gravity = false 🔗
If true, the gravity sensor is enabled and Input.get_gravity() returns valid data.
bool input_devices/sensors/enable_gyroscope = false 🔗
If true, the gyroscope sensor is enabled and Input.get_gyroscope() returns valid data.
bool input_devices/sensors/enable_magnetometer = false 🔗
If true, the magnetometer sensor is enabled and Input.get_magnetometer() returns valid data.
String internationalization/locale/fallback = "en" 🔗
The locale to fall back to if a translation isn't available in a given language. If left empty, en (English) will be used.
Note: Not to be confused with TextServerFallback.
bool internationalization/locale/include_text_server_data = false 🔗
If true, text server break iteration rule sets, dictionaries and other optional data are included in the exported project.
Note: "ICU / HarfBuzz / Graphite" text server data includes dictionaries for Burmese, Chinese, Japanese, Khmer, Lao and Thai as well as Unicode Standard Annex #29 and Unicode Standard Annex #14 word and line breaking rules. Data is about 4 MB large.
Note: TextServerFallback does not use additional data.
int internationalization/locale/line_breaking_strictness = 0 🔗
Default strictness of line-breaking rules. Can be overridden by adding @lb={auto,loose,normal,strict} to the language code.
Auto (
0) - strictness is based on the length of the line.Loose (
1) - the least restrictive set of line-breaking rules. Typically used for short lines.Normal (
2) - the most common set of line-breaking rules.Strict (
3) - the most stringent set of line-breaking rules.
See Line Breaking Strictness: the line-break property for more info.
String internationalization/locale/test = "" 🔗
If non-empty, this locale will be used instead of the automatically detected system locale.
Note: This setting also applies to the exported project. To only affect testing within the editor, override this setting with an editor feature tag for localization testing purposes.
bool internationalization/pseudolocalization/double_vowels = false 🔗
進行偽當地語系化時，重複字串中的母音字母，類比由於當地語系化而造成的文字加長。
float internationalization/pseudolocalization/expansion_ratio = 0.0 🔗
進行偽當地語系化時的擴充比率。取 0.3 能夠適應大部分實際使用，會將每個字串加長 30%。
bool internationalization/pseudolocalization/fake_bidi = false 🔗
If true, emulate bidirectional (right-to-left) text when pseudolocalization is enabled. This can be used to spot issues with RTL layout and UI mirroring that will crop up if the project is localized to RTL languages such as Arabic or Hebrew. See also internationalization/rendering/force_right_to_left_layout_direction.
bool internationalization/pseudolocalization/override = false 🔗
將字串中的所有字元替換為 *。用於搜尋不可當地語系化的字串。
String internationalization/pseudolocalization/prefix = "[" 🔗
將被前置到偽當地語系化字串的前綴。
bool internationalization/pseudolocalization/replace_with_accents = true 🔗
進行偽當地語系化時，將所有字元替換為其重音變體。
bool internationalization/pseudolocalization/skip_placeholders = true 🔗
進行偽當地語系化時，跳過 %s 或 %f 等用於字串格式的占位符。有助於識別需要額外的控制字元才能正確顯示的字串。
String internationalization/pseudolocalization/suffix = "]" 🔗
將被追加到偽當地語系化字串的後綴。
bool internationalization/pseudolocalization/use_pseudolocalization = false 🔗
如果為 true，則為專案啟用偽當地語系化。這可用於發現，一旦專案被本地化為字串比來源語言更長的語言時，可能出現的無法翻譯的字串或佈局問題。
注意：只有在專案啟動時該屬性才會被讀取。要在運作時切換偽當地語系化，請改用 TranslationServer.pseudolocalization_enabled。
bool internationalization/rendering/force_right_to_left_layout_direction = false 🔗
Force layout direction and text writing direction to RTL for all controls, even if the current locale is intended to use a left-to-right layout and text writing direction. This should be enabled for testing purposes only. See also internationalization/pseudolocalization/fake_bidi.
bool internationalization/rendering/root_node_auto_translate = true 🔗
If true, root node will use Node.AUTO_TRANSLATE_MODE_ALWAYS, otherwise Node.AUTO_TRANSLATE_MODE_DISABLED will be used.
Note: This property is only read when the project starts. To change the auto translate mode at runtime, set Node.auto_translate_mode of SceneTree.root instead.
int internationalization/rendering/root_node_layout_direction = 0 🔗
根節點的預設佈局方向。
String internationalization/rendering/text_driver = "" 🔗
Specifies the TextServer to use. If left empty, the default will be used.
"ICU / HarfBuzz / Graphite" (TextServerAdvanced) is the most advanced text driver, supporting right-to-left typesetting and complex scripts (for languages like Arabic, Hebrew, etc.). The "Fallback" text driver (TextServerFallback) does not support right-to-left typesetting and complex scripts.
Note: The driver in use can be overridden at runtime via the --text-driver command line argument.
Note: There is an additional Dummy text driver available, which disables all text rendering and font-related functionality. This driver is not listed in the project settings, but it can be enabled when running the editor or project using the --text-driver Dummy command line argument.
2D 導覽層 1 的可選名稱。留空則會顯示為“層 1”。
2D 導覽層 2 的可選名稱。留空則會顯示為“層 2”。
2D 導覽層 3 的可選名稱。留空則會顯示為“層 3”。
2D 導覽層 4 的可選名稱。留空則會顯示為“層 4”。
2D 導覽層 5 的可選名稱。留空則會顯示為“層 5”。
2D 導覽層 6 的可選名稱。留空則會顯示為“層 6”。
2D 導覽層 7 的可選名稱。留空則會顯示為“層 7”。
2D 導覽層 8 的可選名稱。留空則會顯示為“層 8”。
2D 導覽層 9 的可選名稱。留空則會顯示為“層 9”。
2D 導覽層 10 的可選名稱。留空則會顯示為“層 10”。
2D 導覽層 11 的可選名稱。留空則會顯示為“層 11”。
2D 導覽層 12 的可選名稱。留空則會顯示為“層 12”。
2D 導覽層 13 的可選名稱。留空則會顯示為“層 13”。
2D 導覽層 14 的可選名稱。留空則會顯示為“層 14”。
2D 導覽層 15 的可選名稱。留空則會顯示為“層 15”。
2D 導覽層 16 的可選名稱。留空則會顯示為“層 16”。
2D 導覽層 17 的可選名稱。留空則會顯示為“層 17”。
2D 導覽層 18 的可選名稱。留空則會顯示為“層 18”。
2D 導覽層 19 的可選名稱。留空則會顯示為“層 19”。
2D 導覽層 20 的可選名稱。留空則會顯示為“層 20”。
2D 導覽層 21 的可選名稱。留空則會顯示為“層 21”。
2D 導覽層 22 的可選名稱。留空則會顯示為“層 22”。
2D 導覽層 23 的可選名稱。留空則會顯示為“層 23”。
2D 導覽層 24 的可選名稱。留空則會顯示為“層 24”。
2D 導覽層 25 的可選名稱。留空則會顯示為“層 25”。
2D 導覽層 26 的可選名稱。留空則會顯示為“層 26”。
2D 導覽層 27 的可選名稱。留空則會顯示為“層 27”。
2D 導覽層 28 的可選名稱。留空則會顯示為“層 28”。
2D 導覽層 29 的可選名稱。留空則會顯示為“層 29”。
2D 導覽層 30 的可選名稱。留空則會顯示為“層 30”。
2D 導覽層 31 的可選名稱。留空則會顯示為“層 31”。
2D 導覽層 32 的可選名稱。留空則會顯示為“層 32”。
String layer_names/2d_physics/layer_1 = "" 🔗
2D 實體層 1 的可選名稱。留空則會顯示為“層 1”。
String layer_names/2d_physics/layer_2 = "" 🔗
2D 實體層 2 的可選名稱。留空則會顯示為“層 2”。
String layer_names/2d_physics/layer_3 = "" 🔗
2D 實體層 3 的可選名稱。留空則會顯示為“層 3”。
String layer_names/2d_physics/layer_4 = "" 🔗
2D 實體層 4 的可選名稱。留空則會顯示為“層 4”。
String layer_names/2d_physics/layer_5 = "" 🔗
2D 實體層 5 的可選名稱。留空則會顯示為“層 5”。
String layer_names/2d_physics/layer_6 = "" 🔗
2D 實體層 6 的可選名稱。留空則會顯示為“層 6”。
String layer_names/2d_physics/layer_7 = "" 🔗
2D 實體層 7 的可選名稱。留空則會顯示為“層 7”。
String layer_names/2d_physics/layer_8 = "" 🔗
2D 實體層 8 的可選名稱。留空則會顯示為“層 8”。
String layer_names/2d_physics/layer_9 = "" 🔗
2D 實體層 9 的可選名稱。留空則會顯示為“層 9”。
String layer_names/2d_physics/layer_10 = "" 🔗
2D 實體層 10 的可選名稱。留空則會顯示為“層 10”。
String layer_names/2d_physics/layer_11 = "" 🔗
2D 實體層 11 的可選名稱。留空則會顯示為“層 11”。
String layer_names/2d_physics/layer_12 = "" 🔗
2D 實體層 12 的可選名稱。留空則會顯示為“層 12”。
String layer_names/2d_physics/layer_13 = "" 🔗
2D 實體層 13 的可選名稱。留空則會顯示為“層 13”。
String layer_names/2d_physics/layer_14 = "" 🔗
2D 實體層 14 的可選名稱。留空則會顯示為“層 14”。
String layer_names/2d_physics/layer_15 = "" 🔗
2D 實體層 15 的可選名稱。留空則會顯示為“層 15”。
String layer_names/2d_physics/layer_16 = "" 🔗
2D 實體層 16 的可選名稱。留空則會顯示為“層 16”。
String layer_names/2d_physics/layer_17 = "" 🔗
2D 實體層 17 的可選名稱。留空則會顯示為“層 17”。
String layer_names/2d_physics/layer_18 = "" 🔗
2D 實體層 18 的可選名稱。留空則會顯示為“層 18”。
String layer_names/2d_physics/layer_19 = "" 🔗
2D 實體層 19 的可選名稱。留空則會顯示為“層 19”。
String layer_names/2d_physics/layer_20 = "" 🔗
2D 實體層 20 的可選名稱。留空則會顯示為“層 20”。
String layer_names/2d_physics/layer_21 = "" 🔗
2D 實體層 21 的可選名稱。留空則會顯示為“層 21”。
String layer_names/2d_physics/layer_22 = "" 🔗
2D 實體層 22 的可選名稱。留空則會顯示為“層 22”。
String layer_names/2d_physics/layer_23 = "" 🔗
2D 實體層 23 的可選名稱。留空則會顯示為“層 23”。
String layer_names/2d_physics/layer_24 = "" 🔗
2D 實體層 24 的可選名稱。留空則會顯示為“層 24”。
String layer_names/2d_physics/layer_25 = "" 🔗
2D 實體層 25 的可選名稱。留空則會顯示為“層 25”。
String layer_names/2d_physics/layer_26 = "" 🔗
2D 實體層 26 的可選名稱。留空則會顯示為“層 26”。
String layer_names/2d_physics/layer_27 = "" 🔗
2D 實體層 27 的可選名稱。留空則會顯示為“層 27”。
String layer_names/2d_physics/layer_28 = "" 🔗
2D 實體層 28 的可選名稱。留空則會顯示為“層 28”。
String layer_names/2d_physics/layer_29 = "" 🔗
2D 實體層 29 的可選名稱。留空則會顯示為“層 29”。
String layer_names/2d_physics/layer_30 = "" 🔗
2D 實體層 30 的可選名稱。留空則會顯示為“層 30”。
String layer_names/2d_physics/layer_31 = "" 🔗
2D 實體層 31 的可選名稱。留空則會顯示為“層 31”。
String layer_names/2d_physics/layer_32 = "" 🔗
2D 實體層 32 的可選名稱。留空則會顯示為“層 32”。
String layer_names/2d_render/layer_1 = "" 🔗
2D 算繪層 1 的可選名稱。留空則會顯示為“層 1”。
String layer_names/2d_render/layer_2 = "" 🔗
2D 算繪層 2 的可選名稱。留空則會顯示為“層 2”。
String layer_names/2d_render/layer_3 = "" 🔗
2D 算繪層 3 的可選名稱。留空則會顯示為“層 3”。
String layer_names/2d_render/layer_4 = "" 🔗
2D 算繪層 4 的可選名稱。留空則會顯示為“層 4”。
String layer_names/2d_render/layer_5 = "" 🔗
2D 算繪層 5 的可選名稱。留空則會顯示為“層 5”。
String layer_names/2d_render/layer_6 = "" 🔗
2D 算繪層 6 的可選名稱。留空則會顯示為“層 6”。
String layer_names/2d_render/layer_7 = "" 🔗
2D 算繪層 7 的可選名稱。留空則會顯示為“層 7”。
String layer_names/2d_render/layer_8 = "" 🔗
2D 算繪層 8 的可選名稱。留空則會顯示為“層 8”。
String layer_names/2d_render/layer_9 = "" 🔗
2D 算繪層 9 的可選名稱。留空則會顯示為“層 9”。
String layer_names/2d_render/layer_10 = "" 🔗
2D 算繪層 10 的可選名稱。留空則會顯示為“層 10”。
String layer_names/2d_render/layer_11 = "" 🔗
2D 算繪層 11 的可選名稱。留空則會顯示為“層 11”。
String layer_names/2d_render/layer_12 = "" 🔗
2D 算繪層 12 的可選名稱。留空則會顯示為“層 12”。
String layer_names/2d_render/layer_13 = "" 🔗
2D 算繪層 13 的可選名稱。留空則會顯示為“層 13”。
String layer_names/2d_render/layer_14 = "" 🔗
2D 算繪層 14 的可選名稱。留空則會顯示為“層 14”。
String layer_names/2d_render/layer_15 = "" 🔗
2D 算繪層 15 的可選名稱。留空則會顯示為“層 15”。
String layer_names/2d_render/layer_16 = "" 🔗
2D 算繪層 16 的可選名稱。留空則會顯示為“層 16”。
String layer_names/2d_render/layer_17 = "" 🔗
2D 算繪層 17 的可選名稱。留空則會顯示為“層 17”。
String layer_names/2d_render/layer_18 = "" 🔗
2D 算繪層 18 的可選名稱。留空則會顯示為“層 18”。
String layer_names/2d_render/layer_19 = "" 🔗
2D 算繪層 19 的可選名稱。留空則會顯示為“層 19”。
String layer_names/2d_render/layer_20 = "" 🔗
2D 算繪層 20 的可選名稱。留空則會顯示為“層 20”。
3D 導覽層 1 的可選名稱。留空則會顯示為“層 1”。
3D 導覽層 2 的可選名稱。留空則會顯示為“層 2”。
3D 導覽層 3 的可選名稱。留空則會顯示為“層 3”。
3D 導覽層 4 的可選名稱。留空則會顯示為“層 4”。
3D 導覽層 5 的可選名稱。留空則會顯示為“層 5”。
3D 導覽層 6 的可選名稱。留空則會顯示為“層 6”。
3D 導覽層 7 的可選名稱。留空則會顯示為“層 7”。
3D 導覽層 8 的可選名稱。留空則會顯示為“層 8”。
3D 導覽層 9 的可選名稱。留空則會顯示為“層 9”。
3D 導覽層 10 的可選名稱。留空則會顯示為“層 10”。
3D 導覽層 11 的可選名稱。留空則會顯示為“層 11”。
3D 導覽層 12 的可選名稱。留空則會顯示為“層 12”。
3D 導覽層 13 的可選名稱。留空則會顯示為“層 13”。
3D 導覽層 14 的可選名稱。留空則會顯示為“層 14”。
3D 導覽層 15 的可選名稱。留空則會顯示為“層 15”。
3D 導覽層 16 的可選名稱。留空則會顯示為“層 16”。
3D 導覽層 17 的可選名稱。留空則會顯示為“層 17”。
3D 導覽層 18 的可選名稱。留空則會顯示為“層 18”。
3D 導覽層 19 的可選名稱。留空則會顯示為“層 19”。
3D 導覽層 20 的可選名稱。留空則會顯示為“層 20”。
3D 導覽層 21 的可選名稱。留空則會顯示為“層 21”。
3D 導覽層 22 的可選名稱。留空則會顯示為“層 22”。
3D 導覽層 23 的可選名稱。留空則會顯示為“層 23”。
3D 導覽層 24 的可選名稱。留空則會顯示為“層 24”。
3D 導覽層 25 的可選名稱。留空則會顯示為“層 25”。
3D 導覽層 26 的可選名稱。留空則會顯示為“層 26”。
3D 導覽層 27 的可選名稱。留空則會顯示為“層 27”。
3D 導覽層 28 的可選名稱。留空則會顯示為“層 28”。
3D 導覽層 29 的可選名稱。留空則會顯示為“層 29”。
3D 導覽層 30 的可選名稱。留空則會顯示為“層 30”。
3D 導覽層 31 的可選名稱。留空則會顯示為“層 31”。
3D 導覽層 32 的可選名稱。留空則會顯示為“層 32”。
String layer_names/3d_physics/layer_1 = "" 🔗
3D 實體層 1 的可選名稱。留空則會顯示為“層 1”。
String layer_names/3d_physics/layer_2 = "" 🔗
3D 實體層 2 的可選名稱。留空則會顯示為“層 2”。
String layer_names/3d_physics/layer_3 = "" 🔗
3D 實體層 3 的可選名稱。留空則會顯示為“層 3”。
String layer_names/3d_physics/layer_4 = "" 🔗
3D 實體層 4 的可選名稱。留空則會顯示為“層 4”。
String layer_names/3d_physics/layer_5 = "" 🔗
3D 實體層 5 的可選名稱。留空則會顯示為“層 5”。
String layer_names/3d_physics/layer_6 = "" 🔗
3D 實體層 6 的可選名稱。留空則會顯示為“層 6”。
String layer_names/3d_physics/layer_7 = "" 🔗
3D 實體層 7 的可選名稱。留空則會顯示為“層 7”。
String layer_names/3d_physics/layer_8 = "" 🔗
3D 實體層 8 的可選名稱。留空則會顯示為“層 8”。
String layer_names/3d_physics/layer_9 = "" 🔗
3D 實體層 9 的可選名稱。留空則會顯示為“層 9”。
String layer_names/3d_physics/layer_10 = "" 🔗
3D 實體層 10 的可選名稱。留空則會顯示為“層 10”。
String layer_names/3d_physics/layer_11 = "" 🔗
3D 實體層 11 的可選名稱。留空則會顯示為“層 11”。
String layer_names/3d_physics/layer_12 = "" 🔗
3D 實體層 12 的可選名稱。留空則會顯示為“層 12”。
String layer_names/3d_physics/layer_13 = "" 🔗
3D 實體層 13 的可選名稱。留空則會顯示為“層 13”。
String layer_names/3d_physics/layer_14 = "" 🔗
3D 實體層 14 的可選名稱。留空則會顯示為“層 14”。
String layer_names/3d_physics/layer_15 = "" 🔗
3D 實體層 15 的可選名稱。留空則會顯示為“層 15”。
String layer_names/3d_physics/layer_16 = "" 🔗
3D 實體層 16 的可選名稱。留空則會顯示為“層 16”。
String layer_names/3d_physics/layer_17 = "" 🔗
3D 實體層 17 的可選名稱。留空則會顯示為“層 17”。
String layer_names/3d_physics/layer_18 = "" 🔗
3D 實體層 18 的可選名稱。留空則會顯示為“層 18”。
String layer_names/3d_physics/layer_19 = "" 🔗
3D 實體層 19 的可選名稱。留空則會顯示為“層 19”。
String layer_names/3d_physics/layer_20 = "" 🔗
3D 實體層 20 的可選名稱。留空則會顯示為“層 20”。
String layer_names/3d_physics/layer_21 = "" 🔗
3D 實體層 21 的可選名稱。留空則會顯示為“層 21”。
String layer_names/3d_physics/layer_22 = "" 🔗
3D 實體層 22 的可選名稱。留空則會顯示為“層 22”。
String layer_names/3d_physics/layer_23 = "" 🔗
3D 實體層 23 的可選名稱。留空則會顯示為“層 23”。
String layer_names/3d_physics/layer_24 = "" 🔗
3D 實體層 24 的可選名稱。留空則會顯示為“層 24”。
String layer_names/3d_physics/layer_25 = "" 🔗
3D 實體層 25 的可選名稱。留空則會顯示為“層 25”。
String layer_names/3d_physics/layer_26 = "" 🔗
3D 實體層 26 的可選名稱。留空則會顯示為“層 26”。
String layer_names/3d_physics/layer_27 = "" 🔗
3D 實體層 27 的可選名稱。留空則會顯示為“層 27”。
String layer_names/3d_physics/layer_28 = "" 🔗
3D 實體層 28 的可選名稱。留空則會顯示為“層 28”。
String layer_names/3d_physics/layer_29 = "" 🔗
3D 實體層 29 的可選名稱。留空則會顯示為“層 29”。
String layer_names/3d_physics/layer_30 = "" 🔗
3D 實體層 30 的可選名稱。留空則會顯示為“層 30”。
String layer_names/3d_physics/layer_31 = "" 🔗
3D 實體層 31 的可選名稱。留空則會顯示為“層 31”。
String layer_names/3d_physics/layer_32 = "" 🔗
3D 實體層 32 的可選名稱。留空則會顯示為“層 32”。
String layer_names/3d_render/layer_1 = "" 🔗
3D 算繪層 1 的可選名稱。留空則會顯示為“層 1”。
String layer_names/3d_render/layer_2 = "" 🔗
3D 算繪層 2 的可選名稱。留空則會顯示為“層 2”。
String layer_names/3d_render/layer_3 = "" 🔗
3D 算繪層 3 的可選名稱。留空則會顯示為“層 3”。
String layer_names/3d_render/layer_4 = "" 🔗
3D 算繪層 4 的可選名稱。留空則會顯示為“層 4”。
String layer_names/3d_render/layer_5 = "" 🔗
3D 算繪層 5 的可選名稱。留空則會顯示為“層 5”。
String layer_names/3d_render/layer_6 = "" 🔗
3D 算繪層 6 的可選名稱。留空則會顯示為“層 6”。
String layer_names/3d_render/layer_7 = "" 🔗
3D 算繪層 7 的可選名稱。留空則會顯示為“層 7”。
String layer_names/3d_render/layer_8 = "" 🔗
3D 算繪層 8 的可選名稱。留空則會顯示為“層 8”。
String layer_names/3d_render/layer_9 = "" 🔗
3D 算繪層 9 的可選名稱。留空則會顯示為“層 9”。
String layer_names/3d_render/layer_10 = "" 🔗
3D 算繪層 10 的可選名稱。留空則會顯示為“層 10”。
String layer_names/3d_render/layer_11 = "" 🔗
3D 算繪層 11 的可選名稱。留空則會顯示為“層 11”。
String layer_names/3d_render/layer_12 = "" 🔗
3D 算繪層 12 的可選名稱。留空則會顯示為“層 12”。
String layer_names/3d_render/layer_13 = "" 🔗
3D 算繪層 13 的可選名稱。留空則會顯示為“層 13”。
String layer_names/3d_render/layer_14 = "" 🔗
3D 算繪層 14 的可選名稱。留空則會顯示為“層 14”。
String layer_names/3d_render/layer_15 = "" 🔗
3D 算繪層 15 的可選名稱。留空則會顯示為“層 15”。
String layer_names/3d_render/layer_16 = "" 🔗
3D 算繪層 16 的可選名稱。留空則會顯示為“層 16”。
String layer_names/3d_render/layer_17 = "" 🔗
3D 算繪層 17 的可選名稱。留空則會顯示為“層 17”。
String layer_names/3d_render/layer_18 = "" 🔗
3D 算繪層 18 的可選名稱。留空則會顯示為“層 18”。
String layer_names/3d_render/layer_19 = "" 🔗
3D 算繪層 19 的可選名稱。留空則會顯示為“層 19”。
String layer_names/3d_render/layer_20 = "" 🔗
3D 算繪層 20 的可選名稱。留空則會顯示為“層 20”。
String layer_names/avoidance/layer_1 = "" 🔗
導覽避障層 1 的可選名稱。留空則會顯示為“層 1”。
String layer_names/avoidance/layer_2 = "" 🔗
導覽避障層 2 的可選名稱。留空則會顯示為“層 2”。
String layer_names/avoidance/layer_3 = "" 🔗
導覽避障層 3 的可選名稱。留空則會顯示為“層 3”。
String layer_names/avoidance/layer_4 = "" 🔗
導覽避障層 4 的可選名稱。留空則會顯示為“層 4”。
String layer_names/avoidance/layer_5 = "" 🔗
導覽避障層 5 的可選名稱。留空則會顯示為“層 5”。
String layer_names/avoidance/layer_6 = "" 🔗
導覽避障層 6 的可選名稱。留空則會顯示為“層 6”。
String layer_names/avoidance/layer_7 = "" 🔗
導覽避障層 7 的可選名稱。留空則會顯示為“層 7”。
String layer_names/avoidance/layer_8 = "" 🔗
導覽避障層 8 的可選名稱。留空則會顯示為“層 8”。
String layer_names/avoidance/layer_9 = "" 🔗
導覽避障層 9 的可選名稱。留空則會顯示為“層 9”。
String layer_names/avoidance/layer_10 = "" 🔗
導覽避障層 10 的可選名稱。留空則會顯示為“層 10”。
String layer_names/avoidance/layer_11 = "" 🔗
導覽避障層 11 的可選名稱。留空則會顯示為“層 11”。
String layer_names/avoidance/layer_12 = "" 🔗
導覽避障層 12 的可選名稱。留空則會顯示為“層 12”。
String layer_names/avoidance/layer_13 = "" 🔗
導覽避障層 13 的可選名稱。留空則會顯示為“層 13”。
String layer_names/avoidance/layer_14 = "" 🔗
導覽避障層 14 的可選名稱。留空則會顯示為“層 14”。
String layer_names/avoidance/layer_15 = "" 🔗
導覽避障層 15 的可選名稱。留空則會顯示為“層 15”。
String layer_names/avoidance/layer_16 = "" 🔗
導覽避障層 16 的可選名稱。留空則會顯示為“層 16”。
String layer_names/avoidance/layer_17 = "" 🔗
導覽避障層 17 的可選名稱。留空則會顯示為“層 17”。
String layer_names/avoidance/layer_18 = "" 🔗
導覽避障層 18 的可選名稱。留空則會顯示為“層 18”。
String layer_names/avoidance/layer_19 = "" 🔗
導覽避障層 19 的可選名稱。留空則會顯示為“層 19”。
String layer_names/avoidance/layer_20 = "" 🔗
導覽避障層 20 的可選名稱。留空則會顯示為“層 20”。
String layer_names/avoidance/layer_21 = "" 🔗
導覽避障層 21 的可選名稱。留空則會顯示為“層 21”。
String layer_names/avoidance/layer_22 = "" 🔗
導覽避障層 22 的可選名稱。留空則會顯示為“層 22”。
String layer_names/avoidance/layer_23 = "" 🔗
導覽避障層 23 的可選名稱。留空則會顯示為“層 23”。
String layer_names/avoidance/layer_24 = "" 🔗
導覽避障層 24 的可選名稱。留空則會顯示為“層 24”。
String layer_names/avoidance/layer_25 = "" 🔗
導覽避障層 25 的可選名稱。留空則會顯示為“層 25”。
String layer_names/avoidance/layer_26 = "" 🔗
導覽避障層 26 的可選名稱。留空則會顯示為“層 26”。
String layer_names/avoidance/layer_27 = "" 🔗
導覽避障層 27 的可選名稱。留空則會顯示為“層 27”。
String layer_names/avoidance/layer_28 = "" 🔗
導覽避障層 28 的可選名稱。留空則會顯示為“層 28”。
String layer_names/avoidance/layer_29 = "" 🔗
導覽避障層 29 的可選名稱。留空則會顯示為“層 29”。
String layer_names/avoidance/layer_30 = "" 🔗
導覽避障層 30 的可選名稱。留空則會顯示為“層 30”。
String layer_names/avoidance/layer_31 = "" 🔗
導覽避障層 31 的可選名稱。留空則會顯示為“層 31”。
String layer_names/avoidance/layer_32 = "" 🔗
導覽避障層 32 的可選名稱。留空則會顯示為“層 32”。
int memory/limits/message_queue/max_size_mb = 32 🔗
Godot 使用一個訊息佇列來延遲一些函式呼叫。如果你的空間用完了（你會看到一個錯誤），你可以在這裡增加大小。
2D 導覽地圖的預設儲存格大小。見 NavigationServer2D.map_set_cell_size()。
2D 導覽地圖的預設邊界連結邊距。見 NavigationServer2D.map_set_edge_connection_margin()。
2D 導覽地圖的預設連結連接半徑。見 NavigationServer2D.map_set_link_connection_radius()。
Default merge rasterizer cell scale for 2D navigation maps. See NavigationServer2D.map_set_merge_rasterizer_cell_scale().
Sets which navigation engine to use for 2D navigation.
DEFAULT is equivalent to GodotNavigation2D, but may change in future releases. Select an explicit implementation if you want to ensure that your project stays on the same engine.
GodotNavigation2D is Godot's internal 2D navigation engine.
Dummy is a 2D navigation server that does nothing and returns only dummy values, effectively disabling all 2D navigation functionality.
Third-party modules can add other navigation engines to select with this setting.
如果啟用，則 2D 導覽地區會使用邊緣連接來連接其他導覽地區，使用的是導覽地圖邊緣連接邊距。這項設定只會影響 World2D 的預設導覽地圖。
如果 true，則當在較大尺寸的導航地圖上使用較小單元格尺寸的導航網格時，導航系統將列印警告，因為這通常會導致柵格化錯誤。
如果true，導航系統將列印有關導航區域或地圖中發生的導航網格邊緣合併錯誤的警告。
3D 導覽地圖的預設儲存格高度。見 NavigationServer3D.map_set_cell_height()。
3D 導覽地圖的預設儲存格大小。見 NavigationServer3D.map_set_cell_size()。
3D 導覽地圖的預設邊界連結邊距。見 NavigationServer3D.map_set_edge_connection_margin()。
3D 導覽地圖的預設連結連接半徑。見 NavigationServer3D.map_set_link_connection_radius()。
3D 導覽地圖的預設儲存格大小。見 NavigationServer3D.map_set_cell_size()。
Default merge rasterizer cell scale for 3D navigation maps. See NavigationServer3D.map_set_merge_rasterizer_cell_scale().
Sets which navigation engine to use for 3D navigation.
DEFAULT is equivalent to GodotNavigation3D, but may change in future releases. Select an explicit implementation if you want to ensure that your project stays on the same engine.
GodotNavigation3D is Godot's internal 3D navigation engine.
Dummy is a 3D navigation server that does nothing and returns only dummy values, effectively disabling all 3D navigation functionality.
Third-party modules can add other navigation engines to select with this setting.
如果啟用，則 3D 導覽地區會使用邊緣連接來連接其他導覽地區，使用的是導覽地圖邊緣連接邊距。這項設定只會影響 World3D 的預設導覽地圖。
如果 true，則當在較大尺寸的導航地圖上使用較小單元格尺寸（或 3D 高度）的導航網格時，導航系統將列印警告，因為這通常會導致柵格化錯誤。
如果true，導航系統將列印有關導航區域或地圖中發生的導航網格邊緣合併錯誤的警告。
如果啟用，並且使用多執行緒進行避障計算，則執行的執行緒為高優先順序。
如果啟用，則會使用多執行緒進行避障計算。
如果啟用，並且使用多執行緒進行避障計算，則執行的執行緒為高優先順序。
如果啟用，則會使用多執行緒進行避障計算。
If enabled, and baking would potentially lead to an engine crash, the baking will be interrupted and an error message with explanation will be raised.
Maximum number of threads that can run pathfinding queries simultaneously on the same pathfinding graph, for example the same navigation map. Additional threads increase memory consumption and synchronization time due to the need for extra data copies prepared for each thread. A value of -1 means unlimited and the maximum available OS processor count is used. Defaults to 1 when the OS does not support threads.
If enabled, navigation map synchronization uses an async process that runs on a background thread. This avoids stalling the main thread but adds an additional delay to any navigation map change.
If enabled, navigation region synchronization uses an async process that runs on a background thread. This avoids stalling the main thread but adds an additional delay to any navigation region change.
int network/limits/debugger/max_chars_per_second = 32768 🔗
允許作為除錯器輸出發送的最大字元數。超過該值，內容將被丟棄。這有助於避免除錯器的連接停滯。
int network/limits/debugger/max_errors_per_second = 400 🔗
允許從除錯器發送的最大錯誤數。超過該值，內容將被丟棄。這有助於避免除錯器的連接停滯。
int network/limits/debugger/max_queued_messages = 2048 🔗
除錯器佇列中的最大消息數。超過該值，內容將被丟棄。這有助於限制除錯器的記憶體使用。
int network/limits/debugger/max_warnings_per_second = 400 🔗
允許從除錯器發送的最大警告數。超過此值，內容將被丟棄。這有助於避免除錯器的連接停滯。
int network/limits/packet_peer_stream/max_buffer_po2 = 16 🔗
用於反序列化 Godot 資料的封包對等流 (stream) 的預設大小（以位元組為單位，指定為 2 的冪）。預設值 16 等於 65,536 位元組。超過此大小，資料將被丟棄。
int network/limits/tcp/connect_timeout_seconds = 30 🔗
使用TCP的連接嘗試的超時（以秒為單位）。
int network/limits/unix/connect_timeout_seconds = 30 🔗
Timeout (in seconds) for connection attempts using UNIX domain socket.
int network/limits/webrtc/max_channel_in_buffer_kb = 64 🔗
WebRTCDataChannel 輸入緩衝區的最大尺寸（單位為 kiB）。
String network/tls/certificate_bundle_override = "" 🔗
The CA certificates bundle to use for TLS connections. If this is set to a non-empty value, this will override Godot's default Mozilla certificate bundle. If left empty, the default certificate bundle will be used.
If in doubt, leave this setting empty.
bool network/tls/enable_tls_v1.3 = true 🔗
If true, enable TLSv1.3 negotiation.
Note: Only supported when using Mbed TLS 3.0 or later (Linux distribution packages may be compiled against older system Mbed TLS packages), otherwise the maximum supported TLS version is always TLSv1.2.
float physics/2d/default_angular_damp = 1.0 🔗
The default rotational motion damping in 2D. Damping is used to gradually slow down physical objects over time. RigidBodies will fall back to this value when combining their own damping values and no area damping value is present.
Suggested values are in the range 0 to 30. At value 0 objects will keep moving with the same velocity. Greater values will stop the object faster. A value equal to or greater than the physics tick rate (physics/common/physics_ticks_per_second) will bring the object to a stop in one iteration.
Note: Godot damping calculations are velocity-dependent, meaning bodies moving faster will take a longer time to come to rest. They do not simulate inertia, friction, or air resistance. Therefore heavier or larger bodies will lose speed at the same proportional rate as lighter or smaller bodies.
During each physics tick, Godot will multiply the linear velocity of RigidBodies by 1.0 - combined_damp / physics_ticks_per_second. By default, bodies combine damp factors: combined_damp is the sum of the damp value of the body and this value or the area's value the body is in. See DampMode.
Warning: Godot's damping calculations are simulation tick rate dependent. Changing physics/common/physics_ticks_per_second may significantly change the outcomes and feel of your simulation. This is true for the entire range of damping values greater than 0. To get back to a similar feel, you also need to change your damp values. This needed change is not proportional and differs from case to case.
float physics/2d/default_gravity = 980.0 🔗
2D 預設重力強度（單位為圖元每平方秒）。
注意：這個屬性只在專案啟動時讀取。要在運作時改變預設重力，請使用以下代碼範例：
# 將預設重力強度設定為 980。
PhysicsServer2D.area_set_param(get_viewport().find_world_2d().space, PhysicsServer2D.AREA_PARAM_GRAVITY, 980)
// 將預設重力強度設定為 980。
PhysicsServer2D.AreaSetParam(GetViewport().FindWorld2D().Space, PhysicsServer2D.AreaParameter.Gravity, 980);
Vector2 physics/2d/default_gravity_vector = Vector2(0, 1) 🔗
2D 預設重力方向。
注意：這個屬性只在專案啟動時讀取。要在運作時改變預設重力向量，請使用以下程式碼範例：
# 將預設整理方向設定為 `Vector2(0, 1)`。
PhysicsServer2D.area_set_param(get_viewport().find_world_2d().space, PhysicsServer2D.AREA_PARAM_GRAVITY_VECTOR, Vector2.DOWN)
// 將預設整理方向設定為 `Vector2(0, 1)`。
PhysicsServer2D.AreaSetParam(GetViewport().FindWorld2D().Space, PhysicsServer2D.AreaParameter.GravityVector, Vector2.Down)
float physics/2d/default_linear_damp = 0.1 🔗
The default linear motion damping in 2D. Damping is used to gradually slow down physical objects over time. RigidBodies will fall back to this value when combining their own damping values and no area damping value is present.
Suggested values are in the range 0 to 30. At value 0 objects will keep moving with the same velocity. Greater values will stop the object faster. A value equal to or greater than the physics tick rate (physics/common/physics_ticks_per_second) will bring the object to a stop in one iteration.
Note: Godot damping calculations are velocity-dependent, meaning bodies moving faster will take a longer time to come to rest. They do not simulate inertia, friction, or air resistance. Therefore heavier or larger bodies will lose speed at the same proportional rate as lighter or smaller bodies.
During each physics tick, Godot will multiply the linear velocity of RigidBodies by 1.0 - combined_damp / physics_ticks_per_second, where combined_damp is the sum of the linear damp of the body and this value, or the area's value the body is in, assuming the body defaults to combine damp values. See DampMode.
Warning: Godot's damping calculations are simulation tick rate dependent. Changing physics/common/physics_ticks_per_second may significantly change the outcomes and feel of your simulation. This is true for the entire range of damping values greater than 0. To get back to a similar feel, you also need to change your damp values. This needed change is not proportional and differs from case to case.
String physics/2d/physics_engine = "DEFAULT" 🔗
Sets which physics engine to use for 2D physics.
DEFAULT is currently equivalent to GodotPhysics2D, but may change in future releases. Select an explicit implementation if you want to ensure that your project stays on the same engine.
GodotPhysics2D is Godot's internal 2D physics engine.
Dummy is a 2D physics server that does nothing and returns only dummy values, effectively disabling all 2D physics functionality.
Third-party extensions and modules can add other physics engines to select with this setting.
bool physics/2d/run_on_separate_thread = false 🔗
如果為 true，則 2D 物理伺服器會在單獨的執行緒上運作，能夠更好地利用多核 CPU。如果為 false，則 2D 物理伺服器會在主執行緒上運作。在單獨的執行緒上運作物理伺服器能夠提高性能，但會把對 API 的存取限制在物理處理中。
float physics/2d/sleep_threshold_angular = 0.13962634 🔗
角速度的閾值，在該閾值以下的 2D 物理體將被視為不活動。見 PhysicsServer2D.SPACE_PARAM_BODY_ANGULAR_VELOCITY_SLEEP_THRESHOLD。
float physics/2d/sleep_threshold_linear = 2.0 🔗
Threshold linear velocity under which a 2D physics body will be considered inactive. See PhysicsServer2D.SPACE_PARAM_BODY_LINEAR_VELOCITY_SLEEP_THRESHOLD.
Note: Only supported when using GodotPhysics3D. This project setting is ignored when using Jolt Physics.
float physics/2d/solver/contact_max_allowed_penetration = 0.3 🔗
Maximum distance a shape can penetrate another shape before it is considered a collision. See PhysicsServer2D.SPACE_PARAM_CONTACT_MAX_ALLOWED_PENETRATION.
Note: Only supported when using GodotPhysics3D. This project setting is ignored when using Jolt Physics.
float physics/2d/solver/contact_max_separation = 1.5 🔗
兩個形狀間的最大距離，超過該距離後它們將被視為分離，接觸將被棄置。見 PhysicsServer2D.SPACE_PARAM_CONTACT_MAX_SEPARATION。
float physics/2d/solver/contact_recycle_radius = 1.0 🔗
一對物體在其碰撞狀態被重新計算之前的最大移動距離。見 PhysicsServer2D.SPACE_PARAM_CONTACT_RECYCLE_RADIUS。
float physics/2d/solver/default_constraint_bias = 0.2 🔗
所有物理約束的預設求解器偏置。定義物體對強制約束的反應程度。見 PhysicsServer2D.SPACE_PARAM_CONSTRAINT_DEFAULT_BIAS。
不同的約束可以指定不同的偏置值（見 Joint2D.bias）。
float physics/2d/solver/default_contact_bias = 0.8 🔗
所有物理接觸的預設求解器偏置。定義物體對強制接觸拆分的反應程度。見 PhysicsServer2D.SPACE_PARAM_CONTACT_DEFAULT_BIAS。
不同的形狀可以指定不同的偏置值（見 Shape2D.custom_solver_bias）。
int physics/2d/solver/solver_iterations = 16 🔗
所有接觸和約束的求解器反覆運算次數。反覆運算次數越多，碰撞越精確。然而，大量反覆運算也會消耗大量的 CPU 性能，可能導致性能的降低。見 PhysicsServer2D.SPACE_PARAM_SOLVER_ITERATIONS。
float physics/2d/time_before_sleep = 0.5 🔗
2D 物理物體進入睡眠狀態之前，所需的不活動時間（以秒為單位）。請參閱 PhysicsServer2D.SPACE_PARAM_BODY_TIME_TO_SLEEP。
float physics/3d/default_angular_damp = 0.1 🔗
The default rotational motion damping in 3D. Damping is used to gradually slow down physical objects over time. RigidBodies will fall back to this value when combining their own damping values and no area damping value is present.
Suggested values are in the range 0 to 30. At value 0 objects will keep moving with the same velocity. Greater values will stop the object faster. A value equal to or greater than the physics tick rate (physics/common/physics_ticks_per_second) will bring the object to a stop in one iteration.
Note: Godot damping calculations are velocity-dependent, meaning bodies moving faster will take a longer time to come to rest. They do not simulate inertia, friction, or air resistance. Therefore heavier or larger bodies will lose speed at the same proportional rate as lighter or smaller bodies.
During each physics tick, Godot will multiply the angular velocity of RigidBodies by 1.0 - combined_damp / physics_ticks_per_second. By default, bodies combine damp factors: combined_damp is the sum of the damp value of the body and this value or the area's value the body is in. See DampMode.
Warning: Godot's damping calculations are simulation tick rate dependent. Changing physics/common/physics_ticks_per_second may significantly change the outcomes and feel of your simulation. This is true for the entire range of damping values greater than 0. To get back to a similar feel, you also need to change your damp values. This needed change is not proportional and differs from case to case.
float physics/3d/default_gravity = 9.8 🔗
3D 預設重力強度（單位為米每平方秒）。
注意：這個屬性只在專案啟動時讀取。要在運作時改變預設重力，請使用以下代碼範例：
# 將預設重力強度設定為 9.8。
PhysicsServer3D.area_set_param(get_viewport().find_world().space, PhysicsServer3D.AREA_PARAM_GRAVITY, 9.8)
// 將預設重力強度設定為 9.8。
PhysicsServer3D.AreaSetParam(GetViewport().FindWorld().Space, PhysicsServer3D.AreaParameter.Gravity, 9.8);
Vector3 physics/3d/default_gravity_vector = Vector3(0, -1, 0) 🔗
3D 預設重力方向。
注意：這個屬性只在專案啟動時讀取。要在運作時改變預設重力向量，請使用以下程式碼範例：
# 將預設整理方向設定為 `Vector3(0, -1, 0)`。
PhysicsServer3D.area_set_param(get_viewport().find_world().get_space(), PhysicsServer3D.AREA_PARAM_GRAVITY_VECTOR, Vector3.DOWN)
// 將預設整理方向設定為 `Vector3(0, -1, 0)`。
PhysicsServer3D.AreaSetParam(GetViewport().FindWorld().Space, PhysicsServer3D.AreaParameter.GravityVector, Vector3.Down)
float physics/3d/default_linear_damp = 0.1 🔗
The default linear motion damping in 3D. Damping is used to gradually slow down physical objects over time. RigidBodies will fall back to this value when combining their own damping values and no area damping value is present.
Suggested values are in the range 0 to 30. At value 0 objects will keep moving with the same velocity. Greater values will stop the object faster. A value equal to or greater than the physics tick rate (physics/common/physics_ticks_per_second) will bring the object to a stop in one iteration.
Note: Godot damping calculations are velocity-dependent, meaning bodies moving faster will take a longer time to come to rest. They do not simulate inertia, friction, or air resistance. Therefore heavier or larger bodies will lose speed at the same proportional rate as lighter or smaller bodies.
During each physics tick, Godot will multiply the linear velocity of RigidBodies by 1.0 - combined_damp / physics_ticks_per_second. By default, bodies combine damp factors: combined_damp is the sum of the damp value of the body and this value or the area's value the body is in. See DampMode.
Warning: Godot's damping calculations are simulation tick rate dependent. Changing physics/common/physics_ticks_per_second may significantly change the outcomes and feel of your simulation. This is true for the entire range of damping values greater than 0. To get back to a similar feel, you also need to change your damp values. This needed change is not proportional and differs from case to case.
String physics/3d/physics_engine = "DEFAULT" 🔗
Sets which physics engine to use for 3D physics.
DEFAULT is currently equivalent to GodotPhysics3D, but may change in future releases. Select an explicit implementation if you want to ensure that your project stays on the same engine.
GodotPhysics3D is Godot's internal 3D physics engine.
Jolt Physics is an alternative physics engine that is generally faster and more reliable than GodotPhysics3D. Jolt Physics is the default for projects created starting in Godot 4.6.
Dummy is a 3D physics server that does nothing and returns only dummy values, effectively disabling all 3D physics functionality.
Third-party extensions and modules can add other physics engines to select with this setting.
String physics/3d/physics_interpolation/scene_traversal = "DEFAULT" 🔗
The approach used for 3D scene traversal when physics interpolation is enabled.
DEFAULT: The default optimized method.Legacy: The previous reference method used for scene tree traversal, which is slower.Debug: Swaps betweenDEFAULTandLegacymethods on alternating frames, and provides logging information (which in turn makes it slower). Intended for debugging only; you should use theDEFAULTmethod in most cases.
bool physics/3d/run_on_separate_thread = false 🔗
If true, the 3D physics server runs on a separate thread, making better use of multi-core CPUs. If false, the 3D physics server runs on the main thread. Running the physics server on a separate thread can increase performance, but restricts API access to only physics process.
Note: When physics/3d/physics_engine is set to Jolt Physics, enabling this setting will prevent the 3D physics server from being able to provide any context when reporting errors and warnings, and will instead always refer to nodes as <unknown>.
float physics/3d/sleep_threshold_angular = 0.13962634 🔗
Threshold angular velocity under which a 3D physics body will be considered inactive. See PhysicsServer3D.SPACE_PARAM_BODY_ANGULAR_VELOCITY_SLEEP_THRESHOLD.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
float physics/3d/sleep_threshold_linear = 0.1 🔗
Threshold linear velocity under which a 3D physics body will be considered inactive. See PhysicsServer3D.SPACE_PARAM_BODY_LINEAR_VELOCITY_SLEEP_THRESHOLD.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
float physics/3d/solver/contact_max_allowed_penetration = 0.01 🔗
Maximum distance a shape can penetrate another shape before it is considered a collision. See PhysicsServer3D.SPACE_PARAM_CONTACT_MAX_ALLOWED_PENETRATION.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
float physics/3d/solver/contact_max_separation = 0.05 🔗
Maximum distance a shape can be from another before they are considered separated and the contact is discarded. See PhysicsServer3D.SPACE_PARAM_CONTACT_MAX_SEPARATION.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
float physics/3d/solver/contact_recycle_radius = 0.01 🔗
Maximum distance a pair of bodies has to move before their collision status has to be recalculated. See PhysicsServer3D.SPACE_PARAM_CONTACT_RECYCLE_RADIUS.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
float physics/3d/solver/default_contact_bias = 0.8 🔗
Default solver bias for all physics contacts. Defines how much bodies react to enforce contact separation. See PhysicsServer3D.SPACE_PARAM_CONTACT_DEFAULT_BIAS.
Individual shapes can have a specific bias value (see Shape3D.custom_solver_bias).
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
int physics/3d/solver/solver_iterations = 16 🔗
Number of solver iterations for all contacts and constraints. The greater the number of iterations, the more accurate the collisions will be. However, a greater number of iterations requires more CPU power, which can decrease performance. See PhysicsServer3D.SPACE_PARAM_SOLVER_ITERATIONS.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
float physics/3d/time_before_sleep = 0.5 🔗
Time (in seconds) of inactivity before which a 3D physics body will put to sleep. See PhysicsServer3D.SPACE_PARAM_BODY_TIME_TO_SLEEP.
Note: This project setting is only effective when using GodotPhysics3D. It has no effect when using Jolt Physics.
bool physics/common/enable_object_picking = true 🔗
在根視圖上啟用 Viewport.physics_object_picking。
int physics/common/max_physics_steps_per_frame = 8 🔗
Controls the maximum number of physics steps that can be simulated each rendered frame. The default value is tuned to avoid situations where the framerate suddenly drops to a very low value beyond a certain amount of physics simulation. This occurs because the physics engine can't keep up with the expected simulation rate. In this case, the framerate will start dropping, but the engine is only allowed to simulate a certain number of physics steps per rendered frame. This snowballs into a situation where framerate keeps dropping until it reaches a very low framerate (typically 1-2 FPS) and is called the physics spiral of death.
However, the game will appear to slow down if the rendering FPS is less than 1 / max_physics_steps_per_frame of physics/common/physics_ticks_per_second. This occurs even if delta is consistently used in physics calculations. To avoid this, increase physics/common/max_physics_steps_per_frame if you have increased physics/common/physics_ticks_per_second significantly above its default value.
Note: This property is only read when the project starts. To change the maximum number of simulated physics steps per frame at runtime, set Engine.max_physics_steps_per_frame instead.
bool physics/common/physics_interpolation = false 🔗
If true, the renderer will interpolate the transforms of objects (both physics and non-physics) between the last two transforms, so that smooth motion is seen even when physics ticks do not coincide with rendered frames. See also Node.reset_physics_interpolation().
Note: Although this is a global setting, finer control of individual branches of the SceneTree is possible using Node.physics_interpolation_mode.
Note: This property is only read when the project starts. To toggle physics interpolation at runtime, set SceneTree.physics_interpolation instead.
Note: Property physics/common/physics_jitter_fix is automatically disabled if physics/common/physics_interpolation is set to true, as the two methods are incompatible.
float physics/common/physics_jitter_fix = 0.5 🔗
Controls how much physics ticks are synchronized with real time. For 0 or less, the ticks are synchronized. Such values are recommended for network games, where clock synchronization matters. Higher values cause higher deviation of in-game clock and real clock, but allows smoothing out framerate jitters. The default value of 0.5 should be good enough for most; values above 2 could cause the game to react to dropped frames with a noticeable delay and are not recommended.
Note: Jitter fix is automatically disabled at runtime when physics/common/physics_interpolation is enabled.
Note: When using a custom physics interpolation solution, the physics jitter fix should be disabled by setting physics/common/physics_jitter_fix to 0.0.
Note: This property is only read when the project starts. To change the physics jitter fix at runtime, set Engine.physics_jitter_fix instead.
int physics/common/physics_ticks_per_second = 60 🔗
The number of fixed iterations per second. This controls how often physics simulation and the Node._physics_process() method are run.
CPU usage scales approximately with the physics tick rate. However, at very low tick rates (usually below 30), physics behavior can break down. Input can also become less responsive at low tick rates as there can be a gap between input being registered, and the response on the next physics tick. High tick rates give more accurate physics simulation, particularly for fast moving objects. For example, racing games may benefit from increasing the tick rate above the default 60.
See also application/run/max_fps.
Note: This property is only read when the project starts. To change the physics FPS at runtime, set Engine.physics_ticks_per_second instead.
Note: Only physics/common/max_physics_steps_per_frame physics ticks may be simulated per rendered frame at most. If more physics ticks have to be simulated per rendered frame to keep up with rendering, the project will appear to slow down (even if delta is used consistently in physics calculations). Therefore, it is recommended to also increase physics/common/max_physics_steps_per_frame if increasing physics/common/physics_ticks_per_second significantly above its default value.
Note: Consider enabling physics interpolation if you change physics/common/physics_ticks_per_second to a value that is not a multiple of 60. Using physics interpolation will avoid jittering when the monitor refresh rate and physics update rate don't exactly match.
float physics/jolt_physics_3d/collisions/active_edge_threshold = 0.87266463 🔗
The maximum angle, in radians, between two adjacent triangles in a ConcavePolygonShape3D or HeightMapShape3D for which the edge between those triangles is considered inactive.
Collisions against an inactive edge will have its normal overridden to instead be the surface normal of the triangle. This can help alleviate ghost collisions.
Note: Setting this too high can result in objects not depenetrating properly.
Note: This applies to all shape queries, as well as physics bodies within the simulation.
Note: This does not apply when enabling Jolt's enhanced internal edge removal, which supersedes this.
float physics/jolt_physics_3d/collisions/collision_margin_fraction = 0.08 🔗
The amount of collision margin to use for certain convex collision shapes, such as BoxShape3D, CylinderShape3D and ConvexPolygonShape3D, as a fraction of the shape's shortest axis, with Shape3D.margin as the upper bound. This is mainly used to speed up collision detection with convex shapes.
Note: Collision margins in Jolt do not add any extra size to the shape. Instead the shape is first shrunk by the margin and then expanded by the same amount, resulting in a shape with rounded corners.
Note: Setting this value too close to 0.0 may also negatively affect the accuracy of the collision detection with convex shapes.
int physics/jolt_physics_3d/joints/world_node = 0 🔗
Which of the two nodes bound by a joint should represent the world when one of the two is omitted, as either Joint3D.node_a or Joint3D.node_b. This can be thought of as having the omitted node be a StaticBody3D at the joint's position. Joint limits are more easily expressed when Joint3D.node_a represents the world.
Note: In Godot Physics, only Joint3D.node_b can represent the world.
float physics/jolt_physics_3d/limits/max_angular_velocity = 47.12389 🔗
The maximum angular velocity that a RigidBody3D can reach, in radians per second.
This is mainly used as a fail-safe, to prevent the simulation from exploding, as fast-moving objects colliding with complex physics structures can otherwise cause them to go out of control. Fast-moving objects can also cause a lot of stress on the collision detection system, which can slow down the simulation considerably.
int physics/jolt_physics_3d/limits/max_bodies = 10240 🔗
The maximum number of PhysicsBody3D to support at the same time, awake or sleeping. When this limit is exceeded, an error is reported and anything past that point is undefined behavior.
Note: This limit also applies within the editor.
int physics/jolt_physics_3d/limits/max_body_pairs = 65536 🔗
The maximum number of body pairs to allow processing of. When this limit is exceeded, a warning is reported and collisions will randomly be ignored while bodies pass through each other.
int physics/jolt_physics_3d/limits/max_contact_constraints = 20480 🔗
The maximum number of contact constraints to allow processing of. When this limit is exceeded, a warning is reported and collisions will randomly be ignored while bodies pass through each other.
float physics/jolt_physics_3d/limits/max_linear_velocity = 500.0 🔗
The maximum linear velocity that a RigidBody3D can reach, in meters per second.
This is mainly used as a fail-safe, to prevent the simulation from exploding, as fast-moving objects colliding with complex physics structures can otherwise cause them to go out of control. Fast-moving objects can also cause a lot of stress on the collision detection system, which can slow down the simulation considerably.
int physics/jolt_physics_3d/limits/temporary_memory_buffer_size = 32 🔗
The amount of memory to pre-allocate for the stack allocator used within Jolt, in MiB. This allocator is used within the physics step to store things that are only needed during it, like which bodies are in contact, how they form islands and the data needed to solve the contacts.
float physics/jolt_physics_3d/limits/world_boundary_shape_size = 2000.0 🔗
The size of WorldBoundaryShape3D boundaries, for all three dimensions. The plane is effectively centered within a box of this size, and anything outside of the box will not collide with it. This is necessary as WorldBoundaryShape3D is not unbounded when using Jolt, in order to prevent precision issues.
Note: Setting this value too high can make collision detection less accurate.
Note: Collisions against the effective edges of a WorldBoundaryShape3D will be inconsistent.
float physics/jolt_physics_3d/motion_queries/recovery_amount = 0.4 🔗
Fraction of the total penetration to depenetrate per iteration during motion queries.
Note: This affects methods CharacterBody3D.move_and_slide(), PhysicsBody3D.move_and_collide(), PhysicsBody3D.test_move() and PhysicsServer3D.body_test_motion().
int physics/jolt_physics_3d/motion_queries/recovery_iterations = 4 🔗
The number of iterations to run when depenetrating during motion queries.
Note: This affects methods CharacterBody3D.move_and_slide(), PhysicsBody3D.move_and_collide(), PhysicsBody3D.test_move() and PhysicsServer3D.body_test_motion().
bool physics/jolt_physics_3d/motion_queries/use_enhanced_internal_edge_removal = true 🔗
If true, enables Jolt's enhanced internal edge removal during motion queries. This can help alleviate ghost collisions, but only with edges within a single body, meaning edges between separate bodies can still cause ghost collisions.
Note: This affects methods CharacterBody3D.move_and_slide(), PhysicsBody3D.move_and_collide(), PhysicsBody3D.test_move() and PhysicsServer3D.body_test_motion().
bool physics/jolt_physics_3d/queries/enable_ray_cast_face_index = false 🔗
If true, populates the face_index field in the results of PhysicsDirectSpaceState3D.intersect_ray(), also accessed through RayCast3D.get_collision_face_index(). If false, the face_index field will be left at its default value of -1.
Note: Enabling this setting will increase Jolt's memory usage for ConcavePolygonShape3D by around 25%.
bool physics/jolt_physics_3d/queries/use_enhanced_internal_edge_removal = false 🔗
If true, enables Jolt's enhanced internal edge removal during shape queries. This can help alleviate ghost collisions when using shape queries for things like character movement, but only with edges within a single body, meaning edges between separate bodies can still cause ghost collisions.
Note: This affects methods PhysicsDirectSpaceState3D.cast_motion(), PhysicsDirectSpaceState3D.collide_shape(), PhysicsDirectSpaceState3D.get_rest_info() and PhysicsDirectSpaceState3D.intersect_shape().
Note: Enabling this setting can cause certain shapes to be culled from the results entirely, but you will get at least one intersection per body.
bool physics/jolt_physics_3d/simulation/allow_sleep = true 🔗
If true, RigidBody3D nodes are allowed to go to sleep if their velocity is below the threshold defined in physics/jolt_physics_3d/simulation/sleep_velocity_threshold for the duration set in physics/jolt_physics_3d/simulation/sleep_time_threshold. This can improve physics simulation performance when there are non-moving RigidBody3D nodes, at the cost of some nodes possibly failing to wake up in certain scenarios. Consider disabling this temporarily to troubleshoot RigidBody3D nodes not moving when they should.
float physics/jolt_physics_3d/simulation/baumgarte_stabilization_factor = 0.2 🔗
How much of the position error of a RigidBody3D to fix during a physics step, where 0.0 is none and 1.0 is the full amount. This affects things like how quickly bodies depenetrate.
Note: Setting this value too high can make RigidBody3D nodes unstable.
float physics/jolt_physics_3d/simulation/body_pair_contact_cache_angle_threshold = 0.034906585 🔗
The maximum relative angle by which a body pair can move and still reuse the collision results from the previous physics step, in radians.
float physics/jolt_physics_3d/simulation/body_pair_contact_cache_distance_threshold = 0.001 🔗
The maximum relative distance by which a body pair can move and still reuse the collision results from the previous physics step, in meters.
bool physics/jolt_physics_3d/simulation/body_pair_contact_cache_enabled = true 🔗
If true, enables the body pair contact cache, which removes the need for potentially expensive collision detection when the relative orientation between two bodies hasn't changed much.
float physics/jolt_physics_3d/simulation/bounce_velocity_threshold = 1.0 🔗
The minimum velocity needed before a collision can be bouncy, in meters per second.
float physics/jolt_physics_3d/simulation/continuous_cd_max_penetration = 0.25 🔗
Fraction of a body's inner radius that may penetrate another body while using continuous collision detection.
float physics/jolt_physics_3d/simulation/continuous_cd_movement_threshold = 0.75 🔗
Fraction of a body's inner radius that the body must move per step to make use of continuous collision detection.
bool physics/jolt_physics_3d/simulation/generate_all_kinematic_contacts = false 🔗
If true, a RigidBody3D frozen with RigidBody3D.FREEZE_MODE_KINEMATIC is able to collide with other kinematic and static bodies, and therefore generate contacts for them.
Note: This setting can come at a heavy CPU and memory cost if you allow many/large frozen kinematic bodies with a non-zero RigidBody3D.max_contacts_reported to overlap with complex static geometry, such as ConcavePolygonShape3D or HeightMapShape3D.
float physics/jolt_physics_3d/simulation/penetration_slop = 0.02 🔗
How much bodies are allowed to penetrate each other, in meters.
int physics/jolt_physics_3d/simulation/position_steps = 2 🔗
Number of solver position iterations. The greater the number of iterations, the more accurate the simulation will be, at the cost of CPU performance.
float physics/jolt_physics_3d/simulation/sleep_time_threshold = 0.5 🔗
Time in seconds a RigidBody3D will spend below the sleep velocity threshold before going to sleep.
float physics/jolt_physics_3d/simulation/sleep_velocity_threshold = 0.03 🔗
The linear velocity of specific points on the bounding box of a RigidBody3D, below which it can be put to sleep, in meters per second. These points help capture both the linear and angular motion of a RigidBody3D.
float physics/jolt_physics_3d/simulation/soft_body_point_radius = 0.01 🔗
How big the points of a SoftBody3D are, in meters. A higher value can prevent behavior such as cloth laying perfectly flush against other surfaces and causing Z-fighting.
float physics/jolt_physics_3d/simulation/speculative_contact_distance = 0.02 🔗
Radius around physics bodies, inside which speculative contact points will be detected, in meters. This is mainly used to prevent tunneling/penetration for RigidBody3D nodes during simulation.
Note: Setting this too high may result in ghost collisions, as speculative contacts are based on the closest points during the collision detection step which may not be the actual closest points by the time the two bodies hit.
bool physics/jolt_physics_3d/simulation/use_enhanced_internal_edge_removal = true 🔗
If true, enables Jolt's enhanced internal edge removal for RigidBody3D. This can help alleviate ghost collisions when, for example, a RigidBody3D collides with the edges of two perfectly joined BoxShape3D. The removal only applies to edges internal to a single body, meaning edges between separate bodies can still cause ghost collisions.
int physics/jolt_physics_3d/simulation/velocity_steps = 10 🔗
Number of solver velocity iterations. The greater the number of iterations, the more accurate the simulation will be, at the cost of CPU performance.
Note: This needs to be at least 2 in order for friction to work, as friction is applied using the non-penetration impulse from the previous iteration.
int rendering/2d/batching/item_buffer_size = 16384 🔗
Maximum number of canvas item commands that can be batched into a single draw call.
int rendering/2d/batching/uniform_set_cache_size = 4096 🔗
Maximum number of uniform sets that will be cached by the 2D renderer when batching draw calls.
Note: Increasing this value can improve performance if the project renders many unique sprite textures every frame.
int rendering/2d/sdf/oversize = 1 🔗
控制 2D 帶符號距離場應該覆蓋原始視口中多大的區域。該 SDF 可以在 CanvasItem 著色器中取樣，用於 GPUParticles2D 碰撞。生成帶符號距離場時，較高的值能夠讓部分處於視口外的遮擋器也納入考慮範圍，但會以犧牲性能為代價。如果你注意到 LightOccluder2D 離開視口時粒子會穿過遮擋器，就可以增大這個設定。
每一側每個軸上都會新增指定的百分比。例如，預設設定為 120% 就會讓帶符號距離場覆蓋視口之外每側（上、右、下、左）各 20% 的大小。
注意：這個屬性僅在專案啟動時讀取。要在運作時修改 2D SDF 擴界百分比，請改用 RenderingServer.viewport_set_sdf_oversize_and_scale()。
int rendering/2d/sdf/scale = 1 🔗
2D 帶符號距離場的解析度縮放。設定為較大的值能夠得到更精確的帶符號距離場，也能讓帶符號距離場在相機移動時更為穩定，但代價是犧牲性能。預設值（50%）表示每個軸使用視口大小的一半大小的解析度，因此生成的 SDF 的圖元數量是視口圖元數量的 25%。
注意：該屬性僅在專案啟動時讀取。 要在運作時更改 2D SDF 解析度縮放，請改為使用 RenderingServer.viewport_set_sdf_oversize_and_scale()。
int rendering/2d/shadow_atlas/size = 2048 🔗
2D 陰影合集的大小，單位為圖元。值越高，Light2D 的陰影越精確，但性能消耗和顯存佔用也越大。指定的值會根據最接近的 2 的冪進行向上取整。
注意：這個屬性僅在專案啟動時讀取。要在運作時修改 2D 陰影合集的大小，請改用 RenderingServer.canvas_set_shadow_texture_size()。
bool rendering/2d/snap/snap_2d_transforms_to_pixel = false 🔗
If true, CanvasItem nodes will internally snap to full pixels. Useful for low-resolution pixel art games. Their position can still be sub-pixel, but the decimals will not have effect as the position is rounded. This can lead to a crisper appearance at the cost of less smooth movement, especially when Camera2D smoothing is enabled.
Note: This property is only read when the project starts. To toggle 2D transform snapping at runtime, use RenderingServer.viewport_set_snap_2d_transforms_to_pixel() on the root Viewport instead.
Note: Control nodes are snapped to the nearest pixel by default. This is controlled by gui/common/snap_controls_to_pixels.
Note: It is not recommended to use this setting together with rendering/2d/snap/snap_2d_vertices_to_pixel, as movement may appear even less smooth. Prefer only enabling this setting instead.
bool rendering/2d/snap/snap_2d_vertices_to_pixel = false 🔗
If true, vertices of CanvasItem nodes will snap to full pixels. Useful for low-resolution pixel art games. Only affects the final vertex positions, not the transforms. This can lead to a crisper appearance at the cost of less smooth movement, especially when Camera2D smoothing is enabled.
Note: This property is only read when the project starts. To toggle 2D vertex snapping at runtime, use RenderingServer.viewport_set_snap_2d_vertices_to_pixel() on the root Viewport instead.
Note: Control nodes are snapped to the nearest pixel by default. This is controlled by gui/common/snap_controls_to_pixels.
Note: It is not recommended to use this setting together with rendering/2d/snap/snap_2d_transforms_to_pixel, as movement may appear even less smooth. Prefer only enabling that setting instead.
int rendering/anti_aliasing/quality/msaa_2d = 0 🔗
Sets the number of multisample antialiasing (MSAA) samples to use for 2D/Canvas rendering (as a power of two). MSAA is used to reduce aliasing around the edges of polygons. A higher MSAA value results in smoother edges but can be significantly slower on some hardware, especially integrated graphics due to their limited memory bandwidth. This has no effect on shader-induced aliasing or texture aliasing.
Note: MSAA is only supported in the Forward+ and Mobile rendering methods, not Compatibility.
Note: This property is only read when the project starts. To set the number of 2D MSAA samples at runtime, set Viewport.msaa_2d or use RenderingServer.viewport_set_msaa_2d().
int rendering/anti_aliasing/quality/msaa_3d = 0 🔗
Sets the number of multisample antialiasing (MSAA) samples to use for 3D rendering (as a power of two). MSAA is used to reduce aliasing around the edges of polygons. A higher MSAA value results in smoother edges but can be significantly slower on some hardware, especially integrated graphics due to their limited memory bandwidth. See also rendering/scaling_3d/mode for supersampling, which provides higher quality but is much more expensive. This has no effect on shader-induced aliasing or texture aliasing.
Note: This property is only read when the project starts. To set the number of 3D MSAA samples at runtime, set Viewport.msaa_3d or use RenderingServer.viewport_set_msaa_3d().
int rendering/anti_aliasing/quality/screen_space_aa = 0 🔗
Sets the screen-space antialiasing mode for the default screen Viewport. Screen-space antialiasing works by selectively blurring edges in a post-process shader. It differs from MSAA which takes multiple coverage samples while rendering objects. Screen-space AA methods are typically faster than MSAA and will smooth out specular aliasing, but tend to make scenes appear blurry. The blurriness is partially counteracted by automatically using a negative mipmap LOD bias (see rendering/textures/default_filters/texture_mipmap_bias).
Another way to combat specular aliasing is to enable rendering/anti_aliasing/screen_space_roughness_limiter/enabled.
Note: Screen-space antialiasing is only supported in the Forward+ and Mobile rendering methods, not Compatibility.
Note: This property is only read when the project starts. To set the screen-space antialiasing mode at runtime, set Viewport.screen_space_aa on the root Viewport instead, or use RenderingServer.viewport_set_screen_space_aa().
float rendering/anti_aliasing/quality/smaa_edge_detection_threshold = 0.05 🔗
Sets the sensitivity to edges when using SMAA for antialiasing. Lower values will catch more edges, at a potentially higher performance cost.
Note: This property is only read when the project starts. There is currently no way to change this setting at run-time.
bool rendering/anti_aliasing/quality/use_debanding = false 🔗
If true, uses a fast dithering filter just before transforming floating point color values to integer color values to make banding significantly less visible. Debanding is applied at different steps of the rendering process depending on the rendering method and rendering/viewport/hdr_2d setting.
In some cases, debanding may introduce a slightly noticeable dithering pattern. It's recommended to enable debanding only when actually needed since the dithering pattern will make lossless-compressed screenshots larger.
Note: This property is only read when the project starts and configures RenderingServer.material_set_use_debanding() and Viewport.use_debanding of the root Viewport. When rendering/viewport/hdr_2d is disabled, you should additionally set the Viewport.use_debanding of other viewports in your project. To set debanding at run-time, the property that should be set depends on the renderer: Forward+ only uses Viewport.use_debanding and Mobile uses both RenderingServer.material_set_use_debanding() and Viewport.use_debanding.
bool rendering/anti_aliasing/quality/use_taa = false 🔗
Enables temporal antialiasing for the default screen Viewport. TAA works by jittering the camera and accumulating the images of the last rendered frames, motion vector rendering is used to account for camera and object motion. Enabling TAA can make the image blurrier, which is partially counteracted by automatically using a negative mipmap LOD bias (see rendering/textures/default_filters/texture_mipmap_bias).
Note: The implementation is not complete yet. Some visual instances such as particles and skinned meshes may show ghosting artifacts in motion.
Note: TAA is only supported in the Forward+ rendering method, not Mobile or Compatibility.
Note: This property is only read when the project starts. To set TAA at runtime, set Viewport.use_taa on the root Viewport instead, or use RenderingServer.viewport_set_use_taa().
float rendering/anti_aliasing/screen_space_roughness_limiter/amount = 0.25 🔗
注意：這個屬性僅在專案啟動時讀取。要在運作時控制螢幕空間粗糙度限制器，請呼叫 RenderingServer.screen_space_roughness_limiter_set_active()。
bool rendering/anti_aliasing/screen_space_roughness_limiter/enabled = true 🔗
如果為 true，則啟用空間篩檢程式以限制具有高頻細節的區域的粗糙度。這可以在一定程度上幫助減少鏡面反射鋸齒，儘管不如啟用 rendering/anti_aliasing/quality/use_taa。 該篩檢程式的性能成本很小，因此如果它對你的場景沒有明顯好處，請考慮禁用它。
注意：螢幕空間粗糙度限制器只支援 Forward+ 和 Mobile 算繪方式，不支援 Compatibility。
注意：這個屬性僅在專案啟動時讀取。要在運作時控制螢幕空間粗糙度限制器，請呼叫 RenderingServer.screen_space_roughness_limiter_set_active()。
float rendering/anti_aliasing/screen_space_roughness_limiter/limit = 0.18 🔗
注意：這個屬性僅在專案啟動時讀取。要在運作時控制螢幕空間粗糙度限制器，請呼叫 RenderingServer.screen_space_roughness_limiter_set_active()。
int rendering/camera/depth_of_field/depth_of_field_bokeh_quality = 1 🔗
設定景深效果的品質。品質越高，取樣的數量也越多，也會更慢，但看起來更平滑。
int rendering/camera/depth_of_field/depth_of_field_bokeh_shape = 1 🔗
設定景深的形狀。可以是方形、六邊形或圓形。方形最快。圓形最真實，但計算開銷也是最大的。
bool rendering/camera/depth_of_field/depth_of_field_use_jitter = false 🔗
如果為 true，則抖動 DOF 樣本以使效果稍微模糊，並隱藏低取樣速率建立的線條。當與較低的樣本數一起使用時，這可能會導致外觀略帶顆粒感。
String rendering/driver/depth_prepass/disable_for_vendors = "PowerVR,Mali,Adreno,Apple" 🔗
針對某些供應商禁用 rendering/driver/depth_prepass/enable。預設情況下會針對移動裝置禁用前置深度階段，因為移動裝置由於獨特的架構而無法從前置深度階段中獲益。
bool rendering/driver/depth_prepass/enable = true 🔗
如果 true，則在算繪 3D 材質之前先執行深度階段。當使用複雜的材質和照明時，這樣做能夠顯著提高過度繪製場景的性能。然而，如果場景中被遮擋的表面比較少，前置深度階段可能會降低性能。如果你的遊戲使用的是不易造成過度繪製的固定視角（例如俯視或者橫版視角），請考慮禁用前置深度階段，從而提升性能。這個設置可以在運作時更改，針對目前查看的場景來優化進行。
注意：前置深度階段僅在使用 Forward+ 或 Compatibility 算繪方法時支援。使用 Mobile 算繪方法時，不會執行前置深度階段。
int rendering/driver/threads/thread_model = 1 🔗
實驗性： This setting has several known bugs which can lead to crashing, especially when using particles or resizing the window. Not recommended for use in production at this stage.
The thread model to use for rendering. Rendering on a thread may improve performance, but synchronizing to the main thread can cause a bit more jitter.
Color rendering/environment/defaults/default_clear_color = Color(0.3, 0.3, 0.3, 1) 🔗
預設清屏顏色。Viewport 可使用其 Environment 進行覆蓋。具體見 Environment.background_mode 和 Environment.background_color。要通過程式碼更改此預設顏色，請使用 RenderingServer.set_default_clear_color()。
String rendering/environment/defaults/default_environment = "" 🔗
場景未指定其環境時作為退回環境使用的 Environment。無論是否設定了環境，載入場景時都會載入預設環境。如果你不依賴退回環境，那麼就不需要設定這個屬性。
bool rendering/environment/fog/use_legacy_blending = false 🔗
Enables legacy fog blending behavior from version 4.5 and earlier. This is intended for users who are developing on pre-4.6 versions and want to upgrade to 4.6 with the smallest possible change to their visuals.
int rendering/environment/glow/upscale_mode = 1 🔗
Sets how the glow effect is upscaled before being copied onto the screen. Linear is faster, but looks blocky. Bicubic is slower but looks smooth.
Note: rendering/environment/glow/upscale_mode is only effective when using the Forward+ or Mobile rendering methods, as Compatibility uses a different glow implementation.
int rendering/environment/glow/upscale_mode.mobile = 0 🔗
由於性能或驅動程式支援問題，rendering/environment/glow/upscale_mode 在移動裝置上的低端覆蓋項。
bool rendering/environment/screen_space_reflection/half_size = true 🔗
If true, screen-space reflections will be rendered at half size and then upscaled before being added to the scene. This is faster but may look pixelated or cause flickering. If false, screen-space reflections will be rendered at full size.
float rendering/environment/ssao/adaptive_target = 0.5 🔗
當 rendering/environment/ssao/quality 被設定為 Ultra 時所使用的品質目標。值為 0.0 時的品質和速度與 Medium 相似，而值為 1.0 時的品質比其他任何設定都要高得多，代價是犧牲性能。
int rendering/environment/ssao/blur_passes = 2 🔗
模糊通道數，在計算螢幕空間環境光遮蔽時使用。數值越大，外觀越平滑，但計算速度會變慢，高頻細節也會變少。
float rendering/environment/ssao/fadeout_from = 50.0 🔗
Distance at which the screen-space ambient occlusion effect starts to fade out. Use this hide ambient occlusion from far away.
float rendering/environment/ssao/fadeout_to = 300.0 🔗
Distance at which the screen-space ambient occlusion is fully faded out. Use this hide ambient occlusion from far away.
bool rendering/environment/ssao/half_size = true 🔗
如果為 true，螢幕空間環境光遮蔽將以一半大小算繪，然後在被新增到場景之前放大。這明顯更快，但可能會遺漏一些小細節。如果為 false，螢幕空間環境光遮蔽將以全尺寸算繪。
int rendering/environment/ssao/quality = 2 🔗
設定螢幕空間環境光遮蔽效果的品質。值越高，取樣數量越多，因此最終的品質也越高，但代價是消耗性能。設為 Ultra 會使用 rendering/environment/ssao/adaptive_target 設定。
float rendering/environment/ssil/adaptive_target = 0.5 🔗
當 rendering/environment/ssil/quality 被設定為 Ultra 時所使用的品質目標。值為 0.0 時的品質和速度與 Medium 相似，而值為 1.0 時的品質比其他任何設定都要高得多，代價是犧牲性能。使用適應目標時，性能損耗與場景複雜度成正比。
int rendering/environment/ssil/blur_passes = 4 🔗
模糊通道數，在計算螢幕空間間接光照時使用。數值越大，外觀越平滑，但計算速度會變慢，高頻細節也會變少。
float rendering/environment/ssil/fadeout_from = 50.0 🔗
Distance at which the screen-space indirect lighting effect starts to fade out. Use this to hide screen-space indirect lighting from far away.
float rendering/environment/ssil/fadeout_to = 300.0 🔗
Distance at which the screen-space indirect lighting is fully faded out. Use this to hide screen-space indirect lighting from far away.
bool rendering/environment/ssil/half_size = true 🔗
如果為 true，螢幕空間間接光照將以一半大小算繪，然後在被新增到場景之前放大。這明顯更快，但可能會遺漏一些小細節，部分物件的邊緣也會出現發光的情況。
int rendering/environment/ssil/quality = 2 🔗
設定螢幕空間間接照明效果的品質。值越高，取樣數量越多，因此最終的品質也越高，但代價是消耗性能。設為 Ultra 會使用 rendering/environment/ssil/adaptive_target 設定。
float rendering/environment/subsurface_scattering/subsurface_scattering_depth_scale = 0.01 🔗
對次表面散射效果套用深度進行縮放。較高的取值能夠讓燈光散射進網格的某些部分或者在螢幕空間中距離較近但深度更遠的其他網格。另見 rendering/environment/subsurface_scattering/subsurface_scattering_scale。
注意：這個屬性僅在專案啟動時讀取。要在運作時設定次表面散射深度縮放，請改為呼叫 RenderingServer.sub_surface_scattering_set_scale()。
int rendering/environment/subsurface_scattering/subsurface_scattering_quality = 1 🔗
設定次表面散射效果的品質。值越高越慢，但視覺效果更佳。影響 BaseMaterial3D.subsurf_scatter_enabled 為 true 以及設定了 SSS_STRENGTH 的 ShaderMaterial 的材質的算繪。
注意：這個屬性僅在專案啟動時讀取。要在運作時設定次表面散射品質，請改為呼叫 RenderingServer.sub_surface_scattering_set_quality()。
float rendering/environment/subsurface_scattering/subsurface_scattering_scale = 0.05 🔗
縮放對次表面散射效果進行取樣的距離。更改該值不會影響性能；但較高的值將導致明顯的偽影，因為樣本將變得明顯分散。較低的值會導致散射光的散佈更小。另見 rendering/environment/subsurface_scattering/subsurface_scattering_depth_scale。
注意：這個屬性僅在專案啟動時讀取。如果要在運作時設定次表面散射縮放，請改為呼叫 RenderingServer.sub_surface_scattering_set_scale()。
int rendering/environment/volumetric_fog/use_filter = 1 🔗
在積分之前啟用體積霧效果的篩選。這大大模糊了霧，減少了精細的細節，但也平滑了粗糙的邊緣和鋸齒偽影。當需要更多細節時禁用。
int rendering/environment/volumetric_fog/volume_depth = 64 🔗
沿著片段體素緩衝區的深度使用的切片數量，用於體積霧。較小的數位會更有效，但可能會導致相機移動時出現偽影。另請參閱 Environment.volumetric_fog_length。
int rendering/environment/volumetric_fog/volume_size = 64 🔗
用於確定相機 X 軸和 Y 軸中片段體素緩衝區大小的基本大小。最終大小按螢幕的長寬比縮放，因此實際值可能與設定的值不同。設定較大的大小，以獲得更詳細的霧；設定較小的大小，以獲得更好的性能。
String rendering/gl_compatibility/driver = "opengl3" 🔗
Sets the driver to be used by the renderer when using the Compatibility renderer. Editing this property has no effect in the default configuration, as first-party platforms each have platform-specific overrides. Use those overrides to configure the driver for each platform.
This can be overridden using the --rendering-driver <driver> command line argument.
Supported values are:
opengl3, OpenGL 3.3 on desktop platforms, OpenGL ES 3.0 on mobile platforms, WebGL 2.0 on web.opengl3_angle, OpenGL ES 3.0 using the ANGLE compatibility layer, supported on macOS (over native OpenGL) and Windows (over Direct3D 11).opengl3_es, OpenGL ES 3.0 on Linux/BSD.
Note: The availability of these options depends on whether the engine was compiled with support for them (determined by SCons options opengl3 and angle_libs).
Note: The actual rendering driver may be automatically changed by the engine as a result of a fallback, or a user-specified command line argument. To get the actual rendering driver that is used at runtime, use RenderingServer.get_current_rendering_driver_name() instead of reading this project setting's value.
String rendering/gl_compatibility/driver.android = "opengl3" 🔗
Android override for rendering/gl_compatibility/driver.
Only one option is supported:
opengl3, OpenGL ES 3.0 from native drivers.
String rendering/gl_compatibility/driver.ios = "opengl3" 🔗
iOS override for rendering/gl_compatibility/driver.
Only one option is supported:
opengl3, OpenGL ES 3.0 from native drivers.
String rendering/gl_compatibility/driver.linuxbsd = "opengl3" 🔗
LinuxBSD override for rendering/gl_compatibility/driver.
Two options are supported:
opengl3(default), OpenGL 3.3 from native drivers.opengl3_es, OpenGL ES 3.0 from native drivers. If rendering/gl_compatibility/fallback_to_gles is enabled, this is used as a fallback if OpenGL 3.3 is not supported.
String rendering/gl_compatibility/driver.macos = "opengl3" 🔗
macOS override for rendering/gl_compatibility/driver.
Two options are supported:
opengl3(default), OpenGL 3.3 from native drivers. If rendering/gl_compatibility/fallback_to_native is enabled, this is used as a fallback if ANGLE is configured as the preferred driver but not supported.opengl3_angle, OpenGL ES 3.0 using the ANGLE compatibility layer over native OpenGL drivers. If rendering/gl_compatibility/fallback_to_angle is enabled, this is used as a fallback if OpenGL 3.3 is not supported.
String rendering/gl_compatibility/driver.web = "opengl3" 🔗
Web override for rendering/gl_compatibility/driver.
Only one option is supported:
opengl3, WebGL 2.0. The underlying native API depends on the target OS, browser, and browser configuration.
String rendering/gl_compatibility/driver.windows = "opengl3" 🔗
Windows override for rendering/gl_compatibility/driver.
Two options are supported:
opengl3(default), OpenGL 3.3 from native drivers. If rendering/gl_compatibility/fallback_to_native is enabled, this is used as a fallback if ANGLE is configured as the preferred driver but not supported.opengl3_angle, OpenGL ES 3.0 using the ANGLE compatibility layer over native Direct3D 11 drivers. If rendering/gl_compatibility/fallback_to_angle is enabled, this is used as a fallback if OpenGL 3.3 is not supported. By default, ANGLE is used as the default driver for some devices listed in rendering/gl_compatibility/force_angle_on_devices.
bool rendering/gl_compatibility/fallback_to_angle = true 🔗
If true, the Compatibility renderer will fall back to ANGLE if native OpenGL is not supported or the device is listed in rendering/gl_compatibility/force_angle_on_devices.
Note: This setting is implemented only on Windows.
bool rendering/gl_compatibility/fallback_to_gles = true 🔗
If true, the Compatibility renderer will fall back to OpenGLES if desktop OpenGL is not supported.
Note: This setting is implemented only on Linux/X11.
bool rendering/gl_compatibility/fallback_to_native = true 🔗
If true, the Compatibility renderer will fall back to native OpenGL if ANGLE is not supported, or ANGLE dynamic libraries aren't found.
Note: This setting is implemented on macOS and Windows.
Array rendering/gl_compatibility/force_angle_on_devices 🔗
An Array of devices which should always use the ANGLE renderer.
Each entry is a Dictionary with the following keys: vendor and name. name can be set to * to add all devices with the specified vendor.
Note: This setting is implemented only on Windows.
int rendering/gl_compatibility/item_buffer_size = 16384 🔗
單次視口更新中能夠繪製的畫布專案命令的最大數量。如果發出了更多的算繪命令，就會被忽略。降低這個限制可以在頻寬有限的裝置上提高性能。如果你發現不是所有的對象都在同一影格中被繪製，請提高這個限制。
bool rendering/gl_compatibility/nvidia_disable_threaded_optimization = true 🔗
如果為 true，則禁用 NVIDIA 驅動的多執行緒優化功能，這個功能會導致大多數 OpenGL 套用的卡頓。
注意：這個設定僅在 Windows 上有效，因為在其他平臺上多執行緒優化預設都是關閉的。
bool rendering/global_illumination/gi/use_half_resolution = false 🔗
如果為 true，則以減半的解析度算繪 VoxelGI 和 SDFGI（Environment.sdfgi_enabled）緩衝區（例如，當視口大小為 1920×1080 時為 960×540）。當啟用 VoxelGI 或 SDFGI 時，這會顯著提高性能，但代價是多邊形邊緣上可能會出現可見的偽影。隨著視口解析度的增加，品質損失變得不那麼明顯。LightmapGI 算繪不受該設定的影響。
注意：只有在專案啟動時該屬性才會被讀取。要在運作時設定半解析度 GI，請改為呼叫 RenderingServer.gi_set_use_half_resolution()。
int rendering/global_illumination/sdfgi/frames_to_converge = 5 🔗
用於收斂帶符號距離場全域光照的影格數。值越高，得到的噪點越少，但完全收斂所需的時間也就越長。這意味著場景的全域光照處於過暗狀態的時間可能更長，在相機快速移動時尤為明顯。實際的收斂速度取決於算繪影格率。例如，預設設定為 30 影格，則 60 FPS 的算繪會讓 SDFGI 在 0.5 秒後完全收斂。另見 rendering/global_illumination/sdfgi/frames_to_update_lights 和 rendering/global_illumination/sdfgi/probe_ray_count。
注意：這個屬性僅在專案啟動時讀取。要在運作時控制 SDFGI 的收斂速度，請改為呼叫 RenderingServer.environment_set_sdfgi_frames_to_converge()。
int rendering/global_illumination/sdfgi/frames_to_update_lights = 2 🔗
帶符號距離場全域光照中動態光照更新所需的影格數。值越高，動態燈光造成的間接光照就會花費更多的時間來更新，但動態燈光數量較多時性能會更好。另見 rendering/global_illumination/sdfgi/frames_to_converge 和 rendering/global_illumination/sdfgi/probe_ray_count。
注意：僅影響 Light3D.light_bake_mode 為 Light3D.BAKE_DYNAMIC（預設值）的 Light3D 節點。讓不會移動的燈光使用 Light3D.BAKE_STATIC 烘焙模式能夠提升性能。
注意：這個屬性僅在專案啟動時讀取。要在運作時控制 SDFGI 的燈光更新速度，請改為呼叫 RenderingServer.environment_set_sdfgi_frames_to_update_light()。
int rendering/global_illumination/sdfgi/probe_ray_count = 1 🔗
計算帶符號距離場全域光照時每影格發出的射線。值越高，得到的噪點越少，但性能消耗也越大。另見 rendering/global_illumination/sdfgi/frames_to_converge 和 rendering/global_illumination/sdfgi/frames_to_update_lights。
注意：這個屬性僅在專案啟動時讀取。要在運作時控制 SDFGI 的品質，請改為呼叫 RenderingServer.environment_set_sdfgi_ray_count()。
int rendering/global_illumination/voxel_gi/quality = 0 🔗
要使用的 VoxelGI 品質。高品質下的光照更精確、反射品質更好，但算繪速度較慢。這個設定不影響烘焙資料，套用時不需要重新烘焙 VoxelGI。
注意：這個屬性僅在專案啟動時讀取。要在運作時控制 VoxelGI 的品質，請改為呼叫 RenderingServer.voxel_gi_set_quality()。
int rendering/lightmapping/bake_performance/max_rays_per_pass = 4 🔗
The maximum number of rays that can be thrown per pass when baking lightmaps with LightmapGI. Depending on the scene, adjusting this value may result in higher GPU utilization when baking lightmaps, leading to faster bake times.
Note: Using a value that is too high for your system can cause crashes due to the GPU being unresponsive for long periods of time, and the graphics driver being reset by the OS.
int rendering/lightmapping/bake_performance/max_rays_per_probe_pass = 64 🔗
The maximum number of rays that can be thrown per pass when baking dynamic object lighting in LightmapProbes with LightmapGI. Depending on the scene, adjusting this value may result in higher GPU utilization when baking lightmaps, leading to faster bake times.
Note: Using a value that is too high for your system can cause crashes due to the GPU being unresponsive for long periods of time, and the graphics driver being reset by the OS.
int rendering/lightmapping/bake_performance/max_transparency_rays = 8 🔗
The maximum number of retry rays that can be thrown per pass when hitting a transparent surface when baking lightmaps with LightmapGI. Depending on the scene, reducing this value may lead to faster bake times.
Note: Using a value that is too high for your system can cause crashes due to the GPU being unresponsive for long periods of time, and the graphics driver being reset by the OS.
int rendering/lightmapping/bake_performance/region_size = 512 🔗
The region size to use when baking lightmaps with LightmapGI. The specified value is rounded up to the nearest power of 2.
Note: Using a value that is too high for your system can cause crashes due to the GPU being unresponsive for long periods of time, and the graphics driver being reset by the OS.
int rendering/lightmapping/bake_quality/high_quality_probe_ray_count = 512 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_HIGH 時，用於在 LightmapProbe 中烘焙動態物件光照的光線數。
int rendering/lightmapping/bake_quality/high_quality_ray_count = 512 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_HIGH 時，用於使用 LightmapGI 烘焙光照貼圖的光線數。
int rendering/lightmapping/bake_quality/low_quality_probe_ray_count = 64 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_LOW 時，用於在 LightmapProbe 中烘焙動態物件光照的光線數。
int rendering/lightmapping/bake_quality/low_quality_ray_count = 32 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_LOW 時，用於使用 LightmapGI 烘焙光照貼圖的光線數。
int rendering/lightmapping/bake_quality/medium_quality_probe_ray_count = 256 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_MEDIUM 時，用於在 LightmapProbe 中烘焙動態物件光照的光線數。
int rendering/lightmapping/bake_quality/medium_quality_ray_count = 128 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_MEDIUM 時，用於使用 LightmapGI 烘焙光照貼圖的光線數。
int rendering/lightmapping/bake_quality/ultra_quality_probe_ray_count = 2048 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_ULTRA 時，用於在 LightmapProbe 中烘焙動態物件光照的光線數。
int rendering/lightmapping/bake_quality/ultra_quality_ray_count = 2048 🔗
當 LightmapGI.quality 為 LightmapGI.BAKE_QUALITY_ULTRA 時，用於使用 LightmapGI 烘焙光照貼圖的光線數。
int rendering/lightmapping/denoising/denoiser = 0 🔗
Denoiser tool used for denoising lightmaps.
Using OpenImageDenoise (OIDN) requires configuring a path to an OIDN executable in the editor settings at EditorSettings.filesystem/tools/oidn/oidn_denoise_path. OIDN can be downloaded from OpenImageDenoise's downloads page.
OIDN will use GPU acceleration when available. Unlike JNLM which uses compute shaders for acceleration, OIDN uses vendor-specific acceleration methods. For GPU acceleration to be available, the following libraries must be installed on the system depending on your GPU:
NVIDIA GPUs: CUDA libraries
AMD GPUs: HIP libraries
Intel GPUs: SYCL libraries
If no GPU acceleration is configured on the system, multi-threaded CPU-based denoising will be performed instead. This CPU-based denoising is significantly slower than the JNLM denoiser in most cases.
bool rendering/lightmapping/lightmap_gi/use_bicubic_filter = true 🔗
If true, applies a bicubic filter during lightmap sampling. This makes lightmaps look much smoother, at a moderate performance cost.
Note: The bicubic filter exaggerates the 'bleeding' effect that occurs when a lightmap's resolution is low enough.
float rendering/lightmapping/primitive_meshes/texel_size = 0.2 🔗
用於計算啟用了 PrimitiveMesh.add_uv2 的 PrimitiveMesh 資源上的 Mesh.lightmap_size_hint 的體素大小。
float rendering/lightmapping/probe_capture/update_speed = 15 🔗
The framerate-independent update speed when representing dynamic object lighting from LightmapProbes. Higher values make dynamic object lighting update faster. Higher values can prevent fast-moving objects from having "outdated" indirect lighting displayed on them, at the cost of possible flickering when an object moves from a bright area to a shaded area.
Note: This property is only read when the project starts. To adjust the BVH build quality at runtime, use RenderingServer.lightmap_set_probe_capture_update_speed().
bool rendering/lights_and_shadows/directional_shadow/16_bits = true 🔗
使用 16 位元的平行光陰影深度貼圖。啟用後，陰影的精度會降低，可能造成陰影失真，但能夠在部分裝置上提升性能。
int rendering/lights_and_shadows/directional_shadow/size = 4096 🔗
方向陰影的大小，單位為圖元。值越高，得到的陰影越清晰，但會以性能為代價。取值將四捨五入到最接近的 2 次冪。
int rendering/lights_and_shadows/directional_shadow/size.mobile = 2048 🔗
由於性能和驅動支援，在移動裝置上會對 rendering/lights_and_shadows/directional_shadow/size 以低配數值覆蓋。
int rendering/lights_and_shadows/directional_shadow/soft_shadow_filter_quality = 2 🔗
DirectionalLight3D 投射的陰影的品質設定。品質設定地越高，從陰影貼圖讀取的樣本越多，因此速度也越慢。品質設定地較低時，可能會導致陰影看起來有顆粒感。
注意：Soft Very Low 設定會自動將常數的陰影模糊乘以 0.75 倍，從而減少可見的噪點。這種對自動模糊的更改僅影響 Light3D.shadow_blur 中定義的常數模糊係數，不影響 DirectionalLight3D 的 Light3D.light_angular_distance 執行的可變模糊。
注意：Soft High 和 Soft Ultra 設定，會自動將常數的陰影模糊分別乘以 1.5 倍和 2 倍，從而更好地利用增加的樣本數。這種對模糊的提升還改善了動態物件陰影的穩定性。
int rendering/lights_and_shadows/directional_shadow/soft_shadow_filter_quality.mobile = 0 🔗
由於性能和驅動支援，在移動裝置上會對 rendering/lights_and_shadows/directional_shadow/soft_shadow_filter_quality 以低配數值覆蓋。
bool rendering/lights_and_shadows/positional_shadow/atlas_16_bits = true 🔗
使用 16 位的全向燈/聚光燈陰影深度貼圖。啟用後，陰影的精度會降低，可能造成陰影失真，但能夠在部分裝置上提升性能。
int rendering/lights_and_shadows/positional_shadow/atlas_quadrant_0_subdiv = 2 🔗
陰影圖集第一象限的細分量。有關更多信息，請參閱文件。
int rendering/lights_and_shadows/positional_shadow/atlas_quadrant_1_subdiv = 2 🔗
陰影圖集第二象限的細分量。更多資訊請參閱文件。
int rendering/lights_and_shadows/positional_shadow/atlas_quadrant_2_subdiv = 3 🔗
陰影圖集第三象限的細分量。有關更多信息，請參閱文件。
int rendering/lights_and_shadows/positional_shadow/atlas_quadrant_3_subdiv = 4 🔗
陰影圖集第四象限的細分量。有關更多信息，請參閱文件。
int rendering/lights_and_shadows/positional_shadow/atlas_size = 4096 🔗
用於 OmniLight3D 和 SpotLight3D 節點的陰影圖集的大小。有關更多信息，請參閱 文件。
int rendering/lights_and_shadows/positional_shadow/atlas_size.mobile = 2048 🔗
由於性能和驅動支援，在移動裝置上會對 rendering/lights_and_shadows/positional_shadow/atlas_size 以低配數值覆蓋。
int rendering/lights_and_shadows/positional_shadow/soft_shadow_filter_quality = 2 🔗
OmniLight3D 和 SpotLight3D 投射的陰影的品質設定。品質設定地越高，從陰影貼圖讀取的樣本越多，因此速度也越慢。品質設定地較低時，可能會導致陰影看起來有顆粒感。
注意：Soft Very Low 設定會自動將常數的陰影模糊乘以 0.75 倍，從而減少可見的噪點。這種對自動模糊的更改僅影響 Light3D.shadow_blur 中定義的常數模糊係數，不影響 DirectionalLight3D 的 Light3D.light_angular_distance 執行的可變模糊。
注意：Soft High 和 Soft Ultra 設定，會自動將陰影模糊分別乘以 1.5 倍和 2 倍，從而更好地利用增加的樣本數。這種對模糊的提升還改善了動態物件陰影的穩定性。
int rendering/lights_and_shadows/positional_shadow/soft_shadow_filter_quality.mobile = 0 🔗
由於性能和驅動支援，在移動裝置上會對 rendering/lights_and_shadows/positional_shadow/soft_shadow_filter_quality 以低配數值覆蓋。
bool rendering/lights_and_shadows/tighter_shadow_caster_culling = true 🔗
If true, items that cannot cast shadows into the view frustum will not be rendered into shadow maps.
This can increase performance.
bool rendering/lights_and_shadows/use_physical_light_units = false 🔗
允許對光源使用基於物理的單位。基於物理的單位往往比 Godot 使用的任意單位大得多，但它們可用於將 Godot 內的照明與真實世界的照明相配對。由於自然界中照明條件的動態範圍很大，Godot 在算繪之前會將曝光量烘焙到各種照明量中。大多數光源在運行時根據活動的 CameraAttributes 資源自動烘焙曝光，但 LightmapGI 和 VoxelGI 需要在烘焙時設定 CameraAttributes 資源以減少動態範圍。在運作時，Godot 將自動協調烘焙的曝光與活動的曝光，以確保照明保持一致。
float rendering/limits/cluster_builder/max_clustered_elements = 512 🔗
可以在相機視圖中一次算繪的集群元素（OmniLight3D + SpotLight3D + Decal + ReflectionProbe）的最大數量。如果相機視圖中存在更多的集群元素，其中一些將不會被算繪（導致在相機移動期間彈出）。在燈光和裝飾上啟用距離淡入淡出（Light3D.distance_fade_enabled、Decal.distance_fade_enabled）有助於避免達到該限制。
減小該值可能會提高某些設定的 GPU 性能，即使在專案中從未達到集群元素最大數量也是如此。
注意：該設定僅在使用 Forward+ 算繪方式時有效，對 Mobile 和 Compatibility 無效。
int rendering/limits/global_shader_variables/buffer_size = 65536 🔗
The maximum number of uniforms that can be used by the global shader uniform buffer. Each item takes up one slot. In other words, a single uniform float and a uniform vec4 will take the same amount of space in the buffer.
Note: When using the Compatibility renderer, most mobile devices (and all web exports) will be limited to a maximum size of 1024 due to hardware constraints.
int rendering/limits/opengl/max_lights_per_object = 8 🔗
每個物件可算繪的全向燈和聚光燈的最大數量。預設值為 8，這意味著每個表面最多可受到 8 個全向燈和 8 個聚光燈影響。這進一步受到硬體支援和 rendering/limits/opengl/max_renderable_lights 的限制。將該設定得較低會略微減少記憶體使用，可能會減少著色器編譯時間，並可能導致在低端、移動或 Web 裝置上的算繪速度更快。
注意：該設定僅支援 Compatibility 算繪方式，不支援 Forward+ 和 Mobile。
int rendering/limits/opengl/max_renderable_elements = 65536 🔗
一影格中可算繪的最大元素數。如果每影格可見的元素多於此，則不會繪製它們。請記住，元素指的是網格表面，而不是網格本身。將此值設定得較低會略微減少記憶體使用量，並可能減少著色器編譯時間，尤其是在 Web 上。對於大多數用途，預設值是合適的，但在 Web 匯出時可考慮盡可能降低。
注意：該設定僅在支援 Compatibility 算繪方式，不支援 Forward+ 和 Mobile。
int rendering/limits/opengl/max_renderable_lights = 32 🔗
一影格中可算繪的定位燈的最大數量。如果使用的燈多於此數量，則它們將被忽略。將此值設定得較低會略微減少記憶體使用量，並可能減少著色器編譯時間，尤其是在 Web 上。對於大多數用途，預設值是合適的，但在 Web 匯出時可考慮盡可能降低。
注意：該設定僅在支援 Compatibility 算繪方式，不支援 Forward+ 和 Mobile。
int rendering/limits/spatial_indexer/threaded_cull_minimum_instances = 1000 🔗
The minimum number of instances that must be present in a scene to enable culling computations on multiple threads. If a scene has fewer instances than this number, culling is done on a single thread.
int rendering/limits/spatial_indexer/update_iterations_per_frame = 10 🔗
There is currently no description for this property. Please help us by contributing one!
float rendering/limits/time/time_rollover_secs = 3600 🔗
Maximum time (in seconds) before the TIME shader built-in variable rolls over. The TIME variable increments by delta each frame, and when it exceeds this value, it rolls over to 0.0. Since large floating-point values are less precise than small floating-point values, this should be set as low as possible to maximize the precision of the TIME built-in variable in shaders. This is especially important on mobile platforms where precision in shaders is significantly reduced. However, if this is set too low, shader animations may appear to restart from the beginning while the project is running.
On desktop platforms, values below 4096 are recommended, ideally below 2048. On mobile platforms, values below 64 are recommended, ideally below 32.
float rendering/mesh_lod/lod_change/threshold_pixels = 1.0 🔗
The automatic LOD bias to use for meshes rendered within the ReflectionProbe. Higher values will use less detailed versions of meshes that have LOD variations generated. If set to 0.0, automatic LOD is disabled. Increase rendering/mesh_lod/lod_change/threshold_pixels to improve performance at the cost of geometry detail.
Note: Depending on the mesh's attributes (vertex colors, blend shapes, ...), a mesh may have fewer levels of LOD generated to avoid visible distortion of the mesh once it is affected by vertex colors or blend shapes. Meshes with a very low vertex count will also not have any LODs generated, which means this setting will not affect them at all. In general, this setting makes the largest impact on static meshes with a high vertex count.
Note: rendering/mesh_lod/lod_change/threshold_pixels does not affect GeometryInstance3D visibility ranges (also known as "manual" LOD or hierarchical LOD).
Note: This property is only read when the project starts. To adjust the automatic LOD threshold at runtime, set Viewport.mesh_lod_threshold on the root Viewport.
int rendering/occlusion_culling/bvh_build_quality = 2 🔗
算繪遮擋剔除緩衝區時使用的 BVH 品質。值越高，得到的遮擋剔除越精確，但代價是 CPU 使用率也越高。另見 rendering/occlusion_culling/occlusion_rays_per_thread。
注意：這個屬性僅在專案啟動時讀取。如果要在運作時調整 BVH 建構品質，請使用 RenderingServer.viewport_set_occlusion_culling_build_quality()。
bool rendering/occlusion_culling/jitter_projection = true 🔗
If true, the projection used for rendering the occlusion buffer will be jittered. This can help prevent objects being incorrectly culled when visible through small gaps.
int rendering/occlusion_culling/occlusion_rays_per_thread = 512 🔗
每個 CPU 執行緒所追蹤的剔除射線數量。更高的值將導致更準確的遮擋剔除，但代價是更高的 CPU 使用率。遮擋剔除緩衝區的圖元數大致等於 occlusion_rays_per_thread * number_of_logical_cpu_cores，因此它取決於系統的 CPU。因此，內核較少的 CPU 將使用較低的解析度，來嘗試保持跨裝置的性能成本。另見 rendering/occlusion_culling/bvh_build_quality。
注意：這個屬性僅在專案啟動時讀取。如果要在運作時調整每個執行緒所追蹤的剔除射線數量，請使用 RenderingServer.viewport_set_occlusion_rays_per_thread()。
bool rendering/occlusion_culling/use_occlusion_culling = false 🔗
如果為 true，則 OccluderInstance3D 節點在根視口的 3D 遮擋剔除中可用。對於自訂視口，必須改為將 Viewport.use_occlusion_culling 設為 true。
注意：啟用遮擋剔除會消耗 CPU 資源。請只在打算使用時啟用遮擋剔除。阻擋視線的物件很少或根本不存在的大型開放場景通常不會因遮擋剔除而獲得什麼優化。相對於遮擋剔除，大型開放場景通常能夠從網格 LOD 和可見範圍（GeometryInstance3D.visibility_range_begin 和 GeometryInstance3D.visibility_range_end）中獲益。
int rendering/reflections/reflection_atlas/reflection_count = 64 🔗
儲存在反射合集中的立方體貼圖數量。場景中的 ReflectionProbe 數量受此數量限制。數字越高，所需的顯存越多。
int rendering/reflections/reflection_atlas/reflection_size = 256 🔗
ReflectionProbe 的立方體貼圖面的大小。數位越大，所需的顯存越多，並且反射探針的更新也可能變得越慢。
int rendering/reflections/reflection_atlas/reflection_size.mobile = 128 🔗
移動裝置上 rendering/reflections/reflection_atlas/reflection_size 的低端覆蓋項，出於性能問題或驅動程式支援的考慮。
bool rendering/reflections/sky_reflections/fast_filter_high_quality = false 🔗
使用快速篩選演算法的更高品質變體。明顯比使用預設品質慢，但會產生更平滑的反射。只應在場景特別詳細時使用。
int rendering/reflections/sky_reflections/ggx_samples = 32 🔗
設定在對 Sky 和 ReflectionProbe 使用重要性取樣時要採用的樣本數。較高的值將導致更平滑、更高品質的反射，但會增加計算輻照度對應的時間。一般來說，更簡單的低動態範圍環境需要更少的樣本，而 HDR 環境和具有高細節層次的環境需要更多的樣本。
int rendering/reflections/sky_reflections/ggx_samples.mobile = 16 🔗
移動裝置上 rendering/reflections/sky_reflections/ggx_samples 的低端覆蓋項，出於性能問題或驅動程式支援的考慮。
int rendering/reflections/sky_reflections/roughness_layers = 8 🔗
使用重要性取樣時，限制在輻照度對應中使用的層數。較低的數字會稍微快一些，並且會佔用較少的 VRAM。
bool rendering/reflections/sky_reflections/texture_array_reflections = true 🔗
If true, uses texture arrays instead of mipmaps for reflection probes and panorama backgrounds (sky). This reduces jitter noise and upscaling artifacts on reflections, but is significantly slower to compute and uses rendering/reflections/sky_reflections/roughness_layers times more memory.
Note: Texture array reflections are always disabled on macOS on Intel GPUs due to driver bugs.
bool rendering/reflections/sky_reflections/texture_array_reflections.mobile = false 🔗
移動裝置上 rendering/reflections/sky_reflections/texture_array_reflections 的低端覆蓋項，出於性能問題或驅動程式支援的考慮。
bool rendering/reflections/specular_occlusion/enabled = true 🔗
If true, reduces reflections based on ambient light.
String rendering/renderer/rendering_method = "forward_plus" 🔗
Sets the renderer that will be used by the project. Options are:
forward_plus (Forward+): High-end renderer designed for desktop devices. Has a higher base overhead, but scales well with complex scenes. Not suitable for older devices or mobile.
mobile (Mobile): Modern renderer designed for mobile devices. Has a lower base overhead than Forward+, but does not scale as well to large scenes with many elements.
gl_compatibility (Compatibility): Low-end renderer designed for older devices. Based on the limitations of the OpenGL 3.3 / OpenGL ES 3.0 / WebGL 2 APIs. Lighting calculations are performed on nonlinear sRGB-encoded color data, which produces inaccurate results that may look acceptable for some games.
This can be overridden using the --rendering-method <method> command line argument.
Note: The actual rendering method may be automatically changed by the engine as a result of a fallback, or a user-specified command line argument. To get the actual rendering method that is used at runtime, use RenderingServer.get_current_rendering_method() instead of reading this project setting's value.
String rendering/renderer/rendering_method.mobile = "mobile" 🔗
移動裝置的 rendering/renderer/rendering_method 覆蓋項。
String rendering/renderer/rendering_method.web = "gl_compatibility" 🔗
Web 平臺的 rendering/renderer/rendering_method 覆蓋項。
int rendering/rendering_device/d3d12/agility_sdk_version = 618 🔗
Version code of the Direct3D 12 Agility SDK to use (D3D12SDKVersion). This must match the minor version that is installed next to the editor binary and in the export templates directory for the current editor version. For example, if you have 1.618.5 installed, you need to input 618 here.
int rendering/rendering_device/d3d12/max_resource_descriptors = 65536 🔗
The number of entries in the resource descriptor heap the Direct3D 12 rendering driver uses for most rendering operations.
Depending on the complexity of scenes, this value may be lowered or may need to be raised.
int rendering/rendering_device/d3d12/max_sampler_descriptors = 1024 🔗
The number of entries in the sampler descriptor heap the Direct3D 12 rendering driver uses for most rendering operations.
Depending on the complexity of scenes, this value may be lowered or may need to be raised.
String rendering/rendering_device/driver = "vulkan" 🔗
Sets the driver to be used by the renderer when using a RenderingDevice-based renderer like the Forward+ or Mobile renderers. Editing this property has no effect in the default configuration, as first-party platforms each have platform-specific overrides. Use those overrides to configure the driver for each platform.
This can be overridden using the --rendering-driver <driver> command line argument.
Supported values are:
metal, Metal (supported on Apple Silicon Macs and iOS).vulkan, Vulkan (supported on all desktop and mobile platforms).d3d12, Direct3D 12 (supported on Windows).
Note: The availability of these options depends on whether the engine was compiled with support for them (determined by SCons options vulkan, metal, and d3d12).
Note: If a given platform has no registered drivers, it can fall back to the Compatibility renderer (OpenGL 3) if rendering/rendering_device/fallback_to_opengl3 is enabled. This fallback happens automatically for the Web platform regardless of that property.
Note: The actual rendering driver may be automatically changed by the engine as a result of a fallback, or a user-specified command line argument. To get the actual rendering driver that is used at runtime, use RenderingServer.get_current_rendering_driver_name() instead of reading this project setting's value.
String rendering/rendering_device/driver.android = "vulkan" 🔗
Android override for rendering/rendering_device/driver.
Only one option is supported:
vulkan, Vulkan from native drivers.
Note: If Vulkan was disabled at compile time, there is no alternative RenderingDevice driver.
String rendering/rendering_device/driver.ios = "metal" 🔗
iOS override for rendering/rendering_device/driver.
Two options are supported:
metal(default), Metal from native drivers.vulkan, Vulkan over Metal via MoltenVK.
String rendering/rendering_device/driver.linuxbsd = "vulkan" 🔗
LinuxBSD override for rendering/rendering_device/driver.
Only one option is supported:
vulkan, Vulkan from native drivers.
Note: If Vulkan was disabled at compile time, there is no alternative RenderingDevice driver.
String rendering/rendering_device/driver.macos = "metal" 🔗
macOS override for rendering/rendering_device/driver.
Two options are supported:
metal(default), Metal from native drivers, only supported on Apple Silicon Macs. On Intel Macs, it will automatically fall back tovulkanas Metal support is not implemented.vulkan, Vulkan over Metal via MoltenVK, supported on both Apple Silicon and Intel Macs.
String rendering/rendering_device/driver.visionos = "metal" 🔗
visionOS override for rendering/rendering_device/driver.
Only one option is supported:
metal(default), Metal from native drivers.
String rendering/rendering_device/driver.windows = "vulkan" 🔗
Windows override for rendering/rendering_device/driver.
Two options are supported:
vulkan(default), Vulkan from native drivers. If rendering/rendering_device/fallback_to_vulkan is enabled, this is used as a fallback if Direct3D 12 is not supported.d3d12, Direct3D 12 from native drivers. If rendering/rendering_device/fallback_to_d3d12 is enabled, this is used as a fallback if Vulkan is not supported.
Note: Starting with Godot 4.6, new projects are configured by default to use d3d12 on Windows. Projects created before Godot 4.6 keep vulkan for compatibility reasons, but it is recommended to switch them manually to d3d12.
bool rendering/rendering_device/fallback_to_d3d12 = true 🔗
If true, the Forward+ renderer will fall back to Direct3D 12 if Vulkan is not supported. The fallback is always attempted regardless of this setting if Vulkan driver support was disabled at compile time.
Note: This setting is implemented only on Windows.
bool rendering/rendering_device/fallback_to_opengl3 = true 🔗
If true, the Forward+ renderer will fall back to OpenGL 3 if Direct3D 12, Metal, and Vulkan are not supported.
Note: This setting is implemented on Windows, Android, macOS, iOS, and Linux/X11.
bool rendering/rendering_device/fallback_to_vulkan = true 🔗
If true, the Forward+ renderer will fall back to Vulkan if Direct3D 12 (on Windows) or Metal (on macOS x86_64) are not supported. The fallback is always attempted regardless of this setting if Direct3D 12 (Windows) or Metal (macOS) driver support was disabled at compile time.
Note: This setting is implemented on Windows and macOS.
bool rendering/rendering_device/pipeline_cache/enable = true 🔗
Enable the pipeline cache that is saved to disk if the graphics API supports it.
Note: This property is unable to control the pipeline caching the GPU driver itself does. Only turn this off along with deleting the contents of the driver's cache if you wish to simulate the experience a user will get when starting the game for the first time.
float rendering/rendering_device/pipeline_cache/save_chunk_size_mb = 3.0 🔗
決定管線快取保存到磁片的間隔。值越低，保存地越頻繁。
int rendering/rendering_device/staging_buffer/block_size_kb = 256 🔗
The size of a block allocated in the staging buffers. Staging buffers are the intermediate resources the engine uses to upload or download data to the GPU. This setting determines the max amount of data that can be transferred in a copy operation. Increasing this will result in faster data transfers at the cost of extra memory.
Note: This property is only read when the project starts. There is currently no way to change this value at run-time.
int rendering/rendering_device/staging_buffer/max_size_mb = 128 🔗
The maximum amount of memory allowed to be used by staging buffers. If the amount of data being uploaded or downloaded exceeds this amount, the GPU will stall and wait for previous frames to finish.
Note: This property is only read when the project starts. There is currently no way to change this value at run-time.
int rendering/rendering_device/staging_buffer/texture_download_region_size_px = 64 🔗
The region size in pixels used to download texture data from the GPU when using methods like RenderingDevice.texture_get_data_async().
Note: This property's upper limit is controlled by rendering/rendering_device/staging_buffer/block_size_kb and whether it's possible to allocate a single block of texture data with this region size in the format that is requested.
Note: This property is only read when the project starts. There is currently no way to change this value at run-time.
int rendering/rendering_device/staging_buffer/texture_upload_region_size_px = 64 🔗
The region size in pixels used to upload texture data from the GPU when using methods like RenderingDevice.texture_update().
Note: This property's upper limit is controlled by rendering/rendering_device/staging_buffer/block_size_kb and whether it's possible to allocate a single block of texture data with this region size in the format that is requested.
Note: This property is only read when the project starts. There is currently no way to change this value at run-time.
int rendering/rendering_device/vsync/frame_queue_size = 2 🔗
The number of frames to track on the CPU side before stalling to wait for the GPU.
Try the V-Sync Simulator, an interactive interface that simulates presentation to better understand how it is affected by different variables under various conditions.
Note: This property is only read when the project starts. There is currently no way to change this value at run-time.
int rendering/rendering_device/vsync/swapchain_image_count = 3 🔗
The number of images the swapchain will consist of (back buffers + front buffer).
2 corresponds to double-buffering and 3 to triple-buffering.
Double-buffering may give you the lowest lag/latency but if V-Sync is on and the system can't render at 60 fps, the framerate will go down in multiples of it (e.g. 30 fps, 15, 7.5, etc.). Triple buffering gives you higher framerate (specially if the system can't reach a constant 60 fps) at the cost of up to 1 frame of latency, with DisplayServer.VSYNC_ENABLED (FIFO).
Use double-buffering with DisplayServer.VSYNC_ENABLED. Triple-buffering is a must if you plan on using DisplayServer.VSYNC_MAILBOX mode.
Try the V-Sync Simulator, an interactive interface that simulates presentation to better understand how it is affected by different variables under various conditions.
Note: Changes to this setting will only be applied on startup or when the swapchain is recreated (e.g. when setting the V-Sync mode).
Note: Some platforms may restrict the actual value.
int rendering/rendering_device/vulkan/max_descriptors_per_pool = 64 🔗
The number of descriptors per pool. Godot's Vulkan backend uses linear pools for descriptors that will be created and destroyed within a single frame. Instead of destroying every single descriptor every frame, they all can be destroyed at once by resetting the pool they belong to.
A larger number is more efficient up to a limit, after that it will only waste RAM (maximum efficiency is achieved when there is no more than 1 pool per frame). A small number could end up with one pool per descriptor, which negatively impacts performance.
Note: Changing this property requires a restart to take effect.
float rendering/scaling_3d/fsr_sharpness = 0.2 🔗
決定使用 FSR 放大模式時放大圖像的清晰度。每個整數的銳度減半。值從 0.0（最銳利）到 2.0。高於 2.0 的值不會產生明顯的差異。
int rendering/scaling_3d/mode = 0 🔗
設定縮放 3D 模式。雙線性縮放以不同的解析度算繪，以對視口進行欠取樣或超取樣。FidelityFX 超解析度（FidelityFX Super Resolution） 1.0，縮寫為 FSR，是一種放大技術，可通過使用一種空間感知放大演算法，以快速畫面播放速率生成高品質圖像。FSR 比雙線性稍微貴一點，但它產生的圖像品質明顯更高。在特別低端的 GPU 上，FSR 的性價比過低（與使用具有一個稍高解析度縮放以配對性能的雙線性縮放相比）。
注意：FSR 只在使用 Forward+ 算繪方式時有效，對 Mobile 或 Compatibility 無效。如果使用不相容的算繪方法，FSR 將退回到雙線性縮放。
int rendering/scaling_3d/mode.ios 🔗
iOS override for rendering/scaling_3d/mode. This allows selecting the MetalFX spatial and MetalFX temporal scaling modes, which are exclusive to platforms where the Metal rendering driver is used.
int rendering/scaling_3d/mode.macos 🔗
macOS override for rendering/scaling_3d/mode. This allows selecting the MetalFX spatial and MetalFX temporal scaling modes, which are exclusive to platforms where the Metal rendering driver is used.
float rendering/scaling_3d/scale = 1.0 🔗
Scales the 3D render buffer based on the viewport size uses an image filter specified in rendering/scaling_3d/mode to scale the output image to the full viewport size. Values lower than 1.0 can be used to speed up 3D rendering at the cost of quality (undersampling). Values greater than 1.0 are only valid for bilinear mode and can be used to improve 3D rendering quality at a high performance cost (supersampling). See also rendering/anti_aliasing/quality/msaa_3d for multi-sample antialiasing, which is significantly cheaper but only smooths the edges of polygons.
Note: When using the Nearest scaling mode, to avoid uneven pixel scaling, it's highly recommended to use a value equal to an integer divisor with a dividend of 1. For example, it's best to use a scale of 0.5 (1/2), 0.3333 (1/3), 0.25 (1/4), 0.2 (1/5), and so on.
bool rendering/shader_compiler/shader_cache/compress = true 🔗
There is currently no description for this property. Please help us by contributing one!
bool rendering/shader_compiler/shader_cache/enabled = true 🔗
啟用著色器快取，編譯後的著色器會儲存在磁片上，防止在下次需要該著色器時因為編譯著色器而帶來卡頓。
bool rendering/shader_compiler/shader_cache/strip_debug = false 🔗
There is currently no description for this property. Please help us by contributing one!
bool rendering/shader_compiler/shader_cache/strip_debug.release = true 🔗
There is currently no description for this property. Please help us by contributing one!
bool rendering/shader_compiler/shader_cache/use_zstd_compression = true 🔗
There is currently no description for this property. Please help us by contributing one!
bool rendering/shading/overrides/force_lambert_over_burley = false 🔗
如果為 true，則使用速度更快但品質較低的 Lambert 材質照明模型，不使用 Burley 模型。
bool rendering/shading/overrides/force_lambert_over_burley.mobile = true 🔗
由於性能問題或驅動程式支援，移動裝置上用於 rendering/shading/overrides/force_lambert_over_burley 的低端覆蓋。
bool rendering/shading/overrides/force_vertex_shading = false 🔗
If true, forces vertex shading for all rendering. This can increase performance a lot, but also reduces quality immensely. Can be used to optimize performance on low-end mobile devices.
int rendering/textures/basis_universal/rdo_dict_size = 1024 🔗
The dictionary size for Rate-Distortion Optimization (RDO) when importing textures as Basis Universal and when RDO is enabled, ranging from 64 to 65536. Higher values reduce the file sizes further, but make encoding times significantly longer.
bool rendering/textures/basis_universal/zstd_supercompression = true 🔗
If true, enables Zstandard supercompression to reduce file size when importing textures as Basis Universal.
Note: Basis Universal textures need to be compressed to gain the benefit of smaller file sizes, otherwise they are as large as VRAM-compressed textures.
int rendering/textures/basis_universal/zstd_supercompression_level = 6 🔗
Specify the compression level for Basis Universal Zstandard supercompression, ranging from 1 to 22.
int rendering/textures/canvas_textures/default_texture_filter = 1 🔗
The default texture filtering mode to use for CanvasItems built-in texture. In shaders, this texture is accessed as TEXTURE.
Note: For pixel art aesthetics, see also rendering/2d/snap/snap_2d_vertices_to_pixel and rendering/2d/snap/snap_2d_transforms_to_pixel.
int rendering/textures/canvas_textures/default_texture_repeat = 0 🔗
The default texture repeating mode to use for CanvasItems built-in texture. In shaders, this texture is accessed as TEXTURE.
int rendering/textures/decals/filter = 3 🔗
Decal 節點的篩選品質。使用 Anisotropic（各向異性）篩選模式時，各向異性篩選級別由 rendering/textures/default_filters/anisotropic_filtering_level 控制。
int rendering/textures/default_filters/anisotropic_filtering_level = 2 🔗
Sets the maximum number of samples to take when using anisotropic filtering on textures (as a power of two). A higher sample count will result in sharper textures at oblique angles, but is more expensive to compute. A value of 0 forcibly disables anisotropic filtering, even on materials where it is enabled.
The anisotropic filtering level also affects decals and light projectors if they are configured to use anisotropic filtering. See rendering/textures/decals/filter and rendering/textures/light_projectors/filter.
Note: In 3D, for this setting to have an effect, set BaseMaterial3D.texture_filter to BaseMaterial3D.TEXTURE_FILTER_LINEAR_WITH_MIPMAPS_ANISOTROPIC or BaseMaterial3D.TEXTURE_FILTER_NEAREST_WITH_MIPMAPS_ANISOTROPIC on materials.
Note: In 2D, for this setting to have an effect, set CanvasItem.texture_filter to CanvasItem.TEXTURE_FILTER_LINEAR_WITH_MIPMAPS_ANISOTROPIC or CanvasItem.TEXTURE_FILTER_NEAREST_WITH_MIPMAPS_ANISOTROPIC on the CanvasItem node displaying the texture (or in CanvasTexture). However, anisotropic filtering is rarely useful in 2D, so only enable it for textures in 2D if it makes a meaningful visual difference.
Note: This property is only read when the project starts. To change the anisotropic filtering level at runtime, set Viewport.anisotropic_filtering_level on the root Viewport instead.
float rendering/textures/default_filters/texture_mipmap_bias = 0.0 🔗
Affects the final texture sharpness by reading from a lower or higher mipmap (also called "texture LOD bias"). Negative values make mipmapped textures sharper but grainier when viewed at a distance, while positive values make mipmapped textures blurrier (even when up close).
Enabling temporal antialiasing (rendering/anti_aliasing/quality/use_taa) will automatically apply a -0.5 offset to this value, while enabling FXAA (rendering/anti_aliasing/quality/screen_space_aa) will automatically apply a -0.25 offset to this value. If both TAA and FXAA are enabled at the same time, an offset of -0.75 is applied to this value.
Note: If rendering/scaling_3d/scale is lower than 1.0 (exclusive), rendering/textures/default_filters/texture_mipmap_bias is used to adjust the automatic mipmap bias which is calculated internally based on the scale factor. The formula for this is log2(scaling_3d_scale) + mipmap_bias.
Note: This property is only supported in the Forward+ and Mobile renderers, not Compatibility. In Compatibility, this property is always treated as if it was set to 0.0.
bool rendering/textures/default_filters/use_nearest_mipmap_filter = false 🔗
如果為 true，則在使用 mipmap 時使用最近鄰 mipmap 篩選（也稱為“雙線性篩選”），這將導致在 mipmap 階段之間出現可見的接縫。因為使用的記憶體頻寬更少，這可能會提高移動裝置的性能。如果為 false，則使用線性 mipmap 篩選（也稱為“三線性篩選”）。
注意：只有在專案啟動時該屬性才會被讀取。目前無法在運作時更改該設定。
int rendering/textures/light_projectors/filter = 3 🔗
OmniLight3D 和 SpotLight3D 投影器的篩選品質。使用 Anisotropic（各向異性）篩選模式時，各向異性篩選級別由 rendering/textures/default_filters/anisotropic_filtering_level 控制。
bool rendering/textures/lossless_compression/force_png = false 🔗
如果為 true，紋理匯入器將使用 PNG 格式匯入無損紋理。否則預設使用 WebP。
bool rendering/textures/vram_compression/cache_gpu_compressor = true 🔗
If true, the GPU texture compressor will cache the local RenderingDevice and its resources (shaders and pipelines), making subsequent imports faster at the cost of increased memory usage.
bool rendering/textures/vram_compression/compress_with_gpu = true 🔗
If true, the texture importer will utilize the GPU for compressing textures, improving the import time of large images.
Note: This only functions on a device which supports either Vulkan, Direct3D 12, or Metal as a rendering driver.
Note: Currently this only affects certain compressed formats (BC1, BC3, BC4, BC5, and BC6), all of which are exclusive to desktop platforms and consoles.
bool rendering/textures/vram_compression/import_etc2_astc = false 🔗
If true, the texture importer will import VRAM-compressed textures using the Ericsson Texture Compression 2 algorithm for lower quality textures and normal maps and Adaptable Scalable Texture Compression algorithm for high quality textures (in 4×4 block size).
Note: This setting is an override. The texture importer will always import the format the host platform needs, even if this is set to false.
Note: Changing this setting does not impact textures that were already imported before. To make this setting apply to textures that were already imported, exit the editor, remove the .godot/imported/ folder located inside the project folder then restart the editor (see application/config/use_hidden_project_data_directory).
bool rendering/textures/vram_compression/import_s3tc_bptc = false 🔗
如果為 true，紋理匯入器將使用 S3 紋理壓縮演算法（DXT1-5）匯入 VRAM 壓縮紋理以獲得較低品質的紋理；並使用 BPTC 演算法（BC6H 和 BC7）匯入高品質紋理。該演算法僅在 PC 桌面平臺和主機平臺上受支援。
注意：這是設定覆蓋項。即便設為 false，紋理匯入器也始終會匯入宿主平臺所需的格式。
注意：更改該設定不會影響之前已經匯入的紋理。要使該設定套用於已匯入的紋理，請退出編輯器，移除位於專案檔案夾內的 .godot/imported/ 資料夾，然後重新啟動編輯器（請參閱 application/config/use_hidden_project_data_directory）。
int rendering/textures/webp_compression/compression_method = 2 🔗
WebP 的預設壓縮方法。影響有損和無損 WebP。較高的值會以壓縮速度為代價產生較小的檔案。解壓縮速度基本上不受壓縮方法的影響。支援的值為 0 到 6。請注意，高於 4 的壓縮方法非常慢並且節省的空間很小。
float rendering/textures/webp_compression/lossless_compression_factor = 25 🔗
無損 WebP 的預設壓縮係數。解壓速度通常不受壓縮係數的影響。支援的值為 0 到 100。
bool rendering/viewport/hdr_2d = false 🔗
If true, enables Viewport.use_hdr_2d on the root Viewport. 2D rendering will use a high dynamic range (HDR) RGBA16 format framebuffer. Additionally, 2D rendering will be performed on linear values and will be converted using the appropriate transfer function immediately before blitting to the screen.
Practically speaking, this means that the end result of the Viewport will not be clamped to the 0-1 range and can be used in 3D rendering without color encoding adjustments. This allows 2D rendering to take advantage of effects requiring high dynamic range (e.g. 2D glow) as well as substantially improves the appearance of effects requiring highly detailed gradients.
Note: This property is only read when the project starts. To toggle HDR 2D at runtime, set Viewport.use_hdr_2d on the root Viewport.
bool rendering/viewport/transparent_background = false 🔗
如果為 true ，則在根視口上啟用 Viewport.transparent_bg 。這樣在同時啟用 display/window/size/transparent 和 display/window/per_pixel_transparency/allowed 之後，逐圖元透明就會生效。
為主視口設定預設的可變速率著色（VRS）模式。請參閱 Viewport.vrs_mode 以在運作時更改該設定，並參閱 VRSMode 以獲取可能的值。
String rendering/vrs/texture = "" 🔗
If rendering/vrs/mode is set to Texture, this is the path to default texture loaded as the VRS image.
The texture must use a lossless compression format so that colors can be matched precisely. The following VRS densities are mapped to various colors, with brighter colors representing a lower level of shading precision:
- 1×1 = rgb(0, 0, 0) - #000000
- 1×2 = rgb(0, 85, 0) - #005500
- 2×1 = rgb(85, 0, 0) - #550000
- 2×2 = rgb(85, 85, 0) - #555500
- 2×4 = rgb(85, 170, 0) - #55aa00
- 4×2 = rgb(170, 85, 0) - #aa5500
- 4×4 = rgb(170, 170, 0) - #aaaa00
- 4×8 = rgb(170, 255, 0) - #aaff00 - Not supported on most hardware
- 8×4 = rgb(255, 170, 0) - #ffaa00 - Not supported on most hardware
- 8×8 = rgb(255, 255, 0) - #ffff00 - Not supported on most hardware
float threading/worker_pool/low_priority_thread_ratio = 0.3 🔗
The ratio of WorkerThreadPool's threads that will be reserved for low-priority tasks. For example, if 10 threads are available and this value is set to 0.3, 3 of the worker threads will be reserved for low-priority tasks. The actual value won't exceed the number of CPU cores minus one, and if possible, at least one worker thread will be dedicated to low-priority tasks.
int threading/worker_pool/max_threads = -1 🔗
Maximum number of threads to be used by WorkerThreadPool. On Web, a value of -1 means 1. On other platforms, it means all logical CPU cores available (see OS.get_processor_count()).
bool xr/openxr/binding_modifiers/analog_threshold = false 🔗
If true, enables the analog threshold binding modifier if supported by the XR runtime.
bool xr/openxr/binding_modifiers/dpad_binding = false 🔗
If true, enables the D-pad binding modifier if supported by the XR runtime.
String xr/openxr/default_action_map = "res://openxr_action_map.tres" 🔗
預設載入的動作對應配置。
bool xr/openxr/enabled = false 🔗
If true, Godot will setup and initialize OpenXR on startup.
int xr/openxr/environment_blend_mode = "0" 🔗
指定 OpenXR 應如何融入環境。這特定於某些 AR 和直通裝置，其中相機影像由 XR 合成器混合。
int xr/openxr/extensions/debug_message_types = "15" 🔗
Specifies the message types for which we request debug messages. Requires xr/openxr/extensions/debug_utils to be set and the extension to be supported by the XR runtime.
int xr/openxr/extensions/debug_utils = "0" 🔗
Enables debug utilities on XR runtimes that supports the debug utils extension. Sets the maximum severity being reported (0 = disabled, 1 = error, 2 = warning, 3 = info, 4 = verbose).
bool xr/openxr/extensions/eye_gaze_interaction = false 🔗
Specify whether to enable eye tracking for this project. Depending on the platform, additional export configuration may be needed.
bool xr/openxr/extensions/frame_synthesis = false 🔗
If true the frame synthesis extension will be activated if supported by the platform.
Note: This feature should not be enabled in conjunction with Application Space Warp, if supported this replaces ASW.
bool xr/openxr/extensions/hand_interaction_profile = false 🔗
If true the hand interaction profile extension will be activated if supported by the platform.
bool xr/openxr/extensions/hand_tracking = false 🔗
If true, the hand tracking extension is enabled if available.
Note: By default hand tracking will only work for data sources chosen by the XR runtime. For SteamVR this is the controller inferred data source, for most other runtimes this is the unobstructed data source. There is no way to query this. If a runtime supports the OpenXR data source extension you can use the xr/openxr/extensions/hand_tracking_controller_data_source and/or xr/openxr/extensions/hand_tracking_unobstructed_data_source to indicate you wish to enable these data sources. If neither is selected the data source extension is not enabled and the XR runtimes default behavior persists.
bool xr/openxr/extensions/hand_tracking_controller_data_source = false 🔗
If true, support for the controller inferred data source is requested. If supported, you will receive hand tracking data even if the user has a controller in hand, with finger positions automatically inferred from controller input and/or sensors.
Note: This requires the OpenXR data source extension and controller inferred handtracking to be supported by the XR runtime. If not supported this setting will be ignored. xr/openxr/extensions/hand_tracking must be enabled for this setting to be used.
bool xr/openxr/extensions/hand_tracking_unobstructed_data_source = false 🔗
If true, support for the unobstructed data source is requested. If supported, you will receive hand tracking data based on the actual finger positions of the user often determined by optical tracking.
Note: This requires the OpenXR data source extension and unobstructed handtracking to be supported by the XR runtime. If not supported this setting will be ignored. xr/openxr/extensions/hand_tracking must be enabled for this setting to be used.
bool xr/openxr/extensions/render_model = false 🔗
If true we enable the render model extension if available.
Note: This relates to the core OpenXR render model extension and has no relation to any vendor render model extensions.
int xr/openxr/extensions/spatial_entity/april_tag_dict = "3" 🔗
The April Tag marker types the built-in marker tracking is set to recognize (if April Tag marker tracking is available and enabled).
int xr/openxr/extensions/spatial_entity/aruco_dict = "15" 🔗
The ArUco marker types the built-in marker tracking is set to recognize (if ArUco marker tracking is available and enabled).
bool xr/openxr/extensions/spatial_entity/enable_builtin_anchor_detection = false 🔗
If true, we enable the built-in logic for handling anchors. Godot will query (persistent) anchors and manage OpenXRAnchorTracker instances for you. If disabled you'll need to create your own spatial and persistence context and perform your own discovery queries.
Note: This functionality requires that spatial anchors are supported and enabled.
bool xr/openxr/extensions/spatial_entity/enable_builtin_marker_tracking = false 🔗
If true, we enable the built-in logic for handling marker tracking. Godot will query markers and manage OpenXRMarkerTracker instances for you. If disabled you'll need to create your own spatial context and perform your own discovery queries.
Note: This functionality requires that marker tracking is supported and enabled.
bool xr/openxr/extensions/spatial_entity/enable_builtin_plane_detection = false 🔗
If true, we enable the built-in logic for handling plane detection. Godot will query detected planes (walls, floors, ceilings, etc.) and manage OpenXRPlaneTracker instances for you. If disabled you'll need to create your own spatial context and perform your own discovery queries.
Note: This functionality requires that plane tracking is supported and enabled.
bool xr/openxr/extensions/spatial_entity/enable_marker_tracking = false 🔗
If true, support for the marker tracking extension is requested. If supported, you will be able to query information about markers detected by the XR runtime, e.g. QR codes, aruca markers and april tags.
Note: This requires that the OpenXR spatial entities and marker tracking extensions are supported by the XR runtime. If not supported this setting will be ignored. xr/openxr/extensions/spatial_entity/enabled must be enabled for this setting to be used.
bool xr/openxr/extensions/spatial_entity/enable_persistent_anchors = false 🔗
If true, support for the persistent anchors extension is requested. If supported, you will be able to store spatial anchors and they will be restored on application startup.
Note: This requires that the OpenXR spatial entities, spatial anchors, and spatial persistence extensions are supported by the XR runtime. If not supported this setting will be ignored. xr/openxr/extensions/spatial_entity/enabled and xr/openxr/extensions/spatial_entity/enable_spatial_anchors must be enabled for this setting to be used.
bool xr/openxr/extensions/spatial_entity/enable_plane_tracking = false 🔗
If true, support for the plane tracking extension is requested. If supported, you will be able to query information about planes detected by the XR runtime, e.g. walls, floors, etc.
Note: This requires that the OpenXR spatial entities and plane tracking extensions are supported by the XR runtime. If not supported this setting will be ignored. xr/openxr/extensions/spatial_entity/enabled must be enabled for this setting to be used.
bool xr/openxr/extensions/spatial_entity/enable_spatial_anchors = false 🔗
If true, support for the spatial anchors extension is requested. If supported, you will be able to register anchor locations in the real world that the XR runtime will adjust as needed and/or potentially share with other headsets.
Note: This requires that the OpenXR spatial entities and spatial anchors extensions are supported by the XR runtime. If not supported this setting will be ignored. xr/openxr/extensions/spatial_entity/enabled must be enabled for this setting to be used.
bool xr/openxr/extensions/spatial_entity/enabled = false 🔗
If true, support for the spatial entity extension is requested. If supported, you will be able to access spatial information about the real environment around you. What information is available is dependent on additional extensions.
Note: This requires that the OpenXR spatial entities extension is supported by the XR runtime. If not supported this setting will be ignored.
bool xr/openxr/extensions/user_presence = false 🔗
If true, the user presence extension is enabled if available.
int xr/openxr/form_factor = "0" 🔗
指定是否應為 HMD 或手持裝置配置 OpenXR。
bool xr/openxr/foveation_dynamic = false 🔗
If true and foveation is supported, will automatically adjust foveation level based on framerate up to the level set on xr/openxr/foveation_level.
bool xr/openxr/foveation_eye_tracked = true 🔗
If true and foveation level is set to anything other than "Disabled", eye-tracked foveation will be used, so long as it's supported by the headset.
int xr/openxr/foveation_level = "0" 🔗
Applied foveation level if supported.
Note: On platforms other than Android, if rendering/anti_aliasing/quality/msaa_3d is enabled, this feature will be disabled.
bool xr/openxr/foveation_with_subsampled_images = true 🔗
If true and foveation is also enabled, subsampled images will be used on Vulkan. This can improve the performance gain from foveated rendering, especially when using high foveation levels.
Note:: Using subsampled images is incompatible with many screen-space rendering features or post-processing effects like FXAA or glow. If any such effects are enabled, subsampled images will automatically be disabled and a warning shown in the log.
int xr/openxr/reference_space = "1" 🔗
指定預設參照空間。
bool xr/openxr/startup_alert = true 🔗
如果為 true，則啟動時如果 OpenXR 初始化失敗，Godot 就會顯示警告彈框。
bool xr/openxr/submit_depth_buffer = false 🔗
如果為 true，則 OpenXR 會管理深度緩衝區，使用深度緩衝區進行高級再投影，前提是 XR 運作時支援。請注意，Godot 中的部分算繪功能無法與該功能一同使用。
String xr/openxr/target_api_version = "" 🔗
Optionally sets a specific API version of OpenXR to initialize in major.minor.patch notation. Some XR runtimes gate old behavior behind version checks. This is non-standard OpenXR behavior.
int xr/openxr/view_configuration = "1" 🔗
指定視圖配置，用於配置 OpenXR 設定單視場或立體算繪。
bool xr/shaders/enabled = false 🔗
如果為 true，Godot 將編譯 XR 所需的著色器。
方法說明
void add_property_info(hint: Dictionary) 🔗
Adds a custom property info to a property. The dictionary must contain:
"name": String (the property's name)"type": int (see Variant.Type)optionally
"hint": int (see PropertyHint) and"hint_string": String
ProjectSettings.set("category/property_name", 0)
var property_info = {
"name": "category/property_name",
"type": TYPE_INT,
"hint": PROPERTY_HINT_ENUM,
"hint_string": "one,two,three"
}
ProjectSettings.add_property_info(property_info)
ProjectSettings.Singleton.Set("category/property_name", 0);
var propertyInfo = new Godot.Collections.Dictionary
{
{ "name", "category/propertyName" },
{ "type", (int)Variant.Type.Int },
{ "hint", (int)PropertyHint.Enum },
{ "hint_string", "one,two,three" },
};
ProjectSettings.AddPropertyInfo(propertyInfo);
Note: Setting "usage" for the property is not supported. Use set_as_basic(), set_restart_if_changed(), and set_as_internal() to modify usage flags.
bool check_changed_settings_in_group(setting_prefix: String) const 🔗
檢查已改變的設定中是否存在前綴為 setting_prefix 的設定項。另見 get_changed_settings()。
清除整個配置（不推薦，可能會弄壞東西）。
PackedStringArray get_changed_settings() const 🔗
Gets an array of the settings which have been changed since the last save. Note that internally changed_settings is cleared after a successful save, so generally the most appropriate place to use this method is when processing settings_changed.
Array[Dictionary] get_global_class_list() 🔗
返回已註冊的全域類的 Array。每個全域類都被表示為包含以下條目的 Dictionary：
base是基底類別的名稱；class是被註冊的全域類的名稱；icon是全域類自訂圖示的路徑，如果有的話；language是編寫全域類的程式設計語言的名稱；path是包含全域類的檔的路徑。
注意：腳本和圖示路徑都是專案檔案系統的本地路徑，即它們以 res:// 開頭。
int get_order(name: String) const 🔗
返回配置值的順序（保存到設定檔時會產生影響）。
Variant get_setting(name: String, default_value: Variant = null) const 🔗
Returns the value of the setting identified by name. If the setting doesn't exist and default_value is specified, the value of default_value is returned. Otherwise, null is returned.
print(ProjectSettings.get_setting("application/config/name"))
print(ProjectSettings.get_setting("application/config/custom_description", "No description specified."))
GD.Print(ProjectSettings.GetSetting("application/config/name"));
GD.Print(ProjectSettings.GetSetting("application/config/custom_description", "No description specified."));
Note: This method doesn't take potential feature overrides into account automatically. Use get_setting_with_override() to handle seamlessly.
See also has_setting() to check whether a setting exists.
Variant get_setting_with_override(name: StringName) const 🔗
Similar to get_setting(), but applies feature tag overrides if any exists and is valid.
Example: If the setting override "application/config/name.windows" exists, and the following code is executed on a Windows operating system, the overridden setting is printed instead:
print(ProjectSettings.get_setting_with_override("application/config/name"))
GD.Print(ProjectSettings.GetSettingWithOverride("application/config/name"));
Variant get_setting_with_override_and_custom_features(name: StringName, features: PackedStringArray) const 🔗
Similar to get_setting_with_override(), but applies feature tag overrides instead of current OS features.
String globalize_path(path: String) const 🔗
返回與當地語系化 path（以 res:// 或 user:// 開頭）相對應的絕對原生 OS 路徑。返回的路徑將因作業系統和使用者首選項而異。 請參閱《Godot 專案中的檔路徑》以查看這些路徑轉換成的內容。另請參閱 localize_path()。
注意：對 res:// 呼叫 globalize_path() 在匯出的專案中不會起作用。而是，當從匯出的專案運作時，將可執行檔的基目錄新增到路徑中：
var path = ""
if OS.has_feature("editor"):
# 從編輯器二進位檔案運作。
# `path` 將包含位於專案根目錄中的 `hello.txt` 的絕對路徑。
path = ProjectSettings.globalize_path("res://hello.txt")
else:
# 從匯出的專案運作。
# `path` 將包含可執行檔旁邊的 `hello.txt` 的絕對路徑。
# 這與使用 `ProjectSettings.globalize_path()` 和 `res://` 路徑*不*相同，
# 但在目的上足夠接近。
path = OS.get_executable_path().get_base_dir().path_join("hello.txt")
bool has_setting(name: String) const 🔗
Returns true if a configuration value is present.
Note: In order to be be detected, custom settings have to be either defined with set_setting(), or exist in the project.godot file. This is especially relevant when using set_initial_value().
bool load_resource_pack(pack: String, replace_files: bool = true, offset: int = 0) 🔗
Loads the contents of the .pck or .zip file specified by pack into the resource filesystem (res://). Returns true on success.
Note: If a file from pack shares the same path as a file already in the resource filesystem, any attempts to load that file will use the file from pack unless replace_files is set to false.
Note: The optional offset parameter can be used to specify the offset in bytes to the start of the resource pack. This is only supported for .pck files.
Note: DirAccess will not show changes made to the contents of res:// after calling this function.
String localize_path(path: String) const 🔗
返回對應於絕對本地作業系統 path 的當地語系化路徑（以 res:// 開頭）。另見 globalize_path()。
將配置保存到 project.godot 檔案中。
注意：此方法是為編輯器外掛程式使用的，因為修改後的 ProjectSettings 無法在運作的套用程式中載入回來。如果要更改匯出專案中的專案設定，請使用 save_custom() 保存 override.cfg 檔案。
Error save_custom(file: String) 🔗
將配置保存到自訂檔。檔副檔名必須是 .godot（以基於文字的 ConfigFile 格式保存）或 .binary（以二進位格式保存）。你也可以保存為 override.cfg 檔，它也是文字，但與其他格式不同，可以在導出的專案中使用。
void set_as_basic(name: String, basic: bool) 🔗
決定指定的設定是基本設定還是高級設定。專案設定中始終顯示基本設定。高級設定僅在使用者啟用“高級設定”選項時顯示。
void set_as_internal(name: String, internal: bool) 🔗
決定指定的設定是否為內部設定。內部設定不會在“專案設定”對話方塊中出現。常用於插件，用來儲存內部設定，避免直接暴露給使用者。
void set_initial_value(name: String, value: Variant) 🔗
Sets the specified setting's initial value. This is the value the setting reverts to. The setting should already exist before calling this method. Note that project settings equal to their default value are not saved, so your code needs to account for that.
extends EditorPlugin
const SETTING_NAME = "addons/my_setting"
const SETTING_DEFAULT = 10.0
func _enter_tree():
if not ProjectSettings.has_setting(SETTING_NAME):
ProjectSettings.set_setting(SETTING_NAME, SETTING_DEFAULT)
ProjectSettings.set_initial_value(SETTING_NAME, SETTING_DEFAULT)
If you have a project setting defined by an EditorPlugin, but want to use it in a running project, you will need a similar code at runtime.
void set_order(name: String, position: int) 🔗
設定配置值的順序（保存到設定檔時會產生影響）。
void set_restart_if_changed(name: String, restart: bool) 🔗
設定一個設定是否需要重新開機編輯器才能正確生效。
注意：這只是向使用者顯示的提示，提示必須重新啟動編輯器才能使更改生效。啟用 set_restart_if_changed() 不會延遲在更改時設定的設定。
void set_setting(name: String, value: Variant) 🔗
Sets the value of a setting.
ProjectSettings.set_setting("application/config/name", "Example")
ProjectSettings.SetSetting("application/config/name", "Example");
This can also be used to erase custom project settings. To do this change the setting value to null.