Class TCastleViewport

Warning: this symbol is deprecated: this default is not used; ProjectionFar in orthographic projection is now automatically adjusted to what you display

referencing deprecated in deprecated

Warning: this symbol is deprecated: this default is not used; ProjectionNear in orthographic projection is now automatically adjusted to what you display

Rendering pass, for user purposes. Useful to keep shaders cached when you render the same scene multiple times in the same frame (under different lighting conditions or other things that change shaders). By default this is always 0, the engine doesn't modify this. You can set this field manually.

At design-time (in CGE editor), this is the camera used by the editor. Usually use InternalCamera, to automatically pick either design-time camera or runtime Camera.

Calculate projection parameters. Determines if the view is perspective or orthogonal and exact field of view parameters. Called each time at the beginning of rendering.

The default implementation of this method in TCastleViewport calculates projection based on the Camera parameters.

In turn, the Camera parameters may be automatically calculated (if AutoCamera) based on the nodes in the TCastleRootTransform.MainScene. Nodes like TViewpointNode or TOrthoViewpointNode or TNavigationInfoNode determine the default camera and projection details.

You can override this method, or assign the OnProjection event to adjust the projection settings. But please note: instead of overriding this method, it's usually easier (and more advised) to simply change the Camera properties, like Camera.ProjectionType or Camera.Orthographic.Width or Camera.Perspective.FieldOfView.

Prepare lights shining on everything. GlobalLights contents should be initialized here. The implementation in this class adds headlight determined by the Items.UseHeadlight value.

Warning: this symbol is deprecated: use InitializeGlobalLights

Render everything from given camera view (as TRenderingCamera). Given RenderingCamera.Target says to where we generate the image. This method must take care of making many rendering passes for shadow volumes, but doesn't take care of updating generated textures.

Render the scene, assuming that buffers were already cleared and background was rendered. Called by RenderFromViewEverything at the end. Lights are calculated in Params at this point.

This will change Params.Transparent, Params.InShadow and Params.ShadowVolumesReceivers as needed. Their previous values do not matter.

Render one pass, with current camera and parameters (e.g. only transparent or only opaque shapes). All current camera settings are saved in RenderParams.RenderingCamera.

Parameters

Params

Rendering parameters, see TRenderParams.

Warning: this symbol is deprecated: use RenderOnePass

Called when PointingDevicePress was not handled by any TCastleTransform object. You can override this to make a message / sound signal to notify user that his Input_Interact click was not successful.

Warning: this symbol is deprecated: use Items.MainScene

Assign current camera vectors and projection. This only fills existing Camera with contents, it doesn't change the Camera to new component.

This is automatically used at first rendering if AutoCamera. You can also use it explicitly.

Does the graphic card support our ScreenSpaceAmbientOcclusion shader. This does not depend on the current state of ScreenSpaceAmbientOcclusion property. You can use it e.g. to disable the menu item to switch SSAO in 3D viewer.

Does the graphic card support our ScreenSpaceReflections shader. This does not depend on the current state of ScreenSpaceReflections property. You can use it e.g. to disable the menu item to switch SSR in 3D viewer.

Parameters to prepare items that are to be rendered within this world. This should be passed to TCastleTransform.PrepareResources.

Note: Instead of using TCastleTransform.PrepareResources, and this method, it's usually easier to call TCastleViewport.PrepareResources. Then the appropriate TPrepareParams will be passed automatically.

Warning: this symbol is deprecated: this is internal info, you should not need this; use PrepareParams to get opaque information to pass to TCastleTransform.PrepareResources

Statistics about last rendering frame. See TRenderStatistics docs.

Current 3D triangle under the mouse cursor. Updated in every mouse move. May be Nil.

Utility method to set camera to a suitable state for 2D games.

• Sets both initial and current camera vectors like this:

  • Position camera at (0, 0, 500),

  • Looks along the -Z direction,

  • "Up" vector is in +Y.

  This way the 2D world spans horizontally in X and vertically in Y. The Z (depth) can be used to put things in front/behind each other.

  Since this initialized the camera sensibly, we also set AutoCamera to false.

• Sets orthographic projection for the camera (TCastleCamera.ProjectionType set to ptOrthographic).

  By default our visible X range is [0..viewport width in pixels], visible Y range is [0..viewport height in pixels]. Use the properties of Camera.Orthographic to control the projection. For example set Camera.Orthographic.Width and/or Camera.Orthographic.Height to define visible projection size (horizontal or vertical) explicitly, regardless of the viewport size.

  Setting Camera.Orthographic.Origin is also often useful, e.g. set it to (0.5,0.5) to make the things positioned at (0,0) in the world visible at the middle of the viewport.

  By default our visible Z range is [-1500, 500], because this sets ProjectionNear to -1000, ProjectionFar to 1000, and camera default depth (Camera.Position.Z) is 500. This was chosen to be comfortable for all cases – you can keep camera Z unchanged and comfortably position things around [-500, 500], or set camera Z to zero and then comfortably position things around [-1000, 1000].

Convert 2D position on the viewport into 3D "world coordinates", by colliding camera ray with a plane parallel to the viewport at given Depth. "World coordinates" are coordinates space seen by TCastleTransform / TCastleScene inside viewport Items. Use Depth > 0 for positions in front of the camera.

This is similar to "Unproject" GLU routine. It allows to map points from the 2D viewport back to the 3D space inside viewport.

The interpretation of Position depends on ContainerCoordinates:

• When ContainerCoordinates = True, then Position is relative to the whole container (like TCastleWindow or TCastleControl).

  And it is expressed in real device coordinates, just like TInputPressRelease.Position when mouse is being clicked, or like TInputMotion.Position when mouse is moved.

• When ContainerCoordinates = False, then Position is relative to this UI control.

  And it is expressed in coordinates after UI scaling. IOW, if the size of this control is Width = 100, then Position.X between 0 and 100 reflects the visible range of this control.

Returns true and sets 3D PlanePosition if such intersection is found. Returns false if it's not possible to determine such point (which should not be possible, unless camera field of view is larger than 180 degrees).

Convert 2D position on the viewport into 3D ray.

The interpretation of Position depends on ContainerCoordinates, and is similar to e.g. TCastleTiledMapControl.PositionToTile:

• When ContainerCoordinates = True, then Position is relative to the whole container (like TCastleWindow or TCastleControl).

  And it is expressed in real device coordinates, just like TInputPressRelease.Position when mouse is being clicked, or like TInputMotion.Position when mouse is moved.

• When ContainerCoordinates = False, then Position is relative to this UI control.

  And it is expressed in coordinates after UI scaling. IOW, if the size of this control is Width = 100, then Position.X between 0 and 100 reflects the visible range of this control.

Convert 2D position on the viewport into 3D "world coordinates", by colliding camera ray with a plane at constant Z. "World coordinates" are coordinates space seen by TCastleTransform / TCastleScene inside viewport Items.

This is a more general version of PositionTo2DWorld, that works with any projection (perspective or orthographic). This is often useful if your game is "close to 2D", which means that you use 3D (and maybe even perspective camera), but most of the game world is placed around some plane with constant Z.

The interpretation of Position depends on ContainerCoordinates, and is similar to e.g. TCastleTiledMapControl.PositionToTile:

• When ContainerCoordinates = True, then Position is relative to the whole container (like TCastleWindow or TCastleControl).

  And it is expressed in real device coordinates, just like TInputPressRelease.Position when mouse is being clicked, or like TInputMotion.Position when mouse is moved.

• When ContainerCoordinates = False, then Position is relative to this UI control.

  And it is expressed in coordinates after UI scaling. IOW, if the size of this control is Width = 100, then Position.X between 0 and 100 reflects the visible range of this control.

This intersects the ray cast by Camera with a plane at Z = PlaneZ.

Returns true and sets 3D PlanePosition (the Z component of this vector must always be equal to PlaneZ) if such intersection is found. Returns false if it's not possible to determine such point (when the camera looks in the other direction).

Convert 2D position into "world coordinates", which is the coordinate space seen by TCastleTransform / TCastleScene inside viewport Items, assuming that we use orthographic projection in XY axes.

The interpretation of Position depends on ContainerCoordinates, and is similar to e.g. TCastleTiledMapControl.PositionToTile:

• When ContainerCoordinates = True, then Position is relative to the whole container (like TCastleWindow or TCastleControl).

  And it is expressed in real device coordinates, just like TInputPressRelease.Position when mouse is being clicked, or like TInputMotion.Position when mouse is moved.

• When ContainerCoordinates = False, then Position is relative to this UI control.

  And it is expressed in coordinates after UI scaling. IOW, if the size of this control is Width = 100, then Position.X between 0 and 100 reflects the visible range of this control.

This assumes that camera "up vector" is +Y, and it is looking along the negative Z axis. It also assumes orthographic projection (Camera.ProjectionType equal ptOrthographic). These are default camera direction, up and projection types set by Setup2D.

Invert PositionTo2DWorld, converting the "world coordinates" (the coordinate space seen by TCastleTransform / TCastleScene inside viewport Items) into final screen position.

Just like PositionTo2DWorld, this assumes that camera "up vector" is +Y, and it is looking along the negative Z axis. It also assumes orthographic projection (Camera.ProjectionType equal ptOrthographic). These are default camera direction, up and projection types set by Setup2D.

Project 3D coordinates (the coordinate space seen by TCastleTransform / TCastleScene inside viewport Items) into 2D position on the viewport UI.

The resulting position is relative to this viewport control and it is expressed in coordinates after UI scaling. IOW, if the size of this control is EffectiveWidth = 100, then Position.X between 0 and 100 reflects the visible range of this control. This is like using ContainerCoordinates = False with PositionToRay.

Prepare resources, to make various methods (like Render) execute fast. Call it only when rendering context is initialized (ApplicationProperties.IsGLContextOpen). If DisplayProgressTitle <> '', we will display progress bar during loading.

Current object (TCastleTransform instance) under the mouse cursor. Updated in every mouse move. May be Nil.

Create new camera and make it used by this viewport. This creates new unnamed TCastleCamera, owned by this viewport, assigns it to Camera and adds it to Items.

Call this only when Camera is Nil.

This is done automatically when creating TCastleViewport (not in design-mode, not at deserialization) using the standard TCastleViewport.Create constructor. So you likely don't need to call this in typical applications.

Note: Even when used from CGE editor (CastleDesignMode), this creates non-design camera. This way viewports created with InternalCreateNonDesign have also non-design cameras. This is important, so that e.g. camera has no bounding box, and thus AssignDefaultCamera in sprite sheet editor doesn't put camera further and further away.

Warning: this symbol is deprecated: in most cases, you can instead read Camera parameters, like Camera.Orthographic.EffectiveWidth, Camera.Orthographic.EffectiveHeight

Current projection parameters, calculated by last CalculateProjection call, adjusted by OnProjection. This is read only. To change the projection parameters, override CalculateProjection or handle event OnProjection.

Screen effects are shaders that post-process the rendered screen. If any screen effects are active, we will automatically render screen to a temporary texture, processing it with each shader.

By default, screen effects come from MainScene.ScreenEffects, so the effects may be defined by VRML/X3D author using ScreenEffect nodes (see docs: [https://castle-engine.io/x3d_extensions_screen_effects.php]). Descendants may override GetScreenEffects, ScreenEffectsCount, and ScreenEffectsNeedDepth to add screen effects by code. Each viewport may have it's own, different screen effects.

Background color. Displayed only if Background is Nil and MainScene doesn't define any background either.

Ignored if Transparent.

Dark gray (DefaultBackgroundColor) by default.

Enable built-in SSAO screen effect in the world.

Enable built-in SSR screen effect in the world.

Adjust SSR default surface glossiness.

Called on any camera change.

Current object (TCastleTransform hierarchy) under the mouse cursor. Updated in every mouse move. May be Nil.

The returned list (if not Nil) contains TCastleTransform instances that collided with the ray (from the deepest instance in the Items tree to the root), along with some additional information. See TRayCollision for details.

Do not collide with this object when moving by Navigation. It makes sense to put here player avatar (in 3rd person view) or player collision volume (in 1st person view) to allow player to move, not colliding with its own body.

Warning: this symbol is deprecated: use Items.Paused

See TCastleAbstractRootTransform.Paused.

Warning: this symbol is deprecated: assign Items from one TCastleViewport to another to view the same world from multiple viewports

Warning: this symbol is deprecated: no need to set this, instead add TCastleNavigation like "MyViewport.InsertBack(MyNavigation)"

Navigation method is an optional component that handles the user input to control the camera.

This is a deprecated property. There's no point in assigning current navigation like this. Instead add the TCastleNavigation instance as a child of TCastleViewport and it will affect the viewport automatically. This property is now only a shortcut to

- get the current TCastleNavigation child (if multiple present, it is undefined which is returned)

- make the given set TCastleNavigation as the only child.

You can assign here an instance of TCastleNavigation, like TCastleWalkNavigation or TCastleExamineNavigation. Or you can leave it as Nil.

Transformations and scenes visible in this viewport. You should add here your TCastleTransform and TCastleScene instances.

It is by default created (not Nil), but you can also assign here your own TCastleRootTransform instance. You can also copy a TCastleRootTransform from one TCastleViewport to another, that is multiple TCastleViewport can refer to the same TCastleRootTransform instance.

Note that assigning here Nil is allowed, but under the hood it just creates an empty TCastleRootTransform instance. So this property is never Nil when reading. But try to not depend on it for the future, at some point this property may be allowed to be Nil, for consistency.

Camera determines the viewer position and orientation.

You should create TCastleCamera, assign to this property, and add it somewhere to Items.

Note that camera can be placed anywhere within Items, not necessary as direct child of Items. For example it can be a children of some deeper TCastleTransform, this way you can attach camera e.g. to some bone or to a moving object.

Generally, TCastleCamera is a regular component that can be added and removed freely from Items, renamed, freed etc. And setting this property is an independent action that doesn't add/remove any camera from Items. This property can also be set to Nil.

For convenience, there are however some moments when we set up this property and add camera to Items automatically:

1. When you create TCastleViewport from code (not by deserializing some xxx.castle-user-interface file, not in design-mode in editor) then we automatically add new (unnamed) camera and set Viewport.Camera and add it to Viewport.Items.

2. Only at design-time (in editor): creating new viewport by editor menu also adds a default camera to it. In this case camera component is named and owned by the same owner as viewport. This is done right after TCastleViewport creation and only at design-time.

3. For backward compatibility, after deserializing file from previous engine versions, we add the camera from it to Viewport.Items. This is made strictly for backward compatibility, only when deserializing, and it shouldn't occur when reading new design files.

It's important to note that when deserializing camera from new engine versions, there's no "magic". We just expect that Viewport.Camera and Viewport.Items are good. Trying to do something automatic/smart in this case turned out to be quite troublesome and breaking more than helping.

Should we render with shadow volumes. You can change this at any time, to switch rendering shadows on/off.

This works only if OpenGL context actually can render shadow volumes, checked by GLFeatures.ShadowVolumesPossible, which means that you have to initialize OpenGL context with stencil buffer.

The shadow volumes algorithm is used only if shadow caster is 2-manifold, that is has a correctly closed volume. Also you need a light source marked as the main shadow volumes light (shadowVolumes = shadowVolumesMain = TRUE). See [https://castle-engine.io/x3d_extensions.php#section_ext_shadows] for details.

Actually draw the shadow volumes to the color buffer, for debugging. If shadows are rendered (see GLFeatures.ShadowVolumesPossible and ShadowVolumes), you can use this to actually see shadow volumes, for debug / demo purposes. Shadow volumes will be rendered on top of the scene, as yellow blended polygons.

Background (like a skybox) to display behind the Items. Displayed only when not Transparent. See https://castle-engine.io/background

Fog to use to display viewport contents in Items. See https://castle-engine.io/fog

If True then the background (from Background or MainScene) will be rendered wireframe, over the solid background filled with BackgroundColor. Useful for debugging when you want to see how your background geometry looks like.

Ignored if Transparent.

If True then the viewport will not draw a background and the UI underneath will be visible. The UI underneath is visible at pixels which are not covered by items on Items, and where items Items have partially-transparent materials.

If True, the Background, BackgroundColor, BackgroundWireframe, and any background defined inside MainScene will be ignored.

At the beginning of rendering, viewport by default clears the depth buffer. This makes every viewport draw everything on top of the previous 2D and 3D stuff (including on top of previous viewport), like a layer.

You can disable this, which allows to combine together the 3D objects rendered by various viewports (and by custom OpenGL rendering), such that the 3D positions determime what overlaps what. This only makes sense if you have a number of TCastleViewport instances, that share the same size and position on the screen, same projection and the same camera.

It's your responsibility in such case to clear the depth buffer. E.g. place one viewport in the back that has ClearDepth = True. Or place a TCastleUserInterface descendant in the back, that calls TRenderContext.Clear RenderContext.Clear in overridden TCastleUserInterface.Render.

Note: to disable clearning the color buffer, set Transparent to False.

Note: if you use shadow volumes, we will still clear the stencil buffer at the beginning of rendering.

Warning: this symbol is deprecated: if you need to tweak this, then do not use MainScene; use regular TCastleScene and set CastGlobalLights as needed

Let lights in MainScene shine on every other TCastleScene, not only MainScene. This is an easy way to lit your whole world with lights defined inside MainScene file. Be sure to set X3D lights global=TRUE.

Warning: this symbol is deprecated: configure fog by assigning to TCastleViewport.Fog component; leave deprecated TCastleViewport.MainScene nil

Let the fog defined in MainScene affect all objects, not only MainScene. This is consistent with UseGlobalLights, that allows lights from MainScene to shine on all objects.

Help user to activate pointing device sensors and pick items. Every time you press Input_Interact (by default just left mouse button), we look if current mouse/touch position hits an object (TCastleTransform) that actually does something on activation. The object may do various stuff inside TCastleTransform.PointingDevicePress, generally this causes various picking/interaction with the object (like pulling a level, opening a door), possibly dragging, possibly with the help of VRML/X3D pointing device and drag sensors.

When this is True, we try harder to hit some 3D object that handles PointingDevicePress. If there's nothing interesting under mouse/touch, we will retry a couple of other positions around the current mouse/touch.

This should be usually used when you use TCastleMouseLookNavigation.MouseLook, or other navigation when mouse cursor is hidden. It allows user to only approximately look at interesting item and hit interaction button or key. Otherwise, activating a small object is difficult, as you don't see the cursor.

Warning: this symbol is deprecated: adjust projection by changing Camera.ProjectionType and other projection parameters inside Camera

Adjust the projection parameters. This event is called before every render. See the CalculateProjection for a description how to default projection parameters are calculated.

Enable to drag a parent control, for example to drag a TCastleScrollView that contains this TCastleViewport, even when the scene inside contains clickable elements (using TouchSensor node).

To do this, you need to turn on TCastleScrollView.EnableDragging, and set EnableParentDragging=True here. In effect, viewport will cancel the click operation once you start dragging, which allows the parent to handle all the motion events for dragging.

Warning: this symbol is deprecated: it is simpler to set camera at design-time explicitly, or use CameraViewpointForWholeScene to auto-adjust camera; if you want to animate the camera, attach TCastleCamera to a bone transformation exposed by Scene.ExposeTransforms

Assign initial camera properties (initial position, direction, up, TCastleCamera.ProjectionNear) by looking at the initial world (Items) when rendering the first frame.

The AssignDefaultCamera is automatically called only if this property is True.

Also, only if this property is True, we synchronize camera when X3D Viewpoint node changes, or a new X3D Viewpoint node is bound.

By default it is False, which means you control Camera properties on your own.

Called when bound Viewpoint node changes. Called exactly when TCastleSceneCore.ViewpointStack.OnBoundChanged is called.

Called when bound NavigationInfo changes (to a different node, or just a field changes).

Protect from falling down because of gravity when position is outside of world bounding box. This is a nice thing for general model viewers (like view3dscene), to prevent from accidentally falling down when using "Walk" mode.

This is only used by navigations performing gravity internally, that is right now: TCastleWalkNavigation (when TCastleWalkNavigation.Gravity = True).

BackgroundColor that can be visually edited in Castle Game Engine Editor, Lazarus and Delphi. Normal user code does not need to deal with this, instead read or write BackgroundColor directly.

See also

BackgroundColor

Background color.