SoSeparator Class Reference
[Groups]
Group node that saves and restores traversal state. More...
#include <Inventor/nodes/SoSeparator.h>
Classes | |
| struct | MTstruct |
Public Types | |
| enum | CacheEnabled { OFF = SO_RENDERCACHING_OFF, ON = SO_RENDERCACHING_ON, AUTO = SO_RENDERCACHING_AUTO } |
| enum | FastEditing { DISABLE = 0x01, KEEP_ZBUFFER = 0x02, CLEAR_ZBUFFER = 0x03 } |
Public Member Functions | |
| virtual SoType | getTypeId () const |
| SoSeparator () | |
| SoSeparator (int nChildren) | |
Static Public Member Functions | |
| static SoType | getClassTypeId () |
Public Attributes | |
| SoSFEnum | renderCaching |
| SoSFEnum | boundingBoxCaching |
| SoSFEnum | renderCulling |
| SoSFEnum | pickCulling |
| SoSFEnum | directVizCaching |
| SoSFEnum | fastEditing |
| SoSFInt32 | renderUnitId |
Deprecated | |
|
| |
| static SoDEPRECATED void | setNumRenderCaches (int howMany) |
| static SoDEPRECATED int | getNumRenderCaches () |
Detailed Description
Group node that saves and restores traversal state.
This group node performs a push (save) of the traversal state before traversing its children and a pop (restore) after traversing them. This isolates the separator's children from the rest of the scene graph. A separator can include lights, cameras, coordinates, normals, bindings, and all other properties. Separators are relatively inexpensive (the push/pop operation is highly optimized), so they can be used freely within scenes.
Caching:
The SoSeparator node provides caching of state during rendering and during bounding box computation. This feature can be explicitly enabled or disabled by setting the renderCaching and boundingBoxCaching fields. By default, these fields are set to AUTO, which means that Open Inventor decides whether to build a cache based on internal heuristics.
Note that during the caching process, insufficient memory errors can happen, due to hardware limitations (ex: GPU memory). In this case, to prevent bad rendering or a crash, Open Inventor will disable caching for the current Separator by setting its renderCaching field to OFF. In this case, we recommend to split the Separator's scene graph into several sub-scene graphs each with a new Separator as the head node.
Note that setting the renderCaching policy to OFF will prevent this specific Separator from building a cache but it does not prevent Separator nodes higher in the scene graph from building a cache that includes this Separator. In other words, during traversal it will not invalidate a currently open cache created by a Separator above this one. To invalidate any open cache and prevent this Separator from being cached, create an SoCallback node and use it to call the SoCacheElement's invalidate() method. This may reduce rendering performance.
The directVizCaching field is only used when DirectViz rendering is enabled. It controls the way DirectViz internal objects are built. When an SoSeparator node is traversed with directVizCaching set to ON, DirectViz will build one internal object for all the shapes below, and the directVizCaching field is ignored for all of the SoSeparator nodes below. Default value is OFF.
Culling:
Separators can also perform culling during rendering and picking. Culling skips over traversal of the separator's children if they are not going to be rendered or picked, based on the comparison of the separator's bounding box with the current view volume. Culling is controlled by the renderCulling and pickCulling fields. These are also set to AUTO by default; however, render culling can be expensive (and can interfere with render caching), so the AUTO heuristics leave it disabled unless specified otherwise. Note that if renderCulling is ON, the SoSeparator will prohibit auto-caching, so that no SoSeparator (or other render caching node) above this node in the scene graph will automatically create a render cache.
Fast edit:
The fast editing feature allows you to modify parts of a scene without redrawing the entire scene. For example, you could use it to interactively move a small object in a large scene that takes a long time to redraw. Fast editing must be enabled using SoGLRenderAction::setFastEditSavePolicy.
The fastEditing field enables fast editing for the sub-graph under the SoSeparator. Possible values are DISABLE, KEEP_ZBUFFER, and CLEAR_ZBUFFER. Using KEEP_ZBUFFER means that the fast edit geometry is depth buffered against the other objects in the scene, and using CLEAR_ZBUFFER means that the fast edit geometry will be drawn on top of the other objects of the scene. If several SoSeparators have the CLEAR_ZBUFFER flag set, they are drawn in the order in which they appear in the scene. The last separator in the scene will be topmost on the screen.
Fast edit performance:
- This feature is most efficient when the parts of a scene graph to be fast edited contain a relatively small number of triangles.
- The topmost SoSeparator of your scene graph should not enable the fast editing feature. Otherwise, performance will be very poor.
- Currently, the entire viewport is saved and restored during fast editing. Therefore, fast editing performance is linked to the size of the viewport: the larger the viewport, the longer it will take to save and restore it.
Fast edit limitations:
- Hardware: Fast editing performance depends on your graphics hardware and driver. To use the fast editing feature, the ARB_bufferRegion OpenGL extension is required (refer to SoBufferRegion class description).
- Clip planes: When a fast edit object is moved outside the clip planes limit, fast editing is temporarily disabled and the entire scene is redrawn.
- All fast editing sub-graphs are rendered even if only one fast editing sub-graph has changed.
- Transparency: If the scene contains partially transparent objects, there is a possibility of visual artifacts regardless of what transparency algorithm is used. The fast edit shapes are necessarily rendered on top of the rest of the scene, which may not be the order required for correct transparency.
- Overlay planes: When using overlay planes, fast edit can cause visual artifacts.
In this case, we recommend setting the OIV_BUFFER_REGION_ENABLE environment variable to 0. The rendering performance of fast edit will decrease slightly, but the resulting rendering will be correct. - Stereo rendering: As fast edit is not compatible with stereo rendering, it is automatically disabled when switching to stereo rendering.
FILE FORMAT/DEFAULT
- Separator {
| renderCaching | AUTO |
| boundingBoxCaching | AUTO |
| directVizCaching | OFF |
| renderCulling | AUTO |
| pickCulling | AUTO |
| fastEditing | DISABLE |
| renderUnitID | -1 |
ACTION BEHAVIOR
- SoGLRenderAction, SoCallbackAction, SoGetBoundingBoxAction, SoGetMatrixAction, SoHandleEventAction, SoRayPickAction, SoSearchAction
Saves the current traversal state, traverses all children, and restores the previous traversal state. Sets: SoCacheElement
SEE ALSO
SoSelection, SoTransformSeparator, SoBufferRegion, SoDirectVizManager.
- See related examples:
Member Enumeration Documentation
Fast editing policy enumeration values.
- Enumerator:
DISABLE This SoSeparator node doesn't have Fast Edit behavior.
KEEP_ZBUFFER This SoSeparator node has a Fast Edit behavior and the children nodes are drawn considering the depth Buffer.
CLEAR_ZBUFFER This SoSeparator node has a Fast Edit behavior and the children nodes are drawn after the depth Buffer has been cleared.
Constructor & Destructor Documentation
| SoSeparator::SoSeparator | ( | ) |
Creates a separator node with default settings.
| SoSeparator::SoSeparator | ( | int | nChildren | ) |
Constructor that takes approximate number of children.
Member Function Documentation
| static SoType SoSeparator::getClassTypeId | ( | ) | [static] |
Returns the type identifier for this class.
Reimplemented from SoGroup.
Reimplemented in SoAnnotation, SoBillboard, SoExtSelection, SoFrontBufferGroup, SoLocateHighlight, SoOcclusionCulling, SoOctreeOrdering, SoRenderList, SoRenderToTarget, SoSelection, SoValueOrdering, SoVertexAttribFeedback, SoWWWAnchor, SoMultiDataSeparator, DicomInfo, Gnomon, PlaneGeometryIntersection, Ruler, SceneView, SliceOrientationMarkers, SliceScaleBar, TextBox, and SoVolumeGroup.
| static SoDEPRECATED int SoSeparator::getNumRenderCaches | ( | ) | [static] |
Returns the current number of render caches.
- Deprecated:
-
Deprecated since Open Inventor 8000
Use SoDB::getNumRenderCaches instead.
| virtual SoType SoSeparator::getTypeId | ( | ) | const [virtual] |
Returns the type identifier for this specific instance.
Reimplemented from SoGroup.
Reimplemented in SoAnnotation, SoBillboard, SoExtSelection, SoFrontBufferGroup, SoLocateHighlight, SoOcclusionCulling, SoOctreeOrdering, SoRenderList, SoRenderToTarget, SoSelection, SoValueOrdering, SoVertexAttribFeedback, SoWWWAnchor, SoMultiDataSeparator, DicomInfo, Gnomon, PlaneGeometryIntersection, Ruler, SceneView, SliceOrientationMarkers, SliceScaleBar, TextBox, and SoVolumeGroup.
| static SoDEPRECATED void SoSeparator::setNumRenderCaches | ( | int | howMany | ) | [static] |
Sets the maximum number of render caches.
- Deprecated:
-
Deprecated since Open Inventor 8000
Use SoDB::setNumRenderCaches instead.
Member Data Documentation
Whether to cache during bounding box traversal.
Use enum CacheEnabled. Default is AUTO.
Whether to cache during OpenRTRT rendering traversal.
Use enum CacheEnabled. Default is AUTO.
NOTE: field available since Open Inventor 6.1
Specifies the fast edit mode of the separator node.
Use enum FastEditing. Default is DISABLE.
NOTE: field available since Open Inventor 5.0Whether to cull during picking traversal.
Use enum CacheEnabled. Default is AUTO.
Whether to cache during rendering traversal.
Use enum CacheEnabled. Default is AUTO.
Whether to cull during rendering traversal.
Use enum CacheEnabled. Default is AUTO.
Used during the ScaleViz depth compositing process, this field specifies which render unit (OIRU) will render the sub scene graph below this separator.
- 0, the sub scene graph is rendered on the local machine.
- > 0, the sub scene graph is rendered by the OIRU with the specified ID. If the specified renderUnitId is larger than the number of available render units, the sub scene graph is distributed to the specified render unit ID modulo the number of render units actually available. That is:
RenderUnitUsed = ((renderUnitId-1)%(ActualNumberOfRenderUnits))+1
- -1 (SO_RENDERUNITID_NONE), this separator is not assigned to any OIRU.
NOTE: This is the default value, which means that by default the sub scene graph is *not* distributed to any node. In order to use depth compositing, you must modify renderUnitId from the default, otherwise nothing will be rendered.
- -2 (SO_RENDERUNITID_INHERIT), this separator inherits its value from parent SoSeparator.
- -3 (SO_RENDERUNITID_ALL), the sub scene graph is rendered by all OIRUs.
The documentation for this class was generated from the following file:
- Inventor/nodes/SoSeparator.h