NAME
SoCamera -
The SoCamera class is the abstract base class for camera definition
nodes.
To be able to view a scene, one needs to have a camera in the scene
graph. A camera node will set up the projection and viewing matrices
for rendering of the geometry in the scene.
SYNOPSIS
#include <Inventor/nodes/SoCamera.h>
Inherits SoNode.
Inherited by SoFrustumCamera, SoOrthographicCamera, and
SoPerspectiveCamera.
Public Types
enum ViewportMapping { CROP_VIEWPORT_FILL_FRAME,
CROP_VIEWPORT_LINE_FRAME, CROP_VIEWPORT_NO_FRAME, ADJUST_CAMERA,
LEAVE_ALONE }
enum StereoMode { MONOSCOPIC, LEFT_VIEW, RIGHT_VIEW }
Public Member Functions
void pointAt (const SbVec3f &targetpoint)
void pointAt (const SbVec3f &targetpoint, const SbVec3f &upvector)
virtual void scaleHeight (float scalefactor)=0
virtual SbViewVolume getViewVolume (float useaspectratio=0.0f) const =0
void viewAll (SoNode *const sceneroot, const SbViewportRegion
&vpregion, const float slack=1.0f)
void viewAll (SoPath *const path, const SbViewportRegion &vpregion,
const float slack=1.0f)
SbViewportRegion getViewportBounds (const SbViewportRegion ®ion)
const
void setStereoMode (StereoMode mode)
StereoMode getStereoMode (void) const
void setStereoAdjustment (float adjustment)
float getStereoAdjustment (void) const
void setBalanceAdjustment (float adjustment)
float getBalanceAdjustment (void) const
virtual void doAction (SoAction *action)
virtual void callback (SoCallbackAction *action)
virtual void GLRender (SoGLRenderAction *action)
virtual void audioRender (SoAudioRenderAction *action)
virtual void getBoundingBox (SoGetBoundingBoxAction *action)
virtual void handleEvent (SoHandleEventAction *action)
virtual void rayPick (SoRayPickAction *action)
virtual void getPrimitiveCount (SoGetPrimitiveCountAction *action)
Static Public Member Functions
static void initClass (void)
Public Attributes
SoSFEnum viewportMapping
SoSFVec3f position
SoSFRotation orientation
SoSFFloat aspectRatio
SoSFFloat nearDistance
SoSFFloat farDistance
SoSFFloat focalDistance
Protected Member Functions
SoCamera (void)
virtual ~SoCamera ()
virtual void viewBoundingBox (const SbBox3f &box, float aspect, float
slack)=0
virtual void jitter (int numpasses, int curpass, const SbViewportRegion
&vpreg, SbVec3f &jitteramount) const
Detailed Description
The SoCamera class is the abstract base class for camera definition
nodes.
To be able to view a scene, one needs to have a camera in the scene
graph. A camera node will set up the projection and viewing matrices
for rendering of the geometry in the scene.
This node just defines the abstract interface by collecting common
fields that all camera type nodes needs. Use the non-abstract camera
node subclasses within a scene graph. The ones that are default part of
the Coin library are SoPerspectiveCamera and SoOrthographicCamera,
which uses the two different projections given by their name.
Note that the viewer components of the GUI glue libraries of Coin
(SoXt, SoQt, SoWin, etc) will automatically insert a camera into a
scene graph if none has been defined.
It is possible to have more than one camera in a scene graph. One
common trick is for instance to use a second camera to display static
geometry or overlay geometry (e.g. for head-up displays (’HUD’)), as
shown by this example code:
#include <Inventor/Qt/SoQt.h>
#include <Inventor/Qt/viewers/SoQtExaminerViewer.h>
#include <Inventor/nodes/SoNodes.h>
int
main(int argc, char ** argv)
{
QWidget * mainwin = SoQt::init(argv[0]);
SoSeparator * root = new SoSeparator;
root->ref();
// Adds a camera and a red cone. The first camera found in the
// scene graph by the SoQtExaminerViewer will be picked up and
// initialized automatically.
root->addChild(new SoPerspectiveCamera);
SoMaterial * material = new SoMaterial;
material->diffuseColor.setValue(1.0, 0.0, 0.0);
root->addChild(material);
root->addChild(new SoCone);
// Set up a second camera for the remaining geometry. This camera
// will not be picked up and influenced by the viewer, so the
// geometry will be kept static.
SoPerspectiveCamera * pcam = new SoPerspectiveCamera;
pcam->position = SbVec3f(0, 0, 5);
pcam->nearDistance = 0.1;
pcam->farDistance = 10;
root->addChild(pcam);
// Adds a green cone to demonstrate static geometry.
SoMaterial * greenmaterial = new SoMaterial;
greenmaterial->diffuseColor.setValue(0, 1.0, 0.0);
root->addChild(greenmaterial);
root->addChild(new SoCone);
SoQtExaminerViewer * viewer = new SoQtExaminerViewer(mainwin);
viewer->setSceneGraph(root);
viewer->show();
SoQt::show(mainwin);
SoQt::mainLoop();
delete viewer;
root->unref();
return 0;
}
Member Enumeration Documentation
enum SoCamera::ViewportMapping Enumerates the available possibilities for
how the render frame should map the viewport.
enum SoCamera::StereoMode Enumerates the possible stereo modes.
Enumerator:
MONOSCOPIC
No stereo.
LEFT_VIEW
Left view.
RIGHT_VIEW
Right view.
Constructor & Destructor Documentation
SoCamera::SoCamera (void) [protected] Constructor.
SoCamera::~SoCamera () [protected, virtual] Destructor.
Member Function Documentation
void SoCamera::initClass (void) [static] Sets up initialization for data
common to all instances of this class, like submitting necessary
information to the Coin type system.
Reimplemented from SoNode.
Reimplemented in SoFrustumCamera, SoOrthographicCamera, and
SoPerspectiveCamera.
void SoCamera::pointAt (const SbVec3f & targetpoint) Reorients the camera
so that it points towards targetpoint. The positive y-axis is used as
the up vector of the camera, unless the new camera direction is
parallel to this axis, in which case the positive z-axis will be used
instead.
void SoCamera::pointAt (const SbVec3f & targetpoint, const SbVec3f &
upvector) Reorients the camera so that it points towards targetpoint,
using upvector as the camera up vector.
This function is an extension for Coin, and it is not available in the
original SGI Open Inventor v2.1 API.
void SoCamera::scaleHeight (float scalefactor) [pure virtual] Sets a
scalefactor for the height of the camera viewport. What ’viewport
height’ means exactly in this context depends on the camera model. See
documentation in subclasses.
Implemented in SoFrustumCamera, SoOrthographicCamera, and
SoPerspectiveCamera.
SbViewVolume SoCamera::getViewVolume (float useaspectratio = 0.0f) const
[pure virtual] Returns total view volume covered by the camera under
the current settings.
This view volume is not adjusted to account for viewport mapping. If
you want the same view volume as the one used during rendering, you
should do something like this:
SbViewVolume vv;
float aspectratio = myviewport.getViewportAspectRatio();
switch (camera->viewportMapping.getValue()) {
case SoCamera::CROP_VIEWPORT_FILL_FRAME:
case SoCamera::CROP_VIEWPORT_LINE_FRAME:
case SoCamera::CROP_VIEWPORT_NO_FRAME:
vv = camera->getViewVolume(0.0f);
break;
case SoCamera::ADJUST_CAMERA:
vv = camera->getViewVolume(aspectratio);
if (aspectratio < 1.0f) vv.scale(1.0f / aspectratio);
break;
case SoCamera::LEAVE_ALONE:
vv = camera->getViewVolume(0.0f);
break;
default:
assert(0 && "unknown viewport mapping");
break;
}
.fi
Also, for the CROPPED viewport mappings, the viewport might be changed if the viewport aspect ratio is not equal to the camera aspect ratio. See SoCamera::getView() to see how this is done.
Implemented in SoFrustumCamera, SoOrthographicCamera, and SoPerspectiveCamera.
void SoCamera::viewAll (SoNode *const sceneroot, const SbViewportRegion &
vpregion, const float slack = 1.0f) Position the camera so that all
geometry of the scene from sceneroot is contained in the view volume of
the camera, while keeping the camera orientation constant.
Finds the bounding box of the scene and calls
SoCamera::viewBoundingBox(). A bounding sphere will be calculated from
the scene bounding box, so the camera will ’view all’ even when the
scene is rotated, in any way.
The slack argument gives a multiplication factor to the distance the
camera is supposed to move out from the sceneroot mid-point.
A value less than 1.0 for the slack argument will therefore cause the
camera to come closer to the scene, a value of 1.0 will position the
camera as exactly outside the scene bounding sphere, and a value larger
than 1.0 will give ’extra slack’ versus the scene bounding sphere.
void SoCamera::viewAll (SoPath *const path, const SbViewportRegion &
vpregion, const float slack = 1.0f) Position the camera so all geometry
of the scene in path is contained in the view volume of the camera.
Finds the bounding box of the scene and calls
SoCamera::viewBoundingBox().
SbViewportRegion SoCamera::getViewportBounds (const SbViewportRegion &
region) const Based in the SoCamera::viewportMapping setting, convert
the values of region to the viewport region we will actually render
into.
void SoCamera::setStereoMode (StereoMode mode) Sets the stereo mode.
SoCamera::StereoMode SoCamera::getStereoMode (void) const Returns the
stereo mode.
void SoCamera::setStereoAdjustment (float adjustment) Sets the stereo
adjustment. This is the distance between the left and right ’eye’ when
doing stereo rendering.
When doing stereo rendering, Coin will render two views, one for the
left eye, and one for the right eye. The stereo adjustment is, a bit
simplified, how much the camera is translated along the local X-axis
between the left and the right view.
The default distance is 0.1, which is chosen since it’s the approximate
distance between the human eyes.
To create a nice looking and visible stereo effect, the application
programmer will often have to adjust this value. If all you want to do
is examine simple stand-alone 3D objects, it is possible to calculate a
stereo offset based on the bounding box of the 3D model (or scale the
model down to an appropriate size).
However, if you have a large scene, where you want to fly around in the
scene, and see stereo on different objects as you approach them, you
can’t calculate the stereo offset based on the bounding box of the
scene, but rather use a stereo offset based on the scale of the
individual objects/details you want to examine.
Please note that it’s important to set a sensible focal distance when
doing stereo rendering. See setBalanceAdjustment() for information
about how the focal distance affects the stereo rendering.
See also:
setBalanceAdjustment()
float SoCamera::getStereoAdjustment (void) const Returns the stereo
adjustment.
See also:
setStereoAdjustment()
void SoCamera::setBalanceAdjustment (float adjustment) Sets the stereo
balance adjustment. This is a factor that enables you to move the zero
parallax plane. Geometry in front of the zero parallax plane will
appear to be in front of the screen.
The balance adjustment is multiplied with the focal distance to find
the zero parallax plane. The default value is 1.0, and the zero
parallax plane is then at the focal point.
See also:
SoCamera::focalDistance
float SoCamera::getBalanceAdjustment (void) const Returns the stereo
balance adjustment.
See also:
setBalanceAdjustment()
void SoCamera::doAction (SoAction * action) [virtual] This function
performs the typical operation of a node for any action.
Reimplemented from SoNode.
void SoCamera::callback (SoCallbackAction * action) [virtual] Action method
for SoCallbackAction.
Simply updates the state according to how the node behaves for the
render action, so the application programmer can use the
SoCallbackAction for extracting information about the scene graph.
Reimplemented from SoNode.
void SoCamera::GLRender (SoGLRenderAction * action) [virtual] Action method
for the SoGLRenderAction.
This is called during rendering traversals. Nodes influencing the
rendering state in any way or who wants to throw geometry primitives at
OpenGL overrides this method.
Reimplemented from SoNode.
void SoCamera::audioRender (SoAudioRenderAction * action) [virtual] Action
method for SoAudioRenderAction.
Does common processing for SoAudioRenderAction action instances.
Reimplemented from SoNode.
void SoCamera::getBoundingBox (SoGetBoundingBoxAction * action) [virtual]
Action method for the SoGetBoundingBoxAction.
Calculates bounding box and center coordinates for node and modifies
the values of the action to encompass the bounding box for this node
and to shift the center point for the scene more towards the one for
this node.
Nodes influencing how geometry nodes calculates their bounding box also
overrides this method to change the relevant state variables.
Reimplemented from SoNode.
void SoCamera::handleEvent (SoHandleEventAction * action) [virtual] Picking
actions can be triggered during handle event action traversal, and to
do picking we need to know the camera state.
See also:
SoCamera::rayPick()
Reimplemented from SoNode.
void SoCamera::rayPick (SoRayPickAction * action) [virtual] Action method
for SoRayPickAction.
Checks the ray specification of the action and tests for intersection
with the data of the node.
Nodes influencing relevant state variables for how picking is done also
overrides this method.
Reimplemented from SoNode.
void SoCamera::getPrimitiveCount (SoGetPrimitiveCountAction * action)
[virtual] Action method for the SoGetPrimitiveCountAction.
Calculates the number of triangle, line segment and point primitives
for the node and adds these to the counters of the action.
Nodes influencing how geometry nodes calculates their primitive count
also overrides this method to change the relevant state variables.
Reimplemented from SoNode.
void SoCamera::viewBoundingBox (const SbBox3f & box, float aspect, float
slack) [protected, pure virtual] Convenience method for setting up the
camera definition to cover the given bounding box with the given aspect
ratio. Multiplies the exact dimensions with a slack factor to have some
space between the rendered model and the borders of the rendering area.
If you define your own camera node class, be aware that this method
should not set the orientation field of the camera, only the position,
focal distance and near and far clipping planes.
Implemented in SoFrustumCamera, SoOrthographicCamera, and
SoPerspectiveCamera.
void SoCamera::jitter (int numpasses, int curpass, const SbViewportRegion &
vpreg, SbVec3f & jitteramount) const [protected, virtual] ’Jitter’ the
camera according to the current rendering pass (curpass), to get an
antialiased rendering of the scene when doing multipass rendering.
Member Data Documentation
SoSFEnum SoCamera::viewportMapping Set up how the render frame should map
the viewport. The default is SoCamera::ADJUST_CAMERA.
SoSFVec3f SoCamera::position Camera position. Defaults to <0,0,1>.
SoSFRotation SoCamera::orientation Camera orientation specified as a
rotation value from the default orientation where the camera is
pointing along the negative z-axis, with ’up’ along the positive y-
axis.
E.g., to rotate the camera to point along the X axis:
mycamera->orientation.setValue(SbRotation(SbVec3f(0, 1, 0), M_PI / 2.0f));
For queries, e.g. to get the current ’up’ and ’look at’ vectors of the
camera:
SbRotation camrot = mycamera->orientation.getValue();
SbVec3f upvec(0, 1, 0); // init to default up vector
camrot.multVec(upvec, upvec);
SbVec3f lookat(0, 0, -1); // init to default view direction vector
camrot.multVec(lookat, lookat);
SoSFFloat SoCamera::aspectRatio Aspect ratio for the camera (i.e. width /
height). Defaults to 1.0.
SoSFFloat SoCamera::nearDistance Distance from camera position to the near
clipping plane in the camera’s view volume.
Default value is 1.0. Value must be larger than 0.0, or it will not be
possible to construct a valid viewing volume (for perspective
rendering, at least).
If you use one of the viewer components from the So[Xt|Qt|Win|Gtk] GUI
libraries provided by Kongsberg Oil & Gas Technologies, they will
automatically update this value for the scene camera according to the
scene bounding box. Ditto for the far clipping plane.
See also:
SoCamera::farDistance
SoSFFloat SoCamera::farDistance Distance from camera position to the far
clipping plane in the camera’s view volume.
Default value is 10.0. Must be larger than the SoCamera::nearDistance
value, or it will not be possible to construct a valid viewing volume.
Note that the range [nearDistance, farDistance] decides the dynamic
range of the Z-buffer in the underlying polygon-rendering rasterizer.
What this means is that if the near and far clipping planes of the
camera are wide apart, the possibility of visual artifacts will
increase. The artifacts will manifest themselves in the form of
flickering of primitives close in depth.
It is therefore a good idea to keep the near and far clipping planes of
your camera(s) as closely fitted around the geometry of the scene graph
as possible.
See also:
SoCamera::nearDistance, SoPolygonOffset
SoSFFloat SoCamera::focalDistance Distance from camera position to center
of scene.
Author
Generated automatically by Doxygen for Coin from the source code.