OSG学习:用多通道(multiple passes)实现透明度

osgFX库提供了一个用于多通道渲染(multi-pass rendering)的框架。每个你想要渲染的子图都应该被添加到osgFX::Effect节点,多通道技术的定义和使用都可以在这个节点中完成。你可能已经熟悉一些预定义的效果,例如osgFX::Scribe和osgFX::Outline。但是在这个教程中,我们的任务是我们自己设计一个多通道技术。这就是所谓的多通道透明度,当我们在透明模式(transparent mode)下它可以消除一些错误。

如何使用

  1. 添加必要的头文件:
#include <osg/BlendFunc>
#include <osg/ColorMask>
#include <osg/Depth>
#include <osg/Material>
#include <osgDB/ReadFile>
#include <osgFX/Effect>
#include <osgViewer/Viewer>

2.首先,我们将提供一个新技术,这个技术继承于osgFX::Technique节点。validate()方法是用于检查现在硬件是否支持这个技术,如果一切正常会返回true。

class TransparencyTechnique : public osgFX::Technique
{
public:
TransparencyTechnique() : osgFX::Technique() {}
virtual bool validate( osg::State& ss ) const
{
return true;
}
protected:
virtual void define_passes();
};

3.在这个类中,另一个必须有的方法是define_passes()。这是用于定义多通道的。在这个教程中,我们将有两个通道:一个是取消颜色掩膜(color mask)以及如果深度缓冲区值(depth buffer value)小于现在的就记录下来;另一个将使用颜色缓冲区,但只有在深度值等于已经记录下来的值的时候写进去。(the second one will enable the useage of the color buffer but only to write to it when the depth value equals to the recorded one.)

osg::ref_ptr<osg::StateSet> ss = new osg::StateSet;
ss->setAttributeAndModes( new osg::ColorMask(
false, false, false, false) );
ss->setAttributeAndModes( new osg::Depth(osg::Depth::LESS) );
addPass( ss.get() );
ss = new osg::StateSet;
ss->setAttributeAndModes( new osg::ColorMask(
true, true, true, true) );
ss->setAttributeAndModes( new osg::Depth(osg::Depth::EQUAL) );
addPass( ss.get() );
  1. 我们设计完技术之后,现在我们可以声明effect类,把技术添加到它的define_techniques()方法中。这里的META_Effect宏用于定义effect的基本方法:库名,类名,作者名和描述。
class TransparencyNode : public osgFX::Effect
{
public:
TransparencyNode() : osgFX::Effect() {}
TransparencyNode( const TransparencyNode& copy,
const osg::CopyOp op=osg::CopyOp::SHALLOW_COPY )
: osgFX::Effect(copy, op) {}
META_Effect( osgFX, TransparencyNode, "TransparencyNode",
"", "" );
protected:
virtual bool define_techniques()
{
addTechnique(new TransparencyTechnique);
return true;
}
};
  1. 在main函数中,我们将使用景点的Cessna模型。不过,为了使它透明,我们将在上面添加一个新的材料,把散射颜色(diffuse color)的alpha通道设置为小于1的值,以及把TRANSPARENT_BIN这个hint应用到state set中。
osg::Node* loadedModel = osgDB::readNodeFile( "cessna.osg" );
osg::ref_ptr<osg::Material> material = new osg::Material;
material->setAmbient( osg::Material::FRONT_AND_BACK,
osg::Vec4(0.0f, 0.0f, 0.0f, 1.0f) );
material->setDiffuse( osg::Material::FRONT_AND_BACK,
osg::Vec4(1.0f, 1.0f, 1.0f, 0.5f) );
loadedModel->getOrCreateStateSet()->setAttributeAndModes(
material.get(), osg::StateAttribute::ON|osg::StateAttribute::OVE
RRIDE );
loadedModel->getOrCreateStateSet()->setAttributeAndModes(
new osg::BlendFunc );
loadedModel->getOrCreateStateSet()->setRenderingHint(
osg::StateSet::TRANSPARENT_BIN );

6.创建一个新定义类的实例并添加到模型中。

osg::ref_ptr<TransparencyNode> fxNode = new TransparencyNode;
fxNode->addChild( loadedModel );
//Start the viewer now.
osgViewer::Viewer viewer;
viewer.setSceneData( fxNode.get() );
return viewer.run();

7.现在你将看到适当渲染过的透明的Cessna。
这里写图片描述

全部代码:


#include <osg/BlendFunc>
#include <osg/ColorMask>
#include <osg/Depth>
#include <osg/Material>
#include <osgDB/ReadFile>
#include <osgFX/Effect>
#include <osgViewer/Viewer>

#include "CommonFunctions"

class TransparencyTechnique : public osgFX::Technique
{
public:
    TransparencyTechnique() : osgFX::Technique() {}
    virtual bool validate( osg::State& ss ) const { return true; }

protected:
    virtual void define_passes()
    {
        osg::ref_ptr<osg::StateSet> ss = new osg::StateSet;
        ss->setAttributeAndModes( new osg::ColorMask(false, false, false, false) );
        ss->setAttributeAndModes( new osg::Depth(osg::Depth::LESS) );
        addPass( ss.get() );

        ss = new osg::StateSet;
        ss->setAttributeAndModes( new osg::ColorMask(true, true, true, true) );
        ss->setAttributeAndModes( new osg::Depth(osg::Depth::EQUAL) );
        addPass( ss.get() );
    }
};

class TransparencyNode : public osgFX::Effect
{
public:
    TransparencyNode() : osgFX::Effect() {}
    TransparencyNode( const TransparencyNode& copy, const osg::CopyOp op=osg::CopyOp::SHALLOW_COPY )
    :   osgFX::Effect(copy, op) {}

    META_Effect( osgFX, TransparencyNode, "TransparencyNode", "", "" );

protected:
    virtual bool define_techniques()
    { addTechnique(new TransparencyTechnique); return true; }
};

int main( int argc, char** argv )
{
    osg::Node* loadedModel = osgDB::readNodeFile( "cessna.osg" );

    osg::ref_ptr<osg::Material> material = new osg::Material;
    material->setAmbient( osg::Material::FRONT_AND_BACK, osg::Vec4(0.0f, 0.0f, 0.0f, 1.0f) );
    material->setDiffuse( osg::Material::FRONT_AND_BACK, osg::Vec4(1.0f, 1.0f, 1.0f, 0.5f) );
    loadedModel->getOrCreateStateSet()->setAttributeAndModes(
        material.get(), osg::StateAttribute::ON|osg::StateAttribute::OVERRIDE );
    loadedModel->getOrCreateStateSet()->setAttributeAndModes( new osg::BlendFunc );
    loadedModel->getOrCreateStateSet()->setRenderingHint( osg::StateSet::TRANSPARENT_BIN );

    osg::ref_ptr<TransparencyNode> fxNode = new TransparencyNode;
    fxNode->addChild( loadedModel );

    osgViewer::Viewer viewer;
    viewer.setSceneData( fxNode.get() );
    return viewer.run();
}

原理

多通道透明度技术会绘制物体两次。第一个通道(见后面代码)仅仅更新深度缓冲区(depth buffer),以及找到最前面的多边形:

osg::ref_ptr<osg::StateSet> ss = new osg::StateSet;
ss->setAttributeAndModes( new osg::ColorMask(
false, false, false, false) );
ss->setAttributeAndModes( new osg::Depth(osg::Depth::LESS) );
addPass( ss.get() );

第二个通道将会绘制进颜色缓冲区(color buffer),但由于在第一个通道中设置的深度值,只有前面的多边形能够通过深度检测,它们的颜色将被绘制,与现在的混合。这避免了我们之前讨论过的顺序问题。代码片段如下:

ss = new osg::StateSet;
ss->setAttributeAndModes( new osg::ColorMask(
true, true, true, true) );
ss->setAttributeAndModes( new osg::Depth(osg::Depth::EQUAL) );
addPass( ss.get() );

选自《OSG3 Cookbook》第六章

posted @ 2017-07-17 17:48  huahai  阅读(1099)  评论(0编辑  收藏  举报