Qt Quick 3D - Cascaded Shadow Maps Example

Demonstrates cascaded shadow maps.

This example demonstrates how to use a DirectionalLight with cascaded shadow maps.

Let's have a look at the scene's QML code:

View3D {
    id: view
    x: settings.viewX
    y: 0
    width: parent.width - settings.viewX
    height: parent.height
    camera: camera
    environment: SceneEnvironment {
        clearColor: "lightblue"
        backgroundMode: SceneEnvironment.Color
        antialiasingMode: SceneEnvironment.MSAA
        antialiasingQuality: SceneEnvironment.High
    }
    PerspectiveCamera {
        id: camera
        position: Qt.vector3d(458, 300, 515)
        eulerRotation: Qt.vector3d(-14, 19, 0)
        clipFar: settings.clipFar
    }
    DirectionalLight {
        visible: true
        castsShadow: true
        shadowFactor: settings.shadowFactor
        eulerRotation: settings.eulerRotation
        csmSplit1: settings.csmSplit1
        csmSplit2: settings.csmSplit2
        csmSplit3: settings.csmSplit3
        csmNumSplits: settings.csmNumSplits
        shadowMapQuality: settings.shadowMapQuality
        csmBlendRatio: settings.csmBlendRatio
        shadowBias: settings.shadowBias
        pcfFactor: settings.pcfFactor
        softShadowQuality: settings.softShadowQuality
        shadowMapFar: settings.shadowMapFar
        lockShadowmapTexels: settings.lockShadowmapTexels
    }
    Model {
        id: ground
        source: "#Cube"
        scale: Qt.vector3d(25, 0.01, 135)
        z: -5500
        materials: PrincipledMaterial {
            baseColor: "gray"
        }
        castsShadows: false
    }
    Node {
        id: shapeSpawner
        Component.onCompleted: {
            var conesAndCylinderTrio = Qt.createComponent("ConesAndCylinderTrio.qml")
            var z_pos = 0
            for (var i = 0; i < 25; i++) {
                conesAndCylinderTrio.incubateObject(shapeSpawner, {
                                                    "z_positions": [
                                                            z_pos,
                                                            z_pos - 125,
                                                            z_pos - 250
                                                        ]})
                z_pos -= 450
            }
        }
    }
}
					

The scene consists of a DirectionalLight PerspectiveCamera , a floor Model , and a long array of cones and cylinder Model s. This long array is laid out like this to show how cascades can help increase the graphical fidelity of the scene. With just one shadow map it will cause artifacts but with cascades these can be remedied. The DirectionalLight contains a lot of properties that are tweakable through a settings pane in the UI.

The reader is encouraged to see the 阴影映射 article for more information on the different properties and to try out the example to see how they interact.

范例工程 @ code.qt.io