pygfx.renderers.wgpu.EffectPass

class pygfx.renderers.wgpu.EffectPass

Bases: FullQuadPass

Base class to do post-processing effect passes, converting one image into another.

USES_DEPTH = False: Overloadable class attribute to state whether bindings to the depth buffer are needed.

uniform_type = {'time': 'f4'}

Overloadable class attribute that defines the structure of the uniform struct.

In a subclass you should use something like this:

uniform_type = dict(
    EffectPass.uniform_type,
    color="4xf4",
    strength="f4",
)

wgsl = 'EffectPass_needs_to_be_subclassed(and_its_wgsl_attr_overloaded);'

Overloadable class attribute that contains the WGSL shading code for the fragment shader.

Use the following template:

@fragment
fn fs_main(varyings: Varyings) -> @location(0) vec4<f32> {

    // Available variables:
    // colorTex - the texture containing the rendered image, or the previous effect pass.
    // depthTex - the texture containing the renderd depth values.
    // texSampler - a sampler to use for the above.
    // varyings.position - the position in physical pixels (a vec4f).
    // varyings.texCoord - the coordinate in the textures (a vec2f).
    // u_effect.time - the current time in seconds, changes each frame.
    // u_effect.xx - whatever uniforms you added.

    // Calculate the pixel index, e.g. if you want to use textureLoad().
    let texIndex = vec2i(varyings.position.xy);

    // To simply copy the image:
    return textureSample(colorTex, texSampler, varyings.texCoord);
}