DirectX post-processing pipeline

Question

I'm trying to make a simple post processing pipeline, depicted as follows with 3 effects in this case:

        |------|    |------|    |------|
   I -->|Shader|-O->|Shader|-I->|Shader|-O->
        |------|    |------|    |------|

I and O stand for a pair of 2 ShaderResourceViews/RenderTargetViews.

My plan was to do the following:

1. Create a shader resource and render target for each the input and the output
2. Apply the shader
3. Swap input and output to be used for the next shader
4. Goto 2.

This sounds somewhat confusing, but basically the same 2 ShaderResourceViews /RenderTargetViews should be reused for each stage of the post processing pipeline.

Is this possible, or do I have to create separate input and output ShaderResourceViews /RenderTargetViews for each stage of the postprocessing pipeline?

Answer 1

You'll need to create the textures with both the D3D11_BIND_RENDER_TARGET and D3D11_BIND_SHADER_RESOURCE bind flags, and you must take care to unbind them from their previous assignment before swapping them, but it should work fine.

Also, make sure the format you want to use is listed in the supported render target formats, and not listed in the formats not supported for Texture2D here: http://msdn.microsoft.com/en-us/library/windows/desktop/ff471325(v=vs.85).aspx