[DX10] DXGI_FORMAT_R16G16B16A16_UINT problems
Hello,
I'm having problems setting the values in a texture that has DXGI_FORMAT_R16G16B16A16_UINT type.
When rendering the 3D model I set the values for this texture manually to 1 (for example, or 2 or whatever...) in the pixel shader, so that this value will get written to the texture that uses DXGI_FORMAT_R16G16B16A16_UINT format. But it doesn't get written to it.
In another pixel shader (the one I use for the fullscreen quad), the values for this texture are all 0 (so I presume since I tried multiplying the final output color by the value stored in the DXGI_FORMAT_R16G16B16A16_UINT texture and it's black).
This texture is bound with 3 others (multiple render targets).
Any ideas? (I'd post some code but since all other texture work as they should I don't know what to post...). Does this format has some limitations (if I change the format to DXGI_FORMAT_R16G16B16A16_UNORM it works normally)?
Ok, I haven't solved this but lets leave this then...
I have another question: how to read from a texture an int (0-255), write it to a render target and than read it from this render target in another render target BUT without converting it to float or clamping the int to [0,1]. Is this possible?
I'd like this so that I can check in the pixel shader for a certain RGB value (0-255) and do some branching based on it (I tried it with floats and it (partially) works, so I'd like ints).
I have another question: how to read from a texture an int (0-255), write it to a render target and than read it from this render target in another render target BUT without converting it to float or clamping the int to [0,1]. Is this possible?
I'd like this so that I can check in the pixel shader for a certain RGB value (0-255) and do some branching based on it (I tried it with floats and it (partially) works, so I'd like ints).
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement