I work in a broadcast environment that delivers video signals over SDI as Fill (RGB) and Key (A) signals. My SDI rendering method produces a Texture for each SDI input and applies it to a material. I'm currently using a shader to populate two texture slots on the material in order to use the Key signal as an alpha mask.
This works pretty well, but it would be ideal to produce a single RGBA Texture from the two, and feed the material with the single Texture. A shader would not produce a new Texture I could get a handle on, so I'm looking for an alternate solution. So far, my google-fu has only produced shader-based approaches, or use somewhat inefficient getPixel/setPixel loops. Is there any affordance for generating a new Texture and defining channel-based inputs?