Originally posted by mikemike I think the principle of foveon is sound and I will not write-it off based only on the 2 generations of cameras released by Sigma.
Maybe. But maybe not.
I am concerned that the foveon principle implies filtering of light. With a perfectly transparent sensor, colors couldn't be separated. So, the intransparency is vltal to foveon (except for an almost monochrome sensor with huge chroma noise). It seriously lowers its quantum efficiency. Doing the math, I wouldn't be surprised if foveon doesn't even have a theoretical advantage.
The quantum efficiency for Bayer sensors is about 40% per color and with green carrying about half the energy, its 20% for green and 13% overall. I am not sure that foveon is better by principle. I've seen no source claiming it.
This is why I said I am more keen in seeing research heading towards bringing overall QE above 50%.