hm
http://www.ddisoftware.com/sd14-5d/
I was impressed with this review, at the time, in my ignorance, but given a strong interest in obtaining realistic image-quality and some real effort in pursuing it I have to say that this article is fatuous.
The real problem with the 5D compared to the Foveon-sensored Sigma cameras isn't the Bayer-blur. It's the limited gamut of the 5D, itself. Especially given AHD demosaicing which works *really* well in restoring detail especially with high-MP cameras, the Bayer-blur isn't that much of an issue...especially not compared to a 4MP Foveon sensor with a comparable pixel-pitch. The big issue is the fact that the 5D simply isn't capturing the blues and reds with much resolution due to the limited passband of the RGB filters. And in the greens also the SD14 has a much wider gamut.
It's just a much-better "color camera" than the 5D especially at low ISO.
It would have been a good idea to try another DSLR from another company first, or just to generate ICC profiles for these two cameras, before making such sweeping conclusions wrt Bayer vs Foveon. A Nikon DSLR would have done much better than a Canon, for that simple reason. Though still it wouldn't have matched the Sigma in color-gamut.
I'm posting this after doing just that:
http://www.flickr.com/photos/************/**********/
This was posted on Friday at 4:33pm EST
Let's see how long before I get banned from this forum for saying this.
...Steve, you're wrong. Plain and simple.
Though perhaps at the time you didn't have a Nikon fullframe DSLR to try, you still could have tried a subframe to see if it gave the same level of IQ and the same IQ issues that the 5D gave you.