In article , David J.
Littleboy wrote:

Are there any differences that are noticeable? It'd be pretty easy to
compare in-camera with ACR/LR. And there are test chart results on Dpreview
and the other sites for just about every camera. So we (if we look) have a
very good idea of what these things do. Most of the difference I see is more
in the AA filter, with cameras like the D70 and 5D letting more color
aliasing into the picture.

FWIW, here's the old 5D with old Lightroom vs. RawShooterProfessional. RSP's
low-ISO sharpening was overmuch in the extreme, but it kept high ISO noise
under better control. (Maybe it'd be fun to redo this with LR 3.2 vs. 5D2's
in-camera processing.)

http://www.pbase.com/davidjl/image/85596346/original

Again, my basic axe to grind here is that (a) assuming you want accurate
imaging you have to have an AA filter, and (b) once you have an AA filter,
the difference between Bayer and any of the 3-color per pixel methods isn't
going to be all that large*. Oh, yes. (c) Bayer images really look great.
Why aren't we being insanely happy at how good 5D2 images look printed at
12x18 (remember, 35mm film looks like god-awful crap by comparison) than
screaming how terrible Bayer is?

*: How much "stronger" an AA filter is needed to suppress color artifacts in
Bayer vs. only luminance artifacts in a 3-color sensor is, of course, the
question. My guess is that that difference wouldn't be all that large, but
others' intuitions will differ.

Enough techno-babble! Who gives a ****? Sigma/Foveon is crap. Just look
at the images the thing produces. Sigma is a company which revolves
around screwing people. Move on...nothing to see here.