It has no bayer mask = no interpolation = really effective pixels, not those "effective pixels" all other manufacturers claim, while they use bayer mask and thus each pixel is for only ONE of the three RGB colors (on Sigmas DP1 Foveon each pixel records all 3 colors per pixel). Sigma claims to be equivalent of 16Mpix sensor with Bayer mask, because of the square RGGB color mask in front of those sensors - in fact it bahaves more like 12Mpix to be honest, but all those reviews saying "its just 4Mpix" are simly complete nonsense. Many of 10Mpix DSLRs with average lenses (and I dont mean cheap kit lenses with word average) will be in big trouble against Sigmas result.

It has superb optics and picture quality, however fixed focal length and its pretty slow. I can imagine DP1 for landscape shots, but definitely walkaround camera. Its definitely well above cheap setups like cheap DSLRs with kit lenses, I would place it about semi-pro. I also saw some reviews and tests (I think it was on Popco), where guys tested Sigma DP1 against Canon 5D + 17-40/4L and Sigma was obviously winner in the colors, detail and dynamic range. There were some issues with green color cast on higher iso, but I think it was later solved with firmware update.

Conclusion: Yep, pretty good for landscapes or other uses, where the slow speed and wide fixed focel length arent problem. It delivers excelent picture quality.

Most microstock agencies accept this size. More is obviously better, but occassionally I still submit a 4 MP image.

but interpolation is not allowed.

OxfordSquare:
You probably missed the point - EVERY sensor except Foveon uses brutal firmware interpolation in camera - every pixel records ONLY one from RGB color so you must interpolate the rest to get normal picture! So in fact Foveon is the only one widely available non-interpolated sensor. You guys better do some reading about how it works. Interpolation is not allowed? Then you should not post any picture from any camera except Sigmas Foveon sensor equipped ones, right? I will tell it once again - to get the same size of the picture as from any other sensor, you need to interpolate image recorded by Sigma DP1 4x and its roughly the same interpolation and size recorded by common camera. Its impossible to get uninterpolated picture from anything else, you just do that in PC later and not in camera! Every other camera DOES interpolate during firmware trasformation of raw data into picture. Even raw is interpolated, if not, you will get just crazy map of red, green and blue dots!   

Dook:
DP1 will very likely smash any low end DSLR with kit or average lenses in detail, colors, dynamic range and sharpness, so what are you talking about? Its completely different technology and much more advanced in some points, you cannot compare that based on common "sensor size" myths. Eg. "small" 4/3 sensor in E-3 has about same noise as "large" 40D up to iso800 - which is far beyond the iso you should use for stock. So what is the advance of "larger" sensor? None, right? Its not that simple as just comparing megapixels or sensor size. It depends on many more things then just this.

Some comparisons:

http://www.pbase.com/hughden/dp1comp

http://www.seriouscompacts.com/2008/04/dp1-shootout-pt-3-landscape-detail.html

http://www.prime-junta.net/pont/Reviews/040_Sigma_DP1/_Sigma_DP1.html?page=4

Extensive info: http://photography-on-the.net/forum/showthread.php?s=e61eb9f509a5233e3d8c51ff815cf46b&t=445792

It is not good idea to shoot stock with that small sensors, in general. Your acceptance ratio will be much smaller, thats for sure.

Do you even have a clue what the sensor size is on DP1 ? (Hint: It's 7X bigger than in Canon G-series, and same size as in some DSLRs)

edit: DSLR!

Huh
lets revitalize this old post. I want to buy used Canon 5D until end of this year but today my friend put bug in my ear about Sigma SD14 with Foveon chip.
In my first look at forum at iStock, and they vision of this camera is more than disaster.
I really dont know if there "big" guys are scared of cheap price of that camera and always pointing at they TOS that enlargement is not allowed (maybe because they invest big money on they toys for "big" guys).
From my POW enlargements in Foveon or Color Cmos interpolation in camera is the same thing, result is what is count at the end.
I am not max pixel fanatic, I just want to make very big, technically and stockworhty images with cheap gear (I think you too want that). For now I have plastic Canon 350D with 8Mpix and last year most of my shoots are panoramic simple studio shots (to get more Mpixs for Alamy) shooted with 2 or more images and joined in PTGui.
After that joining images in PTGui there is no exif or other data whit camera is that image shooted and they are accepted on all sites (I mean iStock too).
I even try much more perversion to shoot very close up in 3 and more focus ranges and for every focus range shoot different EV from -3 to +3, then make HDR of every focus range and finally join them in multifocus application to get full focus with HDR tone range of big CloseUp and 2, 3 or more time bigger than resolution what my camera can bring (also approved on iStock so there is no problems with that issue about they TOS, only if after this post they reviewers look first in exif or first Exif scrip scan camera model and reject it without any human view on image).

I would bee happy if someone of you can tell me can I give with that camera same quality like on Canon 5D.
Last year I was shooting few days with 5D and from my POW I am impressed.

Normally before buying I will make my own tests samples with that Sigma.
For now before go to shopping I dont want to make any prejudice. I am so nervous to compare this two cameras.

Price comparison is on Sigmas side, new one is double cheaper than 5 or more years used 5D with 20000-30000 or more shoots or with fairy tales for kids from second hand sellers that model is from shop window with few hundreds of shoots?!?
THNX for reply

THNX
I just realize that this Faveon chip and this story is too old, I think its somehow new story about max year old.
Anyhow I will test it for free but for now 5D is better option as I see.