I just took a look at the image you directed me to. But I didn't stop there, because without a comparison it doesn't mean a lot. Is the defect due to the jpegging, is it the limit of a digital sensor, or the resolution? I couldn't tell. So I went to another digital review site www.imaging-resource.com, and I used the "Comparometer", and took a look at the image of the house and the musicians, both which contain numerous angular lines within them.
I used the Canon D30, since it has a similar resolution, and I was unable to see anything drastically difference between them in terms of this particular defect. I was somewhat surprised by the red/green color "fringing" on the Sigma, however, but that may have more to do with the lens, or even the software. If you can show me an A:B comparison of a camera of similar resolution that shows this problem clearly, I'm all eyes. BTW, I am not stating that your are wrong, you may well be correct. It is just that without having a comparison of another image technology to compare it to, I cannot identify that the probelm is unique to the Foveon chip. Art David J. Littleboy wrote: > "Arthur Entlich" <[EMAIL PROTECTED]> wrote: > >>The Foveon technology looks interesting, but their current implementation >>doesn't have any diffuser over the sensor, which makes it appear sharper >>than most competitive sensors, but it is prone to aliasing and moire. > > > I've only seen samples on the web (some full file size) but I seem to > recall seeing more aliasing with the CCD/bayer interpolation pattern > than with the Foveon. > > > http://www.steves-digicams.com/2002_reviews/sd9/samples/IMG00128n.jpg > > Look closely at just about any close to horizontal or vertical line, e.g. > the white border around the ONE WAY sign. (Diagonals seem to survive (see > the finer wires) but the fat wires are a jaggy mess (those are jaggies, not > twisted pair cables).) At first glance there seems to be a lot of detail, > but it's all aliasing artifacts. > > David J. Littleboy > [EMAIL PROTECTED] > Tokyo, Japan > ---------------------------------------------------------------------------------------- Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
