Hi Truman, > The others also have 1/3 of the > resolution that they claim. That's not true. They have the FULL spatial resolution they claim, it's the color fidelity that is decreased. It's a matter of number of photosites, not pixels or M bytes. > You cannot invent additional sensors than are not there. Well, actually you can. That's what interpolation is...but it wouldn't work with the Foveon in the same way as it does with the Bayer pattern imaging sensors, simply because the Bayer pattern imaging sensors have every output pixel based on some actual data from that photosite. Interpolating a Foveon would be the same as doing it in PhotoShop...but the Bayer pattern interpolation for color information is entirely different. > Of course if you step a red then blue then green laser across a > conventional sensor you will get 1/3 the number of sensors advertised. Actually, 1/2 for green. The Bayer pattern imaging sensors have %25 red, %25 blue and %50 green. But, the fact is, not much exists in real life that is %100 red, %100 blue or %100 green, and this is why the Bayer pattern imaging sensors give really good images, and why a Foveon, even of the exact same spatial resolution size, doesn't give much of an increase in color fidelity over a Bayer pattern imaging sensor of the same spatial resolution. People are confusing two resolutions. One is spatial, the other is color (tonality). Spatial is purely based on the number of photosites. Color is based on the depth of information received at any one photosite. For the Foveon, it gets full RGB at each photosite, and for the Bayer, it gets either R, G or B at each photosite. To relate this to your computer monitor. You can have a 1200 x 1600 monitor with 8 bits/pixel (256 colors), and the same monitor can be 1200 x 1600 with 24 bits/pixel...yet they are the exact same spatial resolution. A 768 x 1024 monitor with 24 bits/pixel does not have the same spatial resolution as a 1200 x 1600 monitor with 8 bits/pixel, though it has higher pixel depth...but pixel depth is irrelevant as far as spatial resolution goes. In fact, TV works a similar way to the Bayer sensor concept, where the color information (Chrominance) is only 1/4th as much "data" as the luminance information. This works because of how our (human) visual acquisition/processing system works. We are more acute to edge/luminance difference than to color differences. This is why there is such a marginal (if any) true visual benefit to having full color information, like the Foveon gives. Austin
Message
RE: [Digital BW] Re: OT Canon 10D
2003-02-26 by Austin Franklin
Attachments
- No local attachments were found for this message.