Sigma/Foveon Questions
in [Digital Cameras]
|
Prev: SI Facescape
Next: FF camera with mirrorless design
From: nospam on 8 May 2010 13:28 In article <cd394f93-1a29-419a-a191-daa56c3709cb(a)i9g2000vbh.googlegroups.com>, Bubba <digitalrube(a)yahoo.com> wrote: > Do manufacturers tell you if a sensor is back-illuminated? more than likely, since they generally want to differentiate it from everything else > For example, is the Canon G11 back-illuminated? no
From: dmaster on 12 May 2010 13:58 On Apr 29, 1:08 am, Alfred Molon <alfred_mo...(a)yahoo.com> wrote: > In article <280420101626373798%nos...(a)nospam.invalid>, nospam says... > > > > > > It's three separate spectral measurements per pixel, while a Bayer > > > sensor has only one. > > > and bayer calculates the other two. at the end of the day, both produce > > rgb, and oddly enough, bayer is more accurate. > > Obviously Bayer is not more accurate. And how could it be, with just 1/3 > of the needed information? > > > > Also, Bayer does not measure luminance at the pixel level, while a full > > > colour sensor does. > > > neither measures full luminance, however, both calculate it. > > A full colour sensor measures luminance at the pixel level. Bayer does > not. > > > > In any case we were talking about *spatial* interpolation, which Bayer > > > does to generate the final image. > > > nope. it fills in the holes. > > With guesses, which often are wrong. > -- > > Alfred Molon > ------------------------------ > Olympus E-series DSLRs and micro 4/3 forum athttp://tech.groups.yahoo.com/group/MyOlympus/http://myolympus.org/photo sharing site This entire arguement is silly on several levels. 1. You reference "full colour sensors" (previously you refered to "true samples"), but you overlook that neither Bayer sensors nor Foveon sensors provide either. Neither is "full colour" at each sensor; neither is even RGB at each sensor, although it appears that you think "RGB" is "full colour" or a "true sample". RGB is an imperfect representation of the true colour, and even if Foveon sensors produced RGB, they wouldn't be "true colors". No sample is "truer" than any other sample; they are samples. 2. Most of the arguements in this chain seem to be centered on whether or not Bayer sensor results are interolated, as if that is somehow good or bad. It is neither; it is an approach. Both Bayer and Foveon sensors are imperfect approaches to sampling the actual scene. There is no point in arguing whether the approaches are "good" or "bad". The only issue is whether the results are good. As millions or billions of examples have shown, Bayer sensors produce good results. If you think the Foveon sensor camera produces a better result for the money, buy it. If you think the Bayer sensor camera produces a better result for the money, buy it. Personally, I prefer the results from the Bayer sensor cameras for my money. If you disagree, so what? I'm happy with my choice; be happy with yours. Dan (Woj...)
From: dmaster on 12 May 2010 14:03 On Apr 30, 7:32 am, Alfred Molon <alfred_mo...(a)yahoo.com> wrote: .... > But the question is, what the performance of the sensor is. It may well > be the case that the lens sucks and the AA filter additionally > contributes to blur everything, but this is a separate issue. > > The key question is, can the sensor resolve pixel-level changes of > colour. A Bayer sensor *cannot*, while a Foveon sensor or other full- > colour imaging device which does not have to interpolate the data from > adjacent pixels can. .... > Alfred Molon > ------------------------------ .... No it is not. The only key question is, does the sensor produce a good picture. The mechanism or approach is completely immaterial. If the sensor uses single color sensors, or non-RGB multi-color sensors, or tiny pixies with tiny arc-welders does not matter in the least. Does it make a good picture? Dan (Woj...)
From: Alfred Molon on 13 May 2010 02:33 In article <ce4842f8-d6d9-49c9-a815-6c26c68fce48 @k2g2000yqe.googlegroups.com>, dmaster says... > The only key question is, does the sensor produce a > good picture. The question is the effective resolution. And here a sensor like Bayer, in which to reconstruct an image you need to use data from neighbouring pixels, fails at the pixel level. Bayer simply does not have the resolution at the pixel level, i.e. a 12MP Bayer sensor has an effective resolution which is lower. Bayer produces good pictures, but the effective resolution is lower than the nominal. -- Alfred Molon ------------------------------ Olympus E-series DSLRs and micro 4/3 forum at http://tech.groups.yahoo.com/group/MyOlympus/ http://myolympus.org/ photo sharing site
From: Alfred Molon on 13 May 2010 02:35
In article <add80f9f-11e6-488c-b5af- 3b959023b44c(a)k19g2000yqm.googlegroups.com>, dmaster says... > 1. You reference "full colour sensors" (previously you refered to > "true samples"), but you overlook that neither Bayer sensors nor > Foveon sensors provide either. Neither is "full colour" at each > sensor; neither is even RGB at each sensor, although it appears that > you think "RGB" is "full colour" or a "true sample". RGB is an > imperfect representation of the true colour, and even if Foveon > sensors produced RGB, they wouldn't be "true colors". No sample is > "truer" than any other sample; they are samples. The key point is that a sensor like Foveon does not rely on interpolation from neighbouring pixels to reconstruct the image. That is a very, very essential difference. -- Alfred Molon ------------------------------ Olympus E-series DSLRs and micro 4/3 forum at http://tech.groups.yahoo.com/group/MyOlympus/ http://myolympus.org/ photo sharing site |