Well there is the fact that your analog signal from the sensor is for one color pixel from the sensor. Then several pixels are analysed, combined according to that analysis, and written out as 4 3-color pixels. And that is just for a RAW image. It is not a simple convert one pixel to x bits operation. Is 22 bits over kill for that? I do not know.
I suspect that they are using a 22 bit ADC simply because that is what was available. But if they are eventually planing 16 bits per pixel output, a 22 bit ADC is not over kill at all. A point to consider is that with modern IC's that ADC is not just an ADC it is most likely a computer chip as well. The point I am trying to make here is I think the analysis of this that folks are making here on the list are way over simplified. -- graywolf http://www.graywolfphoto.com http://webpages.charter.net/graywolf "Idiot Proof" <==> "Expert Proof" ----------------------------------- Digital Image Studio wrote: > On 15/10/06, Thibouille <[EMAIL PROTECTED]> wrote: > >> It may be I'm not into electronics enough to argue about that. However >> the thing about the 22bit (18 or 20 being enough? how could we know) >> is indeed to have a high enough definition not to lose to much in >> calculations. > > We can't know exactly (though it's determinable using very basic > engineering calculations) as we have insufficient data however it's > clear that 18, 20 or 22 bits is overkill. In any case this is the > pre-processor that we are discussing, calculations are very > rudimentary and consist primarily of level adjustments, the output > precision to the RAW to JPG convertor is likely still only 12 bits per > colour channel. > >> Of course it is clear to me that if all the camera had to do was: >> * read the analog signal >> * convert it to digital >> * write it in a RAW file. >> >> Then 22 bit is way overkill. But I really don't think it works that >> way. Specialy if you need JPEGs (and if you don't the camera can do it >> so it is important). > > JPG doesn't figure in the RAW data path nor does RAW conversion and as > I mentioned above the RAW convertor whatever it's internal precision > will likely only have 12 bits per colour channel of data as the seed > for the demosaiced output. > >> As a stupid example (but still applicable IMO), find a old very simple >> calculator write 1 and do whatever calculation forcing in to use a lot >> of floating point numbers. Then do exactly the opposite calculation. >> Wanna bet you want get 1 as answer? The difference between 1 and what >> you got is **lost** information and you can do nothing to recover it, >> except thinking about it before and try to have as much figures pas >> the the point as possible to minimize the ammount of lost >> informations. > > No one is denying the advantage of processing precision in RAW > conversion etc however in the RAW data path only very basic level > shifting is being performed from my understanding. > -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net
