Following this explanation which makes sense to me based on my experience, what would be the optimal design iso performance for the 60Mp backs, which start at 50? I can see the push to 100 as I do it often, but by 200, in my images where you still have good highlight protection, I feel the images really suffer in the shadows, i.e. excessive noise. So I still tend to bracket most series knowing that one of the shots will be the correct one. But to me the other side of the equation is good light for the CCD, i.e correct exposure and this becomes even more critical on shifts/movements. The difference in say 1/60 and 1/125 @ F11 on a 15mm shift can be huge in the difference in details that are captured, and not lost to noise. This is something I see every time I use a Phase back. The image must have the correct or near correct exposure for the critical parts of the image, where as with a CMOS capture I have the confidence I can expose for the highlights and pull up the shadows and still have details in them, with only 1 series of exposures.
The information in the later part of this post has been very informative.
Thanks to all.
Paul
Paul,
The lower you rate the back, the better, although of course you really need to get the highlights right. The reason is that with a shift lens the sin^4 law (vignetting) lowers the exposure more and more in the far edge, and even if one *can* push, one is then eating into the shadow latitude. However, leading edge is exposed brighter, until you've shifted completely past the center, eg. using a 4x5 lens on a 36x38 back to make a pano.
The measured T stop of the lens depends on the measurement position on the sensor, and the degree of shift.
If you really shift all the way, you need to lower the assumed ISO of the back more and more, because now the lightmeter ISO has nothing to do with the back ISO due to the shift. It's not really the back ISO which is changing, in fact it is the real T-stop of the shifted lens, but it's simpler, I guess to think in terms of ISO.
I'm not the sharpest knife in the box here, as we all know, but Erik or Bart will do the sums for you. Anyway, at 5 am. I don't expect much of myself in terms of making sense
Edmund