Jan, not that any of this is relevant to "wowhorse" who started the thread with an apparently innocent question about whether to buy a Rebel or a D70,
Of course it isn't.
"Never go off on tangents, which are lines that intersect a curve at only one point and were discovered by Euclid, who lived in the 6th century, which was an era dominated by the Goths, who lived in what we now know as Poland."
- Unknown from Nov. 1998 issue of Infosystems Executive.
but just for the record, there is about 37% increase of resolution (measured by pixel dimensions) between an 8MP 20D and an 11MP 1Ds.
No, that's "37% increase of total pixel count".
"Total pixel count" and "resolution" are two different things, it makes no sense to consider a camera with twice the number of megapixels as twice the resolution, unless those megapixels are distributed differently.
That's why I was so careful to write angular resolution
in the context of lenses; horizontally, the 1Ds sensor has effectively 4064/35.8 = 113.5 pixels/mm, the 20D 3504/22.5 = 155.7, which is what counts when you use the same lens.
And if you choose to change the focal length for the lens on the 1Ds in order to achieve the same FOV as the 20D would have, the resolution advantage to the 1Ds is 4064/3504 - 1 = 16% horizontally.
Similarly, the advantage for the 1Ds MkII at the same FOV is 42.5% horizontally, but the sensor has "only" 138.7px/mm of effective resolution.
This is what the Nikon folks are talking about when they say that the D2X (180.9px/mm) has greater resolution than the 1DsMkII. And of course they're right, in the sense that the D2X ought to be noticeably more demanding of its lenses.
I'm a bit stunned to see that this hasn't caught on, even though the same arguments were used in the "Digital vs. film" debate, just with digital vs. film instead of digital vs. digital.
As both sensors are 12 bit depth and both have ample resolution at least to size A3, relative to the 1Ds there would have to be really major advances in the 20D's sensor design and software for there to be that much difference in rendering of shadow detail between them.
I was actually thinking of the other way around; that the 1Ds probably doesn't show much better shadow detail than the 20D. But that may also depend on whether we look at the same print size for the same FOV or not.
Once we're dealing with cameras in the range of the 20D - 1Ds, I think shadow detail depends alot on exposure and post-processing strategies rather than fine differences in the hardware.
Now that is something I think is easy to agree on, unless you're experiencing problems like those pom has.
And since you don't seem to be experiencing the same kind of problems, my guess is that he has been unlucky.
FWIW, there are circumstances when banding shows up at ISO 100 on a clear, blue sky even with my 20D, but those are rare, and difficult to reproduce. I'm guessing at electromagnetic disturbance from something, but what?
I've long been thinking that modern DSLRs should incorporate a Faraday cage to reduce the risk of electromagnetic noise from external sources, and maybe they already do. Of course, there's a possible problem with that big hole in the front of the camera if you have to block GHz frequencies. ::