Theory only applies if the sensor technology is the same in both sensors being compared. But when you have different generations of sensors based on different technologies (CMOS vs CCD) made by different manufacturers, with different or no AA filter or microlenses, the only way to compare is with actual images, not theoretical extrapolations based on sensor area or pixel size or whatever.
[a href=\"index.php?act=findpost&pid=190359\"][{POST_SNAPBACK}][/a]
If the means to do such comparisons are available, that is, you have the equipment, then of course you will get a more precise impression using the actual equipment rather than predicting what the outcome is likely to be using other equipment. That hardly needs mentioning .
But science is about predicting what the results will be. If you have knowledge about lens performance and your your sensor's size and pixel count, you (or perhaps I should say
I) can get a pretty good idea of what to expect with the equipment you have'nt got.
I don't think that at base ISO the differences between CCD and CMOS, or the differences between slightly old technology and brand new technology, are going to be nearly as significant as sensor size, pixel count and lens quality.
One can get an fair idea of the subtle effects of removing the AA filter, from sites such as MaxMax and from the images and comments posted already on this forum.
I did some preliminary comparisons yesterday between my 40D with 50/1.4, and 5D with TSE 90/2.8. I needed to check out my 50/1.4 after receiving it back from calibration.
Sadly, from a distance of about 4 ft, my 50/1.4 simply isn't autofocussing accurately. I'll do more tests at different distances and different types of targets, but it looks as though this lens will have to go back again to Canon.
However, comparing the 40D and 50/1.4 at F1.4 with the 5D and 90/2.8 at 2.8, comparable apertures considering the 5D needed to be moved back from the tartget slightly, the results are pretty much as I predicted.
All focussing was manual, of course, and here the 40D has the edge because of LiveView. Despite this edge, the 5D/90mm combination produced clearly superior images at all equivalent apertures I tried, but the differences were greatest at equivalent apertures of F1.4 and F2.8. If I were in a studio shooting models, there's no way I would choose the 40D and 50/1.4 in place of the 5D and TSE 90.
PS.I didn't compare the 5D/TSE 90 at f16 with the 40D/50mm at f8 (that's for the next test ). However, I would predict that at these apertures the differences would be insignificant, just as I would predict that a 1Ds3 at f8 (using a good lens) would produce images on a par with a Mamiya ZD at f11 or f13.