EOS 5D and 24-105 f4L IS USM
EOS 20D and 17-85 f4-5.6 IS USM
PART 2 - More first shots (24 September 2005)
At their widest setting, there was a clear victory for the 5D combo. What about at maximum zoom?
In this difficult lighting, both camera/lens combinations shooting wide open (f4 on the 24-105, f5.6 on the 17-85) show some vignetting, again the EOS 5D and its full frame lens being slightly worse in this regard. Nothing that couldn't be fixed quite simply in Photoshop.
The samples above are 100% crops. The 5D crop has more pixels, but doesn't appear to show any more detail. Perhaps the 17-85 lens is sharper at 68mm (not its maximum zoom) than the 24-105L is at 105 (maximum)?
Here I have upsampled the 20D crop to match the size of the 5D extract. I cannot see any extra detail in the 5D crop. For this situation, the 20D camera/lens combination appears to equal the 5D, although there does appear to be slightly more noise in the 20D result, as might be expected.
Since taking these shots, I have repeated this test in different lighting conditions and with different subjects, and the results are the same. This is not so much a "dead heat" as an actual victory for the 20D/17-85 combination. Whilst the 24-105 was at maximum zoom, the 17-85 still had more to come, as maximum zoom with that lens is equivalent to 135mm on a full frame camera.