For those arguing 5Dvs 40D ISO issues...

I just took a few pictures with the two bodies. Nothing high-tech here.

A single light fixture, above and to the left of the frame.

70-200 2.8L IS on a tripod with the IS disabled. Roughly 110mm for the 40D, 173mm for the 5D.

Shots taken using the timer.

Camera set in Av mode.

No control modifications

AWB.

5D set to Faithful, 40D set to Neutral (remebered after the fact)
The 5D was also set for center point focus (didn't realize until I looked at the focus points).

Shots taken at 100, 400, 800, 1600, and 3200

There are 9 images in the directory, 4 from the 40D, 5 from the 5D. (Forgot to turn ISO expansion on on the 40D) These are the CR2 files. They can all be opened in ZoomBrowserEX 5.8.

As you can tell from my sig, I either have no horse in this race or a lot of ponies, depnding upon your point of view

The files can be downloaded from http://www.chernicoff.​com/pictures/ISOTest/

Draw your own conclusions and post here, if you download the files; it's clearly marked in the file names as to what they are...

Now that I'm looking at he images I realize that the focus point issue does have a significant impact on the results. I might have to redo this later tonight.

Thanks! Appreciate the effort. I especially like that you used the same lens for both and that you provided the RAW images to download and play with.

I had to make a few adjustments to equalize the images - set them both at "neutral" picture style, and "tungsten" lighting.

It appears that the 40D carries on the new trend in Canon cameras of being 1/3 stop less sensitive. The earlier cameras (5D, 30D, 1D2n, etc) were about 1/3 stop more sensitive than their ISO setting measured, with ISO 1600 being equal to a measured ISO 2000. Canon has apparently corrected that with the 400D, 40D, and 1D3 (not 100% sure on the last one). Now, an indicated ISO 1600 is about equal to a measured ISO 1600.

Anyway, when I bumped up the ISO 1600 shot of the 40D by 1/3 stop to match (by eye) the exposure level of the 5D, it did exhibit more noise. However, setting DPR's noise filter at "low" for both chrominance and luminance noise cut the difference down a bit.

There seems to be a bit of a depth-of-field and focus plane difference - the 5D appears to have been focused on the middle box (fishing line) while the 40D's focus plane falls on the front of the Fruit2O bottle.

Tom W wrote in post #3901970
There seems to be a bit of a depth-of-field and focus plane difference - the 5D appears to have been focused on the middle box (fishing line) while the 40D's focus plane falls on the front of the Fruit2O bottle.

If you look at them in ZoomBrowser and check the focus points, you'll see that is exactly the case. One of the reasons I should probably redo this.