What real Canon ISO?

[snoke]

Recent review talk about it ISO problem for 6D Mark II.

Images from DxO show it. Not PRNG scores. Measurement.

What this mean?

If math say shot need 1/100, ISO 100, f/5.6, 6D Mark II need 1/60 exposure correct. Only long time camera series not issue is 7D.

If someone have 6D and 6D Mark II, same settings (use manual - ignore EV meter!) darker on 6D Mark II. Anyone? Experiment?

 

[ejenner]

I noticed something like this when I went from T1i to 5DII. I used to compare sensors, but since upgrading to a 5DIII, I have found it is 'good' enough' for me. Whether a sensor is 1/3 or 2/3 stop better in noise is not important to me anymore.

But yes, I think you have to be careful when comparing ISO's of sensors because I noticed up to 2/3 stop variation. You need to set to manual and compare with the exact same settings, then adjust in post if necessary (I would say adjust ISO, but with the non 1D cameras having native ISOs at about 1 stop intervals, that isn't so practical).

I do think this is a 'trick' canon has played over the years to try to show improvement in sensors when none really existed. Of course it is all relative, there is no 'true' ISO. Film may have had a rating, but there were definitely variations.

 

[Mt Spokane Photography]

DXO can confuse people who read their postings, and those who can understand what they are saying know its often a red herring.

Camera ISO settings are plenty accurate, no matter what brand. DXO is off into theory, and you have to read their fine print. A camera, even with manual settings can only set exposures to 1/3 stop, so adding in camera variation, a better than 1/2 stop accuracy is hopeful. Its the nature of the beast, nothing is perfectly accurate.

They are apparently unable to actually measure sensor full well electron count, but have devised a home brew method of attempting to measure it, and claim 1% accuracy. That's a pretty doubtful claim.

It has nothing to do with the camera giving a proper exposure, or at best its sort of related. Note that they never actually test a camera exposure, just a theoretical measurement that does not take into account any internal camera calibration of shutter speeds.

https://www.dxomark.com/dxomark-camera-sensor-testing-protocol-and-scores/