The "measured ISO" of the GH3: Surprise, surprise

The ISO standard forces manufacturers to make sure that a certain exposure under certain light conditions at a certain ISO setting results in a certain brightness in the out-of-camera jpegs. However, manufacturers are free to decide how the brightness of the OOC jpegs relates to the saturation level in the RAW files. The fact that the OOC jpegs clip at a certain point need not imply that the RAW data clip at the same point. And for those who shoot RAW, the saturation level of the RAW data rather than the brightness of the OOC jpegs is all that really matters.

How RAW level saturation relates to camera ISO is regularly measured by DxOMark, and it is already well known that Olympus and Panasonic have chosen to calibrate their cameras differently in this regard. Whereas the camera ISO settings on Panasonic bodies usually correspond quite well to the measure of RAW file saturation that DxO calls "measured ISO", those on Olympus bodies tend to fall significantly above "measured ISO". The most extreme example of this so far is the E-M5, where camera ISO is nearly twice as high as "measured ISO". For example, at a camera ISO of 200, the "measured ISO" of the E-M5 is merely 107.

I think everyone (me included) expected this difference between the two manufacturers to stay the same with the introduction of the GH3 so that ISO 200 would in this case correspond to a "measured ISO" of approximately 200. As far as I can tell, however, such is not the case. On the contrary, the "measured ISOs" of the GH3 appear to be just as far below the camera ISOs as on the E-M5.

What gives me reason to think so? Well, I have recently had the opportunity to check-out the RAW files from some side-by-side shots with the E-M5 and the GH3, where both cameras were used to capture the same scene at the same camera ISO (ISO 200) under the same lighting conditions using the same shutter speed and the same lens at the same f-stop. And -- surprise, surprise -- the saturation level in the GH3 RAWs is no higher than in the RAWs from the E-M5.

I cannot make the RAWs on which I base these preliminary conclusion available to others since they are not mine to share. I have received them via private channels from another forum member. For this as well as other reasons, I encourage those who have the opportunity to compare the GH3 to some MFT camera already tested by DxO (it doesn't matter which) to repeat the experiment and report their findings in this thread. There is always the possibility of errors somewhere in the "research process" and since the results are unexpected, there are even stronger reasons to verify them than would otherwise be the case.

If someone is willing to do the test but need help or advice on how to do it, just let me know.