Originally posted by FantasticMrFox You can't just say one ISO value is equivalent to another ISO value in a different format.
I just did. And I know what I am talking about. Although, to avoid another debate, I won't discuss. Please, google it up. Thanks.
Originally posted by Simen1 I think you disagreed on something we actually agree on. I just formulated it different. ... Thats why I wrote equal ISO number means apples to apples. If I choose your approach, small sensors (phones/compacts) would have a very high real base ISO.
No, we disagree and we do because you didn't do your homework.
Actually, small sensors *DO* have a very high
35mm-equivalent base ISO although their
base ISO is low. That's not a question of approach, that's how the terms are defined. However, it doesn't affect comparisons as you compare at the lowest available value of ISO anyway. It just explains why the results are so different. BTW, the results for an apples to apples comparison.
Base ISO isn't the same for all sensors, regardless of size. 645Z and D800 roughly have the same full well capacity per unit area and the same base ISO, correct. But the D810 (and here you missed to do your homework) has a different sensor significantly more expensive to build and not used in any other camera. It has a larger full well capacity per unit area which allows it to have base ISO of 64 (50 actually if calibrated the way DxO does). To estimate the effect, I computed the 645Z 35mm-equivalent base ISO to compare the effect of the two more expensive sensor techs: size vs. area capacity. That's an apples to apples comparison, your's was not.
Medium format has a special look (always had) because it captures more light before the medium saturates, film or sensor. Which leads to cleaner images, not only more detailed. In the digital age, this effect is expressed as 35mm-equivalent base ISO. And here, the D810 compares favourably because it has a different sensor tech effectively simulating this medium format effect, at least for cropped MF.