laukkart

@ecka84 I believe you can choose the ISO you like in that tool for your own comparisons.

This very particular site gives you guys an opportunity to compare cameras side by side on the studio comparison tool. I am pretty sure it removes most if not all the needs for arguments in this particular case:

"Without NR" (or RAW):

https://www.dpreview.com/reviews/image-comparison?attr18=lowlight&attr13_0=fujifilm_gfx100&attr13_1=fujifilm_gfx50r&attr13_2=sony_a7riv&attr13_3=fujifilm_xt3&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=25600&attr16_1=25600&attr16_2=25600&attr16_3=25600&normalization=print&widget=1&x=-0.8500366568914957&y=-0.17981773947939447

With NR (JPEG):
https://www.dpreview.com/reviews/image-comparison?attr18=lowlight&attr13_0=fujifilm_gfx100&attr13_1=fujifilm_gfx50r&attr13_2=sony_a7riv&attr13_3=fujifilm_xt3&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=25600&attr16_1=25600&attr16_2=25600&attr16_3=25600&normalization=print&widget=1&x=-0.8500366568914957&y=-0.17981773947939447

I see your comments around tell me more what this has to do with. Enjoy, and have a nice day.

I think you mean steps above typical DPreview effort on sample images :).

There's quite a big difference between Phocus and ACR edited shots, especially in tones and color contrast.

Indeed they do. But frankly, that does not improve the position of this test at all, rather opposite.

It kinda feels weird to defend importance of RAW and postprocessing here - on a "photography" site.

The problem of this article is that it hides the actual intention. The intention (I guess) was to show that one can get almost equal IQ out of Z7, when certain conditions are applied (closed down aperture, matched light gathering, non-edited RAW exported to JPEG).

Anyone who has actually shot these systems, and/or edited the RAW files, knows that the this IQ comparison does not held so well anymore, when other shooting and postprocessing conditions are applied, and the MF will come more clearly ahead.

I can also get almost equal IQ out of 24mp APS-C vs. 24mp FF, when certain conditions are selected. Once again, apply other conditions, and the 24mp FF will come clearly ahead.

Comparison of JPEGs in general is pointless. They do not show the differences, as they are compressed with limited DR and 8-bit colors. The ACR will only "normalize" colors, sharpening, NR, and the JPEG engine. More or less, presenting the images as JPEGs more like hides the potential differences.

You can repeat this test of "no visible difference" to great degree on pretty much any system with nearly same resolution, as long as the lens is decent enough. You can observe this from the 2nd link, which compares 20mpix MFT to APS-C and 24mp APS-C to FF. Feel free to change there JPEG to RAW if you like.

photofan1986 You may have noticed that this article uses JPEG files to compare 35mm to MF.

One can see from the links I have provided how not so good of an idea that is.

You are correct that the situation improves on non NR'd RAWs, especially A7r3 no longer looks so dominating over the Z7: https://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13_0=fujifilm_gfx50s&attr13_1=nikon_z7&attr13_2=sony_a7iii&attr13_3=sony_a7riii&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=6400&attr16_1=6400&attr16_2=6400&attr16_3=6400&attr126_3=1&normalization=full&widget=1&x=-0.18377937578767883&y=-0.5918534795860941

I'd wish Dpreview would provide RAW comparison with standard ACR NR applied as well, as high ISO RAW comparison is quite useless otherwise.

Then there is the way the RAWs behave in postprocessing that these tests cannot show. For that, one must download the RAWs and test themselves. Highly recommended to those who want to study actual differences - especially for MFT - APS-C - FF - MF comparison.

And in certain different use cases, 3:2 and 4:3 aspect ratios will have significant difference.

I'll just leave these two other test scenario links here, courtesy of this very site:

https://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13_0=fujifilm_gfx50s&attr13_1=nikon_z7&attr13_2=sony_a7iii&attr13_3=sony_a7riii&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=6400&attr16_1=6400&attr16_2=6400&attr16_3=6400&normalization=full&widget=1&x=-0.18377937578767883&y=-0.5918534795860941

You can even drop the Z7 to ISO3200, if you want to "compensate" the ISO64 advantage.

https://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13_0=sony_a6500&attr13_1=sony_a7iii&attr13_2=olympus_em1x&attr13_3=nikon_d7200&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=100&attr16_1=100&attr16_2=200&attr16_3=100&attr126_2=1&attr171_2=1&normalization=full&widget=1&x=-0.19826542960871318&y=-0.5975918481638142

Do your own conclusions.

@BlueBomberTurbo

Yes, it is true. Do comparison with ISO200 1/10 0EV for X-T2 and ISO100 1/10 +1EV for Sony, as they are shot at same shutter speeds and apertures. From there you can easily observe that they indeed are the same shot if you push Sony or pull Fuji about 1EV. That is, shoot Fuji at ISO200, and you will have enough highlight headroom to pull it to match Sony's ISO100.

If we compare Sony's ISO200 shot to Fujifilm ISO200 shot, the applied gain of Sony already messes up the comparison, and as the end result there is not that much of difference on highlights, instead, Sony has about 2/3-1 stop worse shadow performance, as Fuji is still on its 0 applied gain maximum DR.

The nonlinear highlight behavior of Sony also indicates that Sony uses non-linear analog gain, and compensates with digital gain, so even on Sony the ISO200 is not "true ISO200" when comapred to its ISO100. This is actually backed by photonstophotos measurement, as ISO200 has only 0.7 EVs less DR than ISO100

Plenty of misinformation here.

A digital sensor has no ISO. It has an amplifier gain. At "base ISO", the gain will be 0dB to give maximum dynamic range. Then the manufacturer approximates the brightness of JPEG to correspond to that "base ISO" (there are couple different standards that can be applied for this). The end result JPEG can have applied digital gain and/or contrast curve that affect the end result brightness.

What Fuji does compared to Sony is that they use different saturation (brightness) point of that same sensor. There's nothing wrong in it, it only moves headroom. That is, on Sony cameras you have very little headroom for highlights, where as on Fuji cameras have about 1 stop of headroom for highs - Fuji tolerates overexposure better and underexposure worse than Sony and viseversa.

This means that using same shutter speed and aperture at base ISO you either need pull Fuji X-T2 ISO200 by 1 stop, or push Sony A6500 ISO100 by 1 stop, this will result to same end result

We can do that here - I have very hard times to see this "step backward at high ISO". And yes, I have set the settings so that they are more realistic comparison than when comparing 1:1 pixel ratio - these are low light shot, NR applied (jpeg), and normalized resolution to something people would actually use (print): https://www.dpreview.com/reviews/image-comparison?attr18=lowlight&attr13_0=fujifilm_xt3&attr13_1=fujifilm_xh1&attr13_2=nikon_d500&attr13_3=sony_a6300&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=12800&attr16_1=12800&attr16_2=12800&attr16_3=12800&attr171_3=1&normalization=print&widget=1&x=-0.17913393087676593&y=-0.6599642542234104

This type of screen is the best for photographers.