About the OM-D E-M10 Mark III

otto k wrote:

Let's agree to disagree. What I see in your photos is that nx photo is brighter (just take a look at the white square) and less saturated (this both could be due to raw developer, profile, curves, etc). Also, take a look at Nikon blue green red squares - blue is bleeding all over the place, red is just a noisy mess. Bottom right, on nx I can clearly see difference between the frame and black and gray squares, on Nikon I can barely see where fields end.

Look, dxo is great for many things but I have learned that I take their simple numbers with a huge grain of salt. For example, their numbers say my nx1000 is around half a stop behind my NX500, but many thousands of photos clearly demonstrate to me that there's a rough comparison ISO 2000 to 6400 from nx1000 to NX500 (so, over a stop). It's not just the amount of noise, it's how it looks to me (others will see it differently) with nx1000 being splotchy and ugly. I also had D300 for almost 8 years and, IMHO, it's at least two stops behind in my practical usage.

But, ali this has next to nothing to do with the topic, so...

I redid the post-processing.

The D200 underexposes about 1/3 stop relative to the NX500, so I bumped the brightness by 1/3 stop.

I ran both the RAW files through Nik Dfine using custom profiles, and the results are much better in terms of noise removal.

The NX500 still has slightly better resolution (which one would expect given the megapixels and the lack of an AA filter on the NX500). But the degradation in contrast, colour saturation and colour fidelity on the NX500 are also visibly greater, and it's not fixable, because the bit depth and DR are not there.

It's basically a question of preservation of detail vs. what the photo looks like at normal viewing distances.

Nikon D200 at ISO1600

Samsung NX500 at ISO 6400

Also of note is that I used ordinary 2700K household LED lighting, which is far from ideal photographically but the point of the exercise was to simulate real-world conditions. The JPEG engine on the D200 struggles with fluorescent and LED lighting, which doesn't show up on the test because I used RAW files and set the white balance in post-processing.