DxO has completed their review of the Canon EOS R6, and found the sensor to be quite capable, even if it doesn't move the needle a ton. DxO gave the lower megapixel brother of the Canon EOS R5 a sensor score of 90, which puts it a point behind the Canon EOS 5D Mark IV and 5 points behind the Canon EOS R5.
The sensor for the Canon EOS R6 is apparently built on the sensor from the Canon EOS-1D X Mark III, and with the latter scoring a 91 from DxO, it appears that they're closely related.
From DxO:
While the 20 MP sensor may seem an odd choice, this is a variant of the one found in the flagship EOS-1D X Mark III, and the results are practically identical. It performs well across the board with excellent dynamic range at all ISOs. Add to that the sensor’s good color and low noise, and the Canon EOS R6 looks set to be a firm favorite among a wide range of photographers who work across many different genres. Read the full review
Only thing I'd like to see improved is sharpness. While it is adequately sharp, I wonder if the sensor would have scored higher if the anti-aliasing filter had not been as strong? I'm not comparing its sharpness to the R5, but the R. Close, but not quite. More like the 5DIII--which was quite good! I'm just balancing my praise here.
But, really, for events and outdoor action, really great. That ability to recover shadows and highlights is so welcome with wide-angle lenses during the day.
I think their scores are pretty heavily affected by resolution, which would seem to imply that it scored better in other categories (as you suggest with your real-life results) than the 30mp sensor in the R and 5DIV. And I'm sure there are some improvements that could be made with filtering, but overall it looks like a really good overall option. Even Jared Polin recently rated it favorably to the Z6 II and the Sony A7III, and that guy doesn't like anything!
-Brian
we should remember what they did with the camera they once came out with for the iPhone a few years ago. While they said that all their RAW tests just included the straight RAW, they actually used the feature in the camera to shoot several images, and combined them for the ratings, and then gave themselves a very high rating for the sensor, without also shooting one image and rating that as well. They were heavily criticized for that one.
And nobody really takes what they say seriously except for consumers who don’t understand any of this.
Their ratings are based solely on sensor performance from what I understand. How many people, I wonder, go to DxO and sort by best score and just get whatever is on the top? For me, a camera is so much more than how much sheer image quality it can produce. I'm still a novice in photography without a lot of experience under my belt, so I had the option of going with any of the big 3 manufacturers (just to be clear that this isn't a cope post). I decided on Canon (Nikon was a very close second), after going through a couple cheap bodies from other brands, despite the fact that Canon's sensors don't usually get as high marks as Sony or Nikon. The deciding factors came down to things like how the camera feels in my hand, the intuitiveness of the menus, the aesthetics of the camera, how the shutter sounds, and other things that some would consider to be superficial. I then considered the technical features, like card slots, IBIS, drive speed etc... Now that I think about it, I didn't even really consider image quality hardly at all. I figured any full frame camera made in 2020 is going to have really good IQ. The point of my rant here is that photography is not a technical profession like engineering or computer science. It's art. And if I have a camera in my hand that feels good to use and is fun to shoot with, I'm going to get so many more shots with it than with a camera that technically scored a few points higher on an epeen site, but is a chore to shoot with (ahem, Sony).