It’s exactly the same sensor and processor as an R6ii. You won’t be disappointed with the dynamic range, despite what any reviews or analysis suggest. Remember that no display can reproduce much more than 8-10 bits of DR, prints even less. And importantly it retains excellent DR at high ISOs. Don’t fret about the math.
Photons to Photos has their own way of doing it. The dynamic range is calculated to omit most of the read noise from the specified noise floor. Click on the key on this page, and you will see a table showing a dynamic range of up to 13.9 stops.
Dynamic range of sensors is not measured like the length of a wooden dowel, which has two clear ends. It is more like the useful length of a dowel that is rotting in mud on the bottom end, with no distinct bottom end, but a distinct top end (highlight clipping).
I have R8 and RP. Haven't touched the RP since. That is a great price at Canon Refurb.
I don't think any review suggests a disappointment in the R8's dynamic range. It's not the best among top end cameras but it's actually quite good by any publicly available measurement.
The DR of the media isn't relevant here because of tone mapping.
Yes, I forgot to mention that Claff's DR measurements are corrected for differences in pixel size. That correction is necessary to compare cameras with different pixel counts. I don't believe most other measurements are corrected.
If 24MP FF is what you want, the R8 is less feature-laden than Canon's other 24MP FF, but for someone using longish lenses with optical image stabilization (with EF lenses, mainly), it completely eliminates any added micro-blur due to IBIS errors, since it has no IBIS. When I bought my R5 and was using the 400/4DO IS II + the EF2xIII, I had less micro-stability than I did in e-shutter mode on the 90D in mechanical shutter/mirror mode. That's with the R5's 137x greater pixel spacing, to boot. I was very frustrated that I couldn't disable the R5 IBIS, which was only causing harm to stability with that lens and TC.
Bill has said that he doesn't directly use the normalized-MP count approach like DxOMark does, where they measure pixel noise and then do the math of what would happen if you resampled to 8MP. I don't remember the details, but I followed along with Bill's process many years ago, and it basically gave results which could have also been done with a theoretically-normalized image size (height or shorter side, actually, I believe, ignoring width, IIRC).
He normalizes for the usual depth-of-field circle of confusion, 18 micrometers or 30 micrometers.
True enough. Noise is measured as the standard deviation of the pixel values in a uniformly exposed area. The DPR test results show visual results that are helpful if one is careful to interpret the results correctly. The standard deviation does give an incomplete picture of the noise, although in recent cameras it usually agrees pretty well with the DPR results, in my opinion. There is always the caveat that the DPR results may have had noise reduction in some cases.
