I am contemplating getting the R6II alongside my R7. I mainly do wildlife, birds, and macro, where I really want lots of pixels on subject, and the R7 is great. But in low light it is deficient, and being a crop sensor, doesn't have as nice of bokeh.
Both those complaints should be irrelevant in the type of work you say you are doing. Do you really think that the R7 sensor is deficient when focal-length-limited, or when stopping down for needed DOF? The R7 has more resolution and less noise than the R3 in those situations. Perhaps you are obsessing on 100% views of sharpened pixels?
I was walking around the zoo. There were indoor things (a platypus, some insects, some fish, etc) where flash was not allowed and there was poor light. Something with better low light performance would have been nicer here.
Probably, but that would really require a bigger lens pupil, with or without a larger sensor or larger pixels. If you actually needed DOF and had to stop down, then your photon noise is due to your DOF; not the sensor (for sensors with similar QE).
The R7 at ISO 6400 (I was shooting wide open at f/2.8) isn't pretty even with good postprocessing.
It's a matter of scale. Yes, if you made a large print of the same composition at the same ISO, you would have a less noisy image with the R6. If you ever thought that I would doubt that, then you clearly have not understood anything that I have said.
However, there is ZERO reason to assume the same ISO, as if the ISO were dictated directly by the lighting environment. The ISOs that you get will depend on the lenses available, and the shutter speeds chosen.
Part of that is because converters assume that most people are looking for pixel-level detail, all the time, even when there isn't sufficient light to magnify them cleanly, so converters sharpen quite a bit at the pixel level with the R7, sharpening a fine level of noise and detail that can't even be recorded at all with R6-size pixels, creating a deep dither effect with more salt and pepper pixels. This is done by software; not the camera!
I do mainly those types of photography but I do the occasional event, artistic, portrait, etc where the R6II would produce better images.
There is no reason to assume that the R6II has less visible noise per unit of sensor area than the R7, except that the sensor design is "newer", but the R3 is newer and has a little bit more visible noise per unit of sensor area than the R7 or R6.
By the way, the differences between these cameras in low light are not just "obsessive pixel peeping":
They are. The reason most people don't compare equitably is because they want to see everything the system captured, and not magnify the image any more than necessary to have each pixel represented on the screen (because doing so makes the view softer and nosier), and that is 100% or 1:1 pixel views, but that is a different magnification for sensors with different pixel densities so people are asking for a lot more from the R7 than from the R6 at 100%; not less, and not even the same, so there is more room for apparent failure. A better way to compare would be to look at the images at the same pixel ratio (both 100%, or both 200%) on separate monitors with the same proportion of pixel density as the sensors, and not sharpen in the conversion.
So the R6II would fill in the gaps. If I had to choose one it would be the R7. You have a similar choice to make.
Optically and noise-wise, a 24MP FF is only going to be superior in IQ when the FF has a larger-pupil lens or you get closer, and you must get shallower DOF with it, or you won't get more light.
The pixels on the R6II are about 4x the size of the pixels on the R7, leading to about two stops better noise performance at the same exposure settings.
No; that's not "noise performance"; it is pixel noise performance, and completely irrelevant when a different number of pixels are used to form a subject.
So I could have shot at ISO 12800 at twice the shutter speed and ended up with a nicer image, in the situation above.
Yes, you can get a less noisy image if you brought along a lens with a larger pupil at the needed angle of view (and you actually want the shallower DOF, or are willing to accept it to get less noise), but that is true for both cameras. A sensor without a lens can not do photography. The lens is what forms the image; the sensor simply decides what fraction of the image circle is captured, and finely it is divvied up.
There are certainly situations where FF can get you more subject/composition photons (usually fast, short focal-length primes), but for others, big pixels and big sensors have no noise benefit for actual photographic needs when shutter speeds needed do not allow base ISO. In any given light, with any given subject and shutter speed, the amount of photons captured from the subject depends only on pupil size and distance; it has nothing whatsoever to do with pixel sizes or sensor sizes. If one's evaluations go against that simple rule, then it might be an indicator that the evaluation occurred in a state of illusion. Yes, there can be differences between cameras in quantum efficiency and added electronic readout noise, but as I demonstrated in the Studio Comparison Tool link in another post, the net effect of QE and readout noise is the same for both the R7 and R6, with the same amount of photons.
