Thanks, Richard, for sharing and the interpretation.
You do give the caveat that the measurements are affected by the conditions, including temperature. Bear in mind that the measurements are made by new purchasers who send them in and we don’t know whether they are done in an unheated outhouse in the arctic winter or in an Australian summer. I also hate seeing numbers without error bars or the means and standard errors from repeat runs. The low iso results are those that are most affected by environmental conditions as the high iso are dependent basically on photon flux.
Thanks for sharing.
Thanks, Richard, for sharing and the interpretation.
So we see that in base/low ISO the BSI sensors are noticeably better, while above that zigzag at about ISO800 everybody looks quite the same, with the A7 V slightly leading.
As I mostly shoot my wildlife at ISO 1600 and above, the R6m3 could be a useful tool instead of my R6m2, delivering same DR at higher MP count and faster readout.
Right now, I see no need to move
Hi Adam!
They all have the same DR above iso 800, which is the important consideration. For noise charts, photonstophotos explicitly states under the plot "These raw values are not appropriate for comparing camera models because they are not adjusted for gain or area." So, ignore them!
Thanks for clearing that up. So is there any way to compare those cameras at high iso? Certainly dynamic range is just a piece of the puzzle when it comes to low light capabilities.
How much difference do you realistically expect with BSI in a full frame sensor? My understanding is that the benefits of BSI are inversely related to pixel size, they are very significant for smartphone sensors with pixel sizes well under 2 µm, but there's much less (or even non meaningful) benefit with larger pixels. BSI was initially developed to enable higher pixel densities for small sensors. The observation that you can achieve similar DR with FSI and BSI full frame sensors is consistent with there being minimal benefit at these large pixel sizes. I've also read that BSI sensors are more expensive to produce, and if that's true it begs the question why use them (other than marketing, which is certainly a valid reason).
Aren't we getting to the point where the dynamic range of both is so good that it doesn't make any practical difference?
I think we're pretty much there at low ISO. At high ISO, we aren't but that may take a leap in technology to really improve, beyond a very small amount. Going on the Photons-to-Photos page and selecting cameras from a few different brands, they are mostly the same at high ISO settings, though Sony seems to have a little 1/3 of a stop special sauce at very high ISO, in some of their bodies. But those numbers re not native ISO according to the chart, so it's hard to say. Could be some noise (and detail) reduction chicanery going on. I'm not very familiar with the Sony bodies so I don't know.
I suspect that you are right. If my understanding is correct, DGO needs a fast sensor to simultaneously read out the values with low and high amplification for high dynamic range, i.e. a (partially) stacked sensor for a photo camera. Just BSI is not “enough” to get a higher dynamic range.
