I mean Photons to Photos has been wrong in the past, too. He had very different numbers for the S1R (basically putting it at almost 12 PDR before eventually settling at something like 11.3 or so) and then ended up changing them with 0 transparency. Never explained what went wrong there. Coincidentally he had his (wrong) numbers up somewhen in March/April 2019 and in May 2019 DXO Mark (which, imo, is a site with a more transparent and reliable testing methodology) put out THEIR PDR numbers. And basically after that eventually Photons to Photos ended up changing the result on his site to match DXO Mark.
I've submitted a second set of data for the S1RII to Bill, but he hasn't added it to the website yet (at least, according to his change log). Will be interesting to see if the numbers change.
And, the PDR number is interesting. Bill says it's supposed to help you understand how much DR (which I read as noise) you will see in an 8x10 print when held at arm's length. He also says that higher PDR levels will be overkill for that test and you won't be able to tell the difference. Which makes sense - I'm assuming that an ISO 1600 shot when printed at 8x10 and held at arm's length would have no visible noise.
Anyway, what is of interest to me is how much noise I see when I squeeze all that DR into something with much less DR, like a 10-bit panel or a 8-bit print (i.e., compressing the DR by lifting shadows). And, again, I think pretty much most-if-not-all FF sensors these days when shot at base ISO will have excellent exposure latitude as long as the shots were reasonably (but not necessarily perfectly) exposed. So I don't consider it a buying consideration. Of course, if somebody is using the camera scientifically and directly extracting data from the raw file that's different. But even for landscape, I just don't think it matters anymore, which is why Nikon, Panasonic, etc. are giving up some DR to get speed.
One thing that I do find interesting, however, is his "
Photographic Dynamic Range Shadow Improvement Chart." Which helps you understand what happens when you shoot at elevated ISOs and then lift shadows. The S1RII sensor does very well here - perhaps best among it's peers. And indeed, if I download an ISO 3200 Sony A7RV image from DPR's image comparison tool, and compare it to the equivalent S1RII image by lifting shadows, I do see noticeably more noise in the Sony image. Which means the S1RII sensor should be better for wildlife & sports (at least from a noise perspective) since those are often shot well above base ISO. Same for raw video, etc. So this may be a very savvy choice that Panasonic made, assuming people can get away from screaming "It's got better PDR at base ISO!!!!" for a minute and actually look at images in the raw processor of their choice.
