While I’ve observed the effect you describe, I’m at a loss as to why Sony would divorce ISO from signal amplifier gain and base a reported ISO on a completely different criteria. There must be a good reason for this, right? Where is the Sony Picture Profile ISO described/documented? I’m interested in learning a bit more.
Late to the party but RawDigger can dump the raw curve. It is linear and equal for the three raws posted earlier.
e.g. 4077 → 65232 each one. To me, that means differences in brightness have nothing to do with the raw tone curve for what that’s worth. I know nothing of Sony cameras and so have nothing to add, sorry.
I run Linux only. Is there something on the Open Source side that might do something similar? Or, minimally, dump all the tags, such as the aforementioned “Sony ISO.”
exiftool -a -u -s -G1
1 Like
Brilliant. Thank you.
If some tags are reported as binary arrays, you then dump specific ones using -b.
Right. Got it.
Indeed, it’s very interesting to see how ISO values can diverge when using Sony Picture Profiles. It’s the S-Log profiles that seem to drive the value divergence.
Because that’s literally what the CIPA standard for digital camera ISO ratings says to do. The only thing “special” Sony is doing is saving out a second piece of metadata that specifies what the ISO would be with default settings.
Note that the JPEG curves are not stored in any metadata I’ve ever seen. That’s what motivated GitHub - Entropy512/camResponseTools: Miscellaneous tools for reverse engineering camera response curves - I wanted to determine the actual behaviors of various Sony PPs
There’s a separate ARW nonlinear mapping curve that is part of their default lossy compression algorithm but I’ve never seen it change.
2 Likes