See my L-mount mirrorless wishlist
In Panasonic S1R: True ISO 50, not a Faux ISO — Kudos to Panasonic, I discussed noise and and the signal/noise ratio derived from images I had taken from ISO 50 to ISO 51200 in 1/3 stops (RawDigger folks analyzed the images). That series was not ideal and not shot for the goal of showing the true S/N ratio because it was shot in-focus, which means that standard deviation calculated on the gray patch is increased by subject detail, including dust.
I re-shot part of that series in whole stops, this time defocusing the image so that subject detail would be blurred away. I was interested in seeing how much the S/N ratio would improve using Multi-Shot High-Res Mode versus single shot. I concluded two things:
- Visually and with careful inspection at enlarged size, I determined that there is a subjective 1.5 stop improvement in noise level by using Multi-Shot High-Res Mode versus single shot. This matches the theoretical improvement in noise from 8X the exposure: √8 = 2.8 which is 1.5 stops (2^1.5 = 2.83).
- Multi-Shot High-Res Mode really only makes sense at ISO 50, is marginal at ISO 100, and no point at ISO 200 on up. Given the time and space required, this can be summarized as “in terms of noise alone, High-Res mode at ISO 50 is the only ISO that makes sense”. At ISO 100, you’re dealing with an 8 exposures huge file hassle for a 1/2 stop gain in noise performance at most—pointless since a single shot at ISO 50 is so close; The S/N graph suggests ISO 50 single shot is as good as ISO 100 MultiShot, but visual inspection repudiates that idea; ISO 100 MultiShot is slightly better.
I expand upon my comments in Panasonic S1R: Signal/Noise ratio for Multi-Shot vs Single Shot, ISO 50 to ISO 1600.