Cloud Highlight Recovery - SD1 Merrill w/filter

Scottelly wrote:

I don't notice "visible coloration" in any of those Ted.

Neither did I, Scott, which is why I said "Do note a complete lack of visible coloration".

The highlights do look well recovered though. I think I'd shoot a little over-exposed, so the shadows are more easily recoverable. Do you think it would make sense to shoot at ISO 400, or would the dynamic range be more limited at a higher ISO like that? (I've read that dynamic range is limited at higher ISO settings, but never really experimented much with that stuff. I just shoot at ISO 100 and 200.)

As always, that would depend on what you mean by "dynamic range". With non-AFE Sigma cameras like your SD14 it would not make sense, IMHO.

Let's say you shoot for the clouds at 100 ISO, using ETTR in a Merrill. The clouds are captured but not blown. Let's say the highest raw value is 4000 and let's say the read noise is maybe 40. One definition of the DR 'low point' is where the read noise is equal to the minimum acceptable signal level (there are other definitions but that one is ISO's). So, from that, we get a DR of log2(4000/40) = 6.64 EV.

Now we shoot at 400 ISO. The highest raw value becomes 1000 but the read noise and therefore the minimum acceptable signal level stay the same. The DR is now log2(1000/40) = 4.64 EV, i.e. two full stops of DR have been lost.

Worse follows: SPP applies a multiplier of 4 for 400 ISO. So, at 100 ISO that read noise is say 40/4095x255 = 2 or 3 after conversion to an 8-bit image but 4x40/4095x255 = 9 or 10.