Whether 'half' or 'whole' - I mean either some part, or all, will 'appear' under-exposed.
I am not so sure, though, that curves lifting exposure back are applied by the RAW converter, it may well be that they are applied in-camera
before creating the RAW.
Anyways, the basic issue of how RAW converters fail has all to do with half vs. full. My observation is that all images that are
sure to contain the same exposure on both halves work properly (except for the loss of detail at 12 MP vs. OOC JPG). All images where supposedly exposure differs between the halves is where RAW software fails to produce a good image (loss of saturation and noise increase).
'Harder' , yes - but 'hard', no.
I really don't see any great difficulty in combining the two different ISO/exposure halves - the combination only needs to follow some very basic, and very straight-forward, logic and rules.
Tell that to Adobe, Phase One and ISL.
My guess is that they don't want to invest the manpower necessary to cover the whole EXR + DR range, while all other cams work with simple Bayer patterns.
Both ISO based DR and EXPosure based are a combination of hardware underexposure + curves. It doesn't really make sense to call one hardware and the other software.
I meant in Jpg. Dr 400 stronger clipping/highlight protection than corresponding Dr 100 shots at 2 stops lower Iso accomplish, appr. 2.5 stops total (until clipping areas match).
I think you mean '2 stops lower exposure' (not
"lower Iso" )?
Both. I mean that a ISO 400 DR 400 JPG image saves about 0.5 stops more clipped highlights than a ISO 100 DR 100 JPG. Working on the RAW of the ISO 100 shot needs 2.5 stops exposure in order to see about the same clipping as the ISO 400 DR 400 JPG.
You can test it yourself. Point a bright (LED) flashlight towards a door-frame at some angle. Keep the door open and the background room dark. Shoot that at higher ISO + DR 400 and then turn down 2 stops of ISO + DR 100. The DR 400 shot will show less area clipping white from the flashlight than the 2 stops lower ISO shot. Open the RAW of the lower ISO + DR 100 shot and pull down exposure until you can match the clipping area with the DR 400 shot.
Anyhow, I see what your saying - although I think it would be clearer/fairer to say that the DR100% setting appears to waste a fraction of a stop of highlight range, and that the DR400% doesn't, so the difference appears larger than just 2 stops - undoubtedly just a matter of differences in respective tone curves.
Yes, you are right, else we could not pull exposure down further in the RAW. But for ooc JPGs it's good to know that you get that extra headroom by using
ISO DR. I have yet to check how EXP DR compares to that.
If 'blooming/orbs' occurs at 'ISO200 DR100%' then increasing the 'DR' to DR200% (the max for ISO200) certainly won't help, because sensor exposure doesn't change.
Just what I said.
But - increasing the so-called 'DR' to DR400% demands that the ISO be increased to ISO400 for DR400%, thereby resulting in lower sensor exposure and thereby reduced 'blooming/orbs'.
For the sake of clarity: No, in two ways! Sorry.
1) ISO DR
reduces ISO by 1 or 2 stops, it does
not increase ISO. The one and only setting where ISO is
nominally increased happen when using A-perture priority mode at ISO below 400 and exposure times longer than 1/4 second. "Nominally" means that the same lifting curves are applied to ISO 100 shots that ISO 400 shots would use after decreasing ISO to 100 internally (at DR 400).
And ironically this very specific situation reveals a firmware bug - or rather user conflict - where the camera may
not reduce exposure time to compensate for the
nominally higher ISO. Metering will still use exposure time based on base ISO (without curves) and may thus considerably overexpose. This happens when ISO and DR are
not set to AUTO.
What happens is that the camera/firmware doesn't know it has to decrease exposure time, because for ISO DR this normally is not done. Would you use AUTO ISO with DR then the camera would choose a minimum of ISO 400 and exposure time would change accordingly. Would you use AUTO DR with ISO 100 then the camera would use DR 100 for any exposure time longer than 1/4s.
This is why I wrote "user conflict" above, the user asks the camera to perform an action that is not really possible with its current programming that says: ISO DR does not change exposure time!
So ISO DR at ISO lower than 400 is an anomaly and only happens because of one reason: to allow the user to freely choose an aperture and
not be limited by the max 1/4s EXP DR or min ISO 400 ISO DR limitations.
All other modes impose those limits. In fact A-mode behaves erratic in other aspects, too, all only because of this special treatment (and the confusion of the firmware programmers).
Now of course, it actually the increased 'ISO' that is reducing the 'blooming/orbs', not the 'DR' as such - but the need for higher 'DR' dictates the higher 'ISO', so the two values can be considered to be 'symbiotic'.
2)
If you are using any of the (semi)-automatic modes then the camera automatically decreases exposure time or closes the aperture to compensate for the increased ISO, this surely helps against blooming. No question about that.
But the ISO increase by itself does nothing to help with blooming (other than smearing the distinct disc outlines with noise and better gradations), light exposure of the sensor doesn't change. Most people seem to automatically connect an ISO increase with an exposure decrease, but that's not a given (and a good reason to do tests in M-anual mode for better control of what's happening).
