Hi, Ted,
Thanks for that. I'll read it soon.
It [is] nonsense to assume that there is an a priori relationship between the average (or otherwise,
representative) exposure of a scene H(average) and the speed point exposure H(speed point) of a photosensitive array.
Oddly enough, I have for several days contemplated making essentially that same point in a note here, but decided it wasn't necessary!
But here it goes (this before I read the Prais paper cited above):
The phtometric Exposure Hm (used in CIPA EC-004, and its counterpart,. Hsos, in ISO 12232-3-1) are not the value of H expected to exist as part of any (possibly hypothetical) photographic "exposure". Rather they are the value if H that, applied to the digital camera of interest, with its "sensitivity" setting at a certain value, will produce a certain (fixed) encoded digital value (that being somewhat arbitrary).
Of course, we could imagine an actual exposure of some scene, using a certain photographic exposure, that would give,, for example, an average phtometric exposure numerically the same as Hm/Hsos.
We might contemplate such a photographic exposure having been made based on a photographic exposure "recommended" by an exposure meter. If that exposure meter a reflected light meter is visualized as complying with ISO 2720, then the resulting photometric exposure (for example, the average phtometric exposure) will depend both on the scene luminance and the value of K chosen for the meter. But thaty would be a totally different matter.
There is little doubt that the authors of, for example, CIPA DC-004, were very conscious of the use of the sensitivity value "measured" per the standard in exposure metering. But CIPA EC-004 itself (for example) does not at all deal with exposure metering, now with the result of an actual photographic exposure of some scene.
I look forward to reading the Prais paper as soon as I can.
Best regards,
Doug
Best regards,
Doug
