telenous said:
It seems to me the only way you see the image as it was captured through the camera lens is in a positive film (and then only when you look directly at the mounted slide). Both scanning and printing through the enlarger lens are bound to affect somewhat the original film. Is there a degree on which one affects the whole process more than the other? I do not know - and more importantly, I think, there is no way to know, because all we ever see are images through scans and printing! At the end of the day. whether one prefers to scan or print is a matter of taste, but I can't see how printing is somehow meant to be an unadulterated way of interpreting the film.
First, for a good example of Rodinal with a slower film, look at this
APX 100/Rodinal 1:50 shot. Scanned well, of course, but I would expect a good optical print to be even better.
I agree that a b&w, first generation positive (i.e. a b&w reversal original), would be the "best" example to evaluate, but that has its problems, too. There could be as many b&w reversal chemistries as one could imagine (well, almost), so you would have to standarize on one as a reference. But then you would undoubtedly find that films looked different in different chemistries. E.g., HP5+ might look better in DR5, Tri-X would look better in some other formula.
So in practical terms for me, I look at negatives under a loupe as a first step, but an optical enlargement under controlled conditions as my final reference/evaluation point. To be truly "controlled", there needs to be voltage control and control of lamp aging. One would also have to either select a particular light source type (point source, condensor, tungsten (or halogen)/diffusion, cold light, LED, etc.), or make multiple testings using a variety of lamp sources.
Of course there are dozens of other variables that one might wish to control, so it could become a formidable effort.
For practicality for most amateur effort, of course, one standardizes on the major variables ... enlarger, lens, paper, developer/processing, etc. In the end it is about an individual's workflow. IMO, the optimum in b&w is still to have everything in the analog domain unless we are talking top-level professional (drum scans, etc.) digital processing which, for most of us, is not practical for our daily hobby/amateur work.
My b&w reference standard settled on Zone VI Brilliant or Ilford Galerie paper developed in Amidol. (Today I could not use Brilliant, of course.) I even did 4x5 and 35mm contact proofs on those papers, at least for the most part. That kept things consistent in my mind. The emotional impact of a 4x5 contact proof on a good FB paper is quite different than a "quick and dirty" proof on an RC paper. Since the final print can be largely about emotion, that's an important factor for me, as it provides continuity. YMMV.