My point regarding extreme contractions is that one must be aware of the shape of the full curve under these exposure/development conditions, not just the density range. Just because one "brings down" Zone XII to Zone VIII doesn't mean it will print with detail. People assume it will because the Zone System tells us "Zone VIII" prints with detail. But the negative curve might show that when applying N-4 to such a situation, the original exposure values for say Zones IX-XIII all end up lying within a relatively small range of densities. In other words, local contrast in the extreme highlights is severely flattened or lost altogether. The negative might fit the paper, but who cares about that if all you get are muddy, featureless highlights (not to mention lower shadow contrast even with more exposure).
The danger of "blown" highlights applies as much to extreme contractions as it does to severe overexposure. This is why when dealing with very wide subject brightness ranges, each situation needs to be considered carefully. In certain cases for example, I may elect to apply a much more mild contraction than a simplistic view of the Zone System indicates. I'll get good separations all the way up to Zone XIII or XIV. Of course burning and dodging or whatever other techniques might be required in printing, but I'll likely have a more satisfying print in the end than if I had contracted the negative to "fit" the paper. Not to mention I'll have a much easier time printing the darker areas, which might be 80% or the image area or more.
When you see people talk about extreme compensating procedures, or N-numbers like N-6 for example, if indeed they are getting good prints with good, clear highlight detail, it is likely because (again to steal Stephen's words) they are not getting what they think they are getting. They are probably getting less contraction than they think they are.
I may not have explained this clearly, so I apologize in advance. If there is interest I could post some curves to demonstrate. But probably a topic for another thread as I'm referring to rather extreme cases (although they are situations I deal with often).
Sorry for sidetracking this from Stephen's testing/methods topic and his questions to Andreas.
Last edited by Michael R 1974; 01-16-2013 at 09:53 PM. Click to view previous post history.