Thanks Jay for starting this thread. I believe this is an important topic. We all have a tool box of techniques that we carry with us when shooting, so why only use the same one or two tools? It is just part of visualization after all. The problem is to find a good example to start.
For very short luminance ranges.
I have this image of a detail of a gravestone. It was late in the day. The weather was dark, windy, and with light rain. It was a typical day in Ireland. The stone had almost no separation. The light was so soft, there was virtually no "shadows" cast.
I wanted the final image to be very dark where it looked almost totally black until you walk up to it. I decided to place the exposure of the stone just under the midpoint, or about Zone IV. Even though I wanted the image dark, I didn't think it would work without a lot of local tonal separation. Extended processing has more of an affect on the upper part of the curve than the lower, so in order to maximized the amount of tonal separation with processing, I placed the exposure higher on the curve than I needed knowing I could print it down. I gave the film +2 development knowing it wouldn't be sufficient, but also knowing I could print it on a grade four paper and use a number of additional techniques if needed.
It did print on a grade 4, and I did just a touch of bleaching to emphasis a few places. Why not just push the film 4 stops? Three reasons: One, extended processing of film increases the higher densities to a greater proportion than the mid tones and shadows, whereas increasing the paper grade expands the mid tones to a greater extent. Two, the undesirable factors that come with pushing film such as grain. Three, it's also extremely difficult to visualize how all the elements will look at such an extreme contrast increase. Not over processing gives you more options.
One of the problems of extremely expanding short luminance ranges it that areas that you don't want to expand, also expand. Sometimes this creates looks of small annoying white spots, and sometimes it makes the overall image look harsh.
Some times it's easy enough to spot out the annoying stuff. I have an image of a burnt wall which I handled in a similar manor as the gravestone. It takes two hours to spot out the new created highlight speculars that now look like dust.
One way to fix that problem and when the image looks overly harsh is to deal more with only the tones you want to change. A good example is rock art. Petroglyphs are drawings created by etching through the desert patina to expose the lighter rock underneath. The contrast range is very limited, but often the sun is hitting the rock creating a normal luminance range or at least expanding the luminance range some. The question is how much can you push the film before the image just looks too contrasty and unrealistic? Using the above technique might work in certain conditions, and look harsh in others. A good way to approach such a scene is to give it normal or +1 processing, and then print it so that the rocks look natural. At this point, the petroglyphs will still look flat.
Off the top of my head, two techniques can help here. One is to use a contrast mask while printing the image. The mask will add density only to the areas you want. This will lighten up the drawings without adding contrast to the rocks. Another way is to bleach. Take a paint brush and paint over the drawings using potassium ferricyanide until they reach the point you want.
The concept that I like with this thread is that this is where theory (such as tone reproduction theory) meets creativity. It is the very essence of photography. Nice idea Jay.
Here's an interesting question. Except for special cirumstances, all scenes look best in a print when there is a full range of tones. Why can't we let a flat scene remain flat in the print (apart from artistic considerations)? It's flat in nature, why not the print? Maybe a question for another thread.