When I first saw Milpool’s straight, linear divided-development curves, I thought: ‘Great. This is what I think I want!’
I read about how testers do not like the slow toe on Tri-X, as it doesn’t give them adequate separation in the shadows.
At the same time, it seems that there is the expectation that a compensating developer will deliver an even slower shoulder. Stand development is seen to deliver a high degree of compensation.
In one of his very informative videos (on Youtube), Analogue Andy created a curve for stand development in a Pyro developer. The curve was described as similar to wet spaghetti thrown against the fridge. This curve is not at all linear, and in the high values it shows the totally opposite of a shoulder. You can probably fit this part of the curve with an exponentially increasing curve.
(101) Pyrocat HD: Edge Effects Via Stand & Semi-Stand Development - YouTube at 1:08 into the video, he looks at the curves.
From a curves-viewpoint, it is not what I expected, but maybe my understanding is limited. It seems that only the mid-tones are compressed. In any case, I wish that I had seen Andy’s curves before trying stand development. His video also demonstrates key points regarding the frequency of agitation. Very informative.
I assume that users are happy with the result of stand development and the associated shoulder/highlights representation. If Andy’s curve is anything near usual for stand development, I would conclude that the straight line divided-development curve also would be acceptable from a high-lights viewpoint.
Currently, I am thinking that I want my developer to yield a long straight, linear curve without too much contrast. Then, for a particular film, I could use only one EI and one development process. For the contrasty scenes, exposure could be placed on Zone VIII and for the rest of the scenes Zone III would be used.
Maybe it would be similar to using Kodachrome 200 (long time ago) or a digital camera at ISO 200, converted to monochrome afterwards. I don’t know about the linearity of Kodachrome, but I think of the digital sensor as having very linear curves in use.
Looking at the various recipes (Stoeckler, D-23, BT2B), I agree with Milpool that they are all very similar. I have been thinking of using bath A at 1:1 (with the same agitation!) to get longer, more controllable development times. Or instead maybe adding a small amount of Potassium Bromide.
I have unsystematically been using the higher definition Kates version with Tri-X and HP5+ and I have not noticed any issues regarding uniformity, but would like to learn more about this.
It seems that there are a number of old threads about Divided-Development that I need to read.