I also missed some thing. Mr. Davis did not tell all. A closer
look at the curves indicates less exposure for maximum black
with extended development. B. Barlow's pdf confirms that but
he does not say any thing about the increase in contrast.
Davis though does mention the change in gradient.
So, to develop to completion, at least to Davis' mind is
to go for maximum contrast; less exposure more time in
the developer. He tested several papers that way.
I wonder where the end points might be. Dan
Ok, that sheds some light on the subject but if it's true that "to develop to completion, at least to Davis' mind is to go for maximum contrast," then I disagree with Mr. Davis. To me, "completion" means there's no more density to be gained by further development time. I think this can only be judged by the bahavior of the silver in the highlights since one may see maximum density in the shadows before all the apparent tone in the highlights is realized.
From the moment that blacks are black onward, as you are still gaining density in the higher values, you are continually losing contrast. Controlling contrast comes from selection of paper grade or filtration and paper exposure time.
It seems to me that the time to "completion" for any paper/developer combination really doesn't have a thing to do with exposure time. I'd imagine a valid test would be to do something like expose a test sheet by moving a card in increments and develop it for a really long time, like six minutes and Identify the first exposure that's discernable from paper white. Then expose a full sheet for that amount of time, cut it into one inch strips and develop strips for 90, 120, 150, 180, 210, 240, 270, 300, 330 and 360 seconds (mark them on the back before development). Find the one that matches the overlong development time, go one increment further just to be safe and that's development to completion.
If you wanted confidence that you were getting maximum black at that time, you could run a set of strips the same way, exposed for the time that produced the first loss of distinction between one time and the next longer time. I'm sure that not only would the time for completion (as determined by the previous test) but that a number of the black tests strips developed for shorter times would be just as black.
I don't know if I've made sense here.
