A bath not proccesing adequately in its is second run not necessarily must have been too bad for yielding in first run results up to standards.
Or do you see that differently?
Yeah, I do see it differently. But keep in mind that I have not actually done the test with developer changing significantly during the process run, so I am guessing at the result.
Here comes one of my long, tedious posts:
I see the developer as going from "good" (ideal) at the beginning of the process run, to substantially deficient at the tail end of the run. So that the actual result is sort of an average of the entire changing process. And I understand that a lot of people will be happy to simply look at that final result, then decide if they are satisfied with it or not. My view on this is that there is nothing to compare it against, so it is easy to be satisfied. But SHOULD one be satisfied? (I'll put some actual experience in the last paragraph.)
So to decide if the "averaging" of the developer condition is "good" (or not), I think that one needs a comparison. It could be either the beginning of the cycle (completely good developer) or at the tail end of the cycle (degraded developer). I just picked the tail end, as I think the result would be more obvious on a control strip, and thus be more eye-opening to someone. From there I think they are in a better position to decide if they want to investigate further or not. Let me demonstrate with an example of personal computers: imagine that you were still using your prior version, and had never seen anyone faster. If someone were to ask, "are you completely happy with it" I'm guessing that you probably would be. But... if you had the opportunity to compare to a more modern one... suddenly the old one is perhaps intolerable. But without a comparison, everything seems perfectly acceptable.
Now, the photo experience I mentioned... let me say that my employer was a studio chain in the us - literally thousands of studios, and every hair color and complexion that exists in the US had probably made it into our studios. (The chain no longer exists, went "belly up" within about a year of Kodak's bankruptcy) But we did portraits, and tested pretty extensively to find "good enough" materials - films and papers. These were always the professional portrait/wedding films, and matching professional papers of the day. Nothing else was adequate, in our view. When we tested a film being considered we shot it with about a half-dozen models, representing the range of our customers. We ran this over a wide range of exposures, not that we planned on shooting this way, but with thousands of studios open 7 days a week, there was always someone who would shoot all week with the lens aperture wide open, etc. So we want to know ahead of time if this is a potential problem.
When we evaluated the results it would be in a large color booth, maybe 12 or 15 feet wide with neutral colored walls. It's large enough to lay out 40 or 50 8x10" prints, and allow 3 or 4, or more people to all sit in for review. We critically evaluate the skin tones for good reproduction and color crossovers. For example, perhaps a person with a very dark complexion gets slightly bluish highlights: we say, where are the other exposures? Let's lay them side by side; does it get worse in certain conditions? Ok, let's color correct out the bluish highlights and see what happens... So and so, would you rerun a set of prints on this? Pull out 1 cc of blue, and let's try minus 2 cc at the same time. It'll be an hour or so to get the prints, so lets meet back here after lunch, ok? And this is how it goes. If it's a paper we're evaluating, then it's also, so and so, would you get some image stability tests started (it's gonna be a couple months for even preliminary results)? And we wanna run latent image shift tests to see if the time delay between print exposure and processing is a problem (we ran 500 to 600 ft rolls of 10" paper, and it can take nearly an hour to expose the full roll; so it could potentially go onto a processor with some images only 10 minutes old, but others nearly an hour).
Something else we include in the test scenes is a selection of colored fabrics - strong red, green, and blue colors, as well as some smocks the models can wear. When we lay out prints in the color booth we'd also bring over the real fabrics. Most of the time they are a close match for the actual fabrics (except that the real fabrics laying on the countertops have folds in them). People here may think, that's not very important. But when a mother has picked out a certain pastel colored Easter outfit for her little girl, she doesn't want a different color on the photo.
I've been digressing, but my point is to give an idea of what we go through. So if someone says, "hey! Let's test with an out-of-spec film process!" well, no, we're not gonna do that. It's such a touchy, finely tuned process that it seems obvious that an out-of-spec film process is gonna bring in color problems somewhere - how could it not?
But to finally get to this point of comparing: we had rolled out a certain printer - a high-quality dye sub machine - in a number of local studios. I'm acting as an in-house product manager for it - I've made special ICC profiles for it, as well as an exposure correction routine for studio use. Anyway, here's something I've done maybe 50 times, always with the same result. I stop by a studio to see either the manager or a senior photographer - whoever they see as their "expert" for onsite printing. I have them pick out a good example of a portrait session on their system - one that they think might be difficult. They're gonna do their best with it - whatever adjustments they think they need to make (they always think that every sitting needs their personal adjustments to be its best). So they make adjustments and spit out an 8x10" print. Are they perfectly happy with it? If they think, maybe a bit more contrast would help, I say to go ahead, then we make another print (I have to let them "charge it" to my department). When they finally get the best they can do, I pull up "my system" on their computer (IT has previously loaded it remotely). Now they just have a few clicks to choose from - it's fairly obvious which is best. We print it, and now the question is, which print do you prefer? Every time they pick mine, and there is no doubt about it - this is the one they want. Up until that moment they thought that they had the "best;" now it is no longer good enough. And the whole thing is - they didn't have "better" to compare against.
So I think the same thing may well happen during the one-shot processing. Now if it's a subject matter, such as a landscape, etc., where there is not a predetermined, set way that it should appear, then it probably doesn't matter much. But when it's a studio portrait, there's really a fairly narrow range that is "best." Most people won't see it with a single print - they can adapt to just about anything. But if they have a handful of variations to choose from they generally pick something in a certain range.
Anyway, I think it's generally a good idea for someone who plans to run C41 with sparse amounts of developer to run at least one set under "good" conditions. And don't think that a control strip will predict the results. They're typically just several neutral patches - they don't show complex interactions between dye layers when different colors predominate. We used to run something like 40 control strips per day, but when we want to know how a certain color reproduces we photograph and print it.