Quote Originally Posted by Photo Engineer View Post
If you repeated that experiment several times, I would guess that these differences would even out and the averages would be virtually identical.
One maxim I learned was "never do any experiment one time".
The speed-loss of that last graph is only 1/6th of a stop. As both you and Rudi suggest, this is tiny and may be solely due to normal experimental variation. I only did one experiment at that ratio of 117. The problem is, it takes me a couple of hours to run one experiment. My steps are:

- put a film-strip in tank, held down with magnets.
- measure and mix chemicals to make 100 ml of test-brew (slow when adjusting pH to a target at 20C).
- bring developer and water-bath for the tank to 20C.
- develop, stop, fix, wash, dry.
- measure densities w/ densitometer.
- type them into computer and graph them.
- if it looks promising: compare grain/sharpness with loupes, and scan strip into computer.
- ponder what to try next.

As a result, one experiment is one evening. So an array of tests, such as trying various amounts of some chemical, means using an array of evenings, and/or all day Saturday. And that's doing each experiment just once. And that's doing just strips, not whole rolls. Back in the day, did Kodak's engineers have testers or technicians available to do much of this tedious work for them? If so, that must have been great.

I'm thinking it might be worthwhile to run longer strips containing two frames instead of just one. That may give most of the advantages of two identical experiments, yet consuming little extra time. Any thoughts about how to speed up experiments?

@kb3lms: We're getting closer! The formula is now fairly stable, so now I'm thinking of things like chelation and improving convenience.

Mark Overton