Changing developing style - recalibrate all or averaging and do some math?
If you have nice film curve family and you have extracted all data from it, having a good developing standard with a one film and developer and one agitation style.
What's the best practice to use when using different agitation style. For example, going from tray development to slosher, rotary etc?
As far as I have noticed via some tests, the correction factor is easy to found, but it's really a kind of average value.
For example, around N-development the correction factor from tray to slosher could be 1.1x but it goes to 1.2x when approaching N+2 contrast. And goes somewhere like 1.05x when doing N- developing.
The change of required correction compared to contrast varies for a every film. If one film has almost linear correction factor (none has really linear), another could have quite extreme, N-2 with correction factor 1.0x and N+2 with factor 1.25x.
When using single correction factor with this kind of film (either averaged or taken from somewhere near N contrast), the error of unlinear nature of film will cause 'significant' error in the resulted contrast when intented development is near extremes, N-2 or N+2.
But does it really matter? As the variable contrast papers will give a lot of room for corrections..
Probably most of films doesn't have a such unlinearity, but respond more linear to changes of agitation style.
So back to the question. Do you prefer to re-calibrate everything (use the 5 sheets and do new curves) if you go from tray to rotary or from tray to slosher or from losher to dip'n'dunk or...
Or is it considered good enough to take a single (or two) reference test and calculate correction factor and use result for calculating times for further developing?
In my tray development -> slosher this would be usually near 1.1x (ranging 1.1x to 1.15x).
Of course this raises an another question. How exact is enough? As the usual methology used for getting data for film curves has so many variables which all has a some degree of error margin.
I guess that a some sloppiness is acceptable
I just found myself wondering same thing again.
I haven't found any data or figures about development time difference between different agitation styles. Only general assumptions like going from intermittent agitation to constant agitation, the new development time should be 10% less than original.
These usually implies that relationship between two different agitation style is somewhat linear.
My tests from random points have always gave unlinear results.
One example of my results, which are way from linear (like 10%) is TMY2 and D-76 1+1. With intermittent agitation (each minute) 8 minutes will give N-2, but with constant agitation I got same result in 5:45.
That's 39% shorter time.
N+1 time with intermittent agitation is 16 minutes and with constant it's 13 minutes. The time is 23% shorter.
So, developing time relation between N-2 and N+1 with two different developing method (yet somewhat similar!) is way from linear. This is confusing. Should it be that way?
You are asking an engineering question, and I'll restate your question using variables and ratios in hopes that one of the engineers on the list will venture a reply.
Let TCA = dev-time with constant agitation to achieve a given density.
Let TIA = dev-time with intermittent (each minute) agitation to achieve the same density.
Let RAT = ratio of agitation times = TCA/TIA. Note that RAT will always be less than 1.0.
Here are your observations:
RAT(N-2) = 0.72 to get N-2 with D-76 1+1.
RAT(N+1) = 0.81 to get N+1 with D-76 1+1.
You are asking why those two RAT values are different. In other words: Why does constant agitation affect development at a different rate when developing to a different density?
Probably only PE knows for sure, but I'll guess that when density is higher (near the end of development), the by-products of development slow down development less, and therefore constant agitation causes less of a boost in development-rate, thus raising the RAT value.
I would do the five tests over again if I changed processing deliberately and planned to keep it that way.
If it was just something I noticed drifting... Then I adjust my agitation until I hit the old test times.
In my case I tray process 6 sheets of 4x5 TMY2 for 13 minutes in D-76 1:1 for N with continual agitation. I track whether or not my times are hitting predicted CI and I look for problems if things drift 10%.
There are many variables here, one being the type of film and how much expansion is possible. It makes sense (to me) that albada's RAT would increase as CI increases dramatically. I would also point out different agitation schemes can result in different curve shapes, not simply a change in slope. There is also always the possibility of error in test results. In general I don't like formulas for changing variables. Depending on all the various differences in every individual's processing techniques and conditions, formulas may or may not work accurately, and variations of +/- 10% are not very large.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
I will do whole tests for different methods just to be on the safe ground. I agree that 10% variation is acceptable. The shutter may not be accurate, agitation can have variations, developer may not be fresh, ...
But after all, it's the negative, not the final result =)
Some folks shoot from the hip, while others prefer to take aim..........I'm a take aim person, as long as there is not a large expenditure of time and materials. Five sheets and a little chemistry, not at all excessive in either category, especially using R. Lambrecht's film testing Excel spreadsheet........recalibrate.
The factor depends much on the type of developer. Devs without redox system (Rodinal, Microdol-X) are much more sensible against changed agitation.
After looking some curves from Kodak's Tmax film documentation, it's quite clear (to me) that even within same developing method, the CI can be approximated only within a marginal. Even with a usual 5 sheet test, the real CI between tested points cannot be calculated.
The CI curve of D-76 1+1 tells a lot.
Makes me wonder why didn't I look carefully these curves before
Originally Posted by Usagi
I use those charts for roll-film, there's a different chart for sheet film.
But when you do your own 5 sheet test, you obtain 5 points that you can use to draw your own Time/CI chart.
When you draw lines between the points, then you can read the chart for the CI results you might get for any development time you want in-between your 5 times. You can even make a good guess what results you might get for the next few minutes longer than your longest test.
Then in any live development run, you can put a one-sheet test with the film. Find the point on the Time/CI chart for that one sheet of film. If it is above or below your 5 sheet test curve... Then you can see if you are "in control" or "out of control". I can still print when I have a deviation of 0.1 CI from what I aimed for, although I would call -0.1 CI "out of control".