Well I guess the fixed flare model is better than no flare model, right?
The Time / CI chart is done. But the answers are the same as my guesstimate.
I'm sure it will come in handy anyway.
Look this over and decide which looks best for you. They are based on LSLR 2.10, NDR 1.05, and 0.40 Flare.
Originally Posted by Bill Burk
This example has the processing times, EFS, and Delta-X speeds using the three different dev models.
Last edited by Stephen Benskin; 09-12-2011 at 07:37 AM. Click to view previous post history.
From the spreadsheet... N and N-1 are the same in all models.
N+1 recommends slightly less development practical vs fixed.
-I might stick with fixed flare model times for plus development because, by the time I identify a neg that needs more than N development, the more the merrier.
N-2 and N-3 recommend slightly more development practical vs fixed.
-I might take the practical flare model times for minus development times to avoid short development times that lead to uneven development.
I have some paperwork to do (find the LER I really want to target).
That's how I like to do things too. Know what the facts are and then make the decisions that work best for you.
I'll send you some more comparison stuff on the different models. As you saw, the differences start to show in the the further out from normal. If I remember correctly, a +2 with the fixed density has the same CI as a +3 with either the variable or practical. That's when the differences become more significant.
Last edited by Stephen Benskin; 09-12-2011 at 12:58 PM. Click to view previous post history.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
Bill, I've attached a few more things. The pdf file is of an unpublished paper I wrote which is basically about this very subject. The other two files are a chart and a graph from the paper.
The graph is a plot of the CIs produced with each flare model. The variables were LSLR 2.20, NDR 1.05, Flare 0.40 (at 2.20).
The chart are the CI values from the graph in a nice little table.
This a continuation of one of the themes of this thread and that is the importance of having a testing method that reflects the conditions of the real world as well as able to accurately predict the results obtained in real world use.
Which flare model is most accurate? Which one will produce the best results in the highest number of cases? There are times when the question can be more than just an academic one. Look at the difference between +2 Fixed Flare and +3 Variable Flare.
Something else to think about concerning the topic of what best reflects "reality". Take a look at the speed values from my last post's attachment. There are the results from two different speed determination methods. The speeds were derived from the same set of data. They only difference was how that data was interpreted. Unlike with the NDR example, the difference in speeds isn't an illusion. The one labelled "EFS", effective film speed, uses the fixed density method where the exposure is calculated at the point on the log-H curve where the negative is at 0.10 Fb+f. The other method is the Delta-X Criteria which is a simplification of the fractional gradient method. The ISO speed determination method is an example of the Delta-X criterion.
The two speed/CI curves show the results of the two methods in graphic form. The differences in results between the two methods raises some interesting questions about what we think we know about film speed.
Last edited by Stephen Benskin; 09-12-2011 at 08:50 PM. Click to view previous post history.
This is example of comparing the placement of Zones.
The curve on the left shows the difference between no flare conditions, which can be considered testing conditions, and the addition of one stop of flare. This is how the Zones would fall if the film was rated at the ISO speed. That is if the film speed from the film/developer combination used in the test was the same as the ISO speed.
The curve on the right shows the same no flare placement (A). Zone Placement B is the 2/3 stop change in the EI rating to bring Zone I up to 0.10 Fb+f. Zone Placement C shows where the Zones would then fall under average one stop flare conditions.
Just out of curiosity Bill, from your real world tests...
What film n developer are you using?
When you did the ASA standardizing tests, what did it come up as?
What times have you come up with for your developing N values so far?
Anyone can make a Digital print, but only a photographer can make a photograph.
Ahh, yes flare again. Seems last time we had a thread on that I got held up with a lack of a good Materials and Methods section in the papers I was reading. I was trying to get a grip on the specifics of original speed papers. Seems they photographed transparencies rather than true scenes. Not sure about the flare characteristics of that system. The paired H&D curves were made, presumably in the standard contact-print method with a step wedge.
A number of different systems have similar CIs for normal, +1, etc, for the simple reason that they work. The Dev Chart CI Comparisons doc compares four different models, No flare, Practical Flare, Way Beyond Monochrome, and Zone System. Except for the no flare curve, which is included only or reference, the different development models are surprising similar. This means that all of them will produce similar negatives for similar conditions.
What's different is that they assume different values for the variables from which they determine the numbers. The second table Variables for Developmental Models has the variables for the different models.
The attached graph illustrates how this works. By simply changing either the aim value for the negative density range, the value for the subject luminance range (at normal), or a combination of the two, the no flare curve can be shifted over matching up its values with the ones incorporating flare. While the results reflect "reality", the variables don't. Changing either the NDR or LSLR variables basically incorporates flare into the results without acknowledging it as a factor.
(a bad analogy) It's possible to predict where the sun will rise whether you use a heliocentric or geocentric model. Both answers will be the same, but only one of the explanations is correct.
Last edited by Stephen Benskin; 09-13-2011 at 10:08 PM. Click to view previous post history.