Here is my impression of what your film curves would look like if they did not include flare.
I compressed the Log E scale in the toe by taking what I thought was a reasonable speed point.
Using that as if it was equal to EI 320 which is by definition -2.6 Log mcs.
Add 0.4 Log E flare to that = -2.2 Log mcs.
The difference (arithmetic) is 0.0038 mcs.
I then built a compressed scale by adding 0.0038 mcs to each 0.1 log increment going from the speed point to the right until it stopped being significant.
Some example values:
-2.6 Log mcs + 0.0038 mcs => -2.2 Log mcs
-2.5 Log mcs + 0.0038 mcs => -2.15 Log mcs
-2.0 Log mcs + 0.0038 mcs => -1.86 Log mcs
-1.5 Log mcs + 0.0038 mcs => -1.45 Log mcs
-1.0 Log mcs + 0.0038 mcs => -0.98 Log mcs
By -1.0 Log mcs the difference caused by flare is 0.02 Log mcs which is insignificant to me.
There are a couple of threads:
Originally Posted by Rafal Lukawiecki
What is the Relationship between Film Speed and Camera Exposure?
Is the K factor relevant to me or should I cancel it out?
Hiding in Plain Sight
These papers are a good start:
Calibration Levels of Films and Exposure Devices, Connelly.pdf
Contrast Index a.pdf
Exposure-Speed Relations and Tone Reproduction.pdf
safety factors in Camera Exposure.pdf
Beyond the Zone System
Photographic Materials and Processes
Theory of the Photographic Process
Which CI & Calculating Flash Power for Transmission Tablet Contact Exposure
I am back from Wyoming, it was a wonderful treat for my eyes and for my soul. I am nursing 72 exposed sheets of 320TXP, a mixture of N-1, N, and N+1. Based on my earlier, flawed and flare-affected testing, described in this thread, and based on various recommendations here, I have treated N-1 as EI 160, N as EI 200, and N+1 as EI 320 (or 250 on a couple of sheets by mistake). Now, all I need to do is to figure out the corresponding development times, for XTol 1+1, to be able to print them on my diffuser (Ilford 500H) enlarger, with my usual papers, MGWT/MGIV FB.
As far as I can see from reading your detailed suggestions, bar Bill's post 41 which I have not yet fully followed, it would be good if I retested for the development times by contacting my 31-step Stouffer and so eliminating flare and other issues. I have just finished a marathon of reading other related threads, also on sensitometers, exposures when contacting and so on, and I am not sure if I would get reliably repeatable exposures by using my enlarger (equiped with Ilford 500H) due to lamp warm-up and start-up times with short exposures. Also, this light is a green/blue mix, not a white light, though this might not matter for this test.
My feeling is that I could use a Speedlight SB-800 in manual mode, checked with a flash meter (Sekonic L-508). I followed Stephen's 8 Feb 2009 calculations for the required contact exposures, on this thread, and elsewhere, but I don't know how to adapt those calculations to a flash exposure. My meter tells me a needed f/stop, and I do not know how this converts to the fcs or other measures of exposure at the surface of the contact printer. I don't mind doing some pre-testing to fine tune the exposure, but it would be nice to start in the right ballpark, and it would be a bonus to understand the math. I can vary the flash output and distance, and I know its various GNs.
My second quest is to figure out which CIs are the ones I should be striving for my N-1, N, and N+1. WBM clearly suggests some, but since they are based on the WBM technique for calculating them, I am confused if I should follow those. In fact, I am not even sure how to find out CIs from my new tests—other than by begging Bill again—since I have previously relied on Ralph's spreadsheet.
I suspect I will have to figure this over the course of trial and error, but I would love to give my new 72 sheets the best I can with the knowledge I have so far. Bear in mind, that until now, my previous few decades of photography, including 4x5 since 2000, were never based on testing but on manufacturer's recommendations, +/- about 30% for N+/-1, add-and-subtract some experience-based-factor for HP5+. However, I lack that experience with 320TXP. Overall, I like the idea of being more in control of the process, so if this stage succeeds, I will also test, for the first time, HP5+, to find out how it compares to my past experience.
Rafal, regarding what CI to target for N-, N+ etc, everyone prints differently and has their own preferences. Some people prefer softer negatives, some prefer less or more contrast in the negative depending on their chosen papers and preferred techniques, some people prefer a thicker negative (ie more exposure) etc. To that end, I'm not convinced there are better ballpark starting points than some of the target densities in Adams. They're as good as anything else. Perhaps start there, make some test images and print them so you can figure it out for yourself based on your own printing. There's nothing like doing a lot of testing and dialing in the process only to find out during printing you're not getting the negatives you thought you were getting, or thought you wanted.
I have to also say I'm generally not a fan of metrics such as CI, Gamma etc. I look at them as "nice to know" results after plotting the curves, but it's really the full H&D curve that tells me what I need to know, not contrast index - particularly when it comes to N minus territory.
Thanks for confirming my other suspicion, Michael, that CI alone might be an oversimplification. I started to fear that when I noticed that a single set of my (flawed) test numbers could be interpreted to produce quite a few different CIs by using different methods: WBM spreadsheet gave one set of CIs, Bill very kindly provided two others, with and without flare compensation. Unfortunately, I do not have the experience of correlating the shape of the curve to my printing experience, as I have never cared for the curves in the past 30 years. I can see, however, this would be useful: for example, I have found it harder to get the local contrasts in skin tones right. Now, I suppose this would be visible from a middle section of a curve, but it will take me a while to make that connection. I look forward to being able to relate my existing printing practice to the curves and tests, hoping for more predictibility and fewer difficult negatives in the future. In the meantime, I would like to figure out the development times through new, better, tests. Thank you for helping me, any additional suggestions are much appreciated.
Originally Posted by Michael R 1974
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
PS. I have started a separate thread to ask for advice on using a flash unit as a sensitometer. Hopefully this means it will be easier to carry the discussion about CI and so on on this thread, without confusing the issue. I hope I am not going to cause any upset by doing that, and thanks for your help.
I guess what I don't understand about things like CI is, why "summarize" the curve with a straight line of best fit when you already have the full curve?? Why not simply plot the curves and compare them?
I don't want to give the impression it is wrong to use CI. I would say that with most of today's films which have relatively long straight line sections, CI is fine as long as the exposure range lies on that straight line. However when dealing with high contrast situations and/or minus development, I need to know what the ends of the curve look like - particularly the shoulder. CI on it's own doesn't tell me enough, and it complicates things for me.
If you've got an understanding of how flare impacts both testing and real world photography (thanks to Stephen and Bill), the rest of the testing for your personal EI and development times should be pretty straight forward. Don't overcomplicate it. And take it from someone who's been down all these roads - don't waste too much time on the n'th degree of precision. There are people out there who rate their film speeds in 1/4 stops, try to get to within 1 density unit of their targeted Zone VIII densities, and determine development times in 10 second increments. I don't care who you are, there is no such precision in the process. Strive for high quality and consistency, but dont get sidetracked into endless testing.
No matter what you do... you have to pick an amount of time in the developer, and that is a single number.
Originally Posted by Michael R 1974
Hiding in Plain Sight thread
Hiding in Plain Sight Post No. 13
Using the highlighted row in this chart and working off the assumption that N = 7 stops, N+1 = 6 stops and N-1 = 8 stops of subject that has to fit on the No. 2 paper.
AND, assuming I'm correct with my flare-chart adjustment...
N time should fall around 6 minutes.
N+1 around 9 1/2 minutes and
N-1 around 4 1/2 minutes.
I would not recommend development times of 4 1/2 minutes because it could lead to uneven processing.
In the case of N-1 I would suggest a change to your developer. I think Michael R 1974 can give you some good ideas there, as it's one of his specialties...
Obviously, but what is to be gained by basing that time on a CI versus a plotted curve?
Originally Posted by Bill Burk