ID11: Batch to Batch Variability Nilgreenspun.com : LUSENET : Large format photography : One Thread
After my last post on this topic, I needed to determine whether or not I was experiencing batch to batch variability for ID11. My last data were confounded with the effects of doing calibrations at the same time, so I really only had a suspician. Before going to the trouble of getting a scale, chemicals, etc. for mixing my own chemistry, I decided to test specifically for batch to batch variability.
I used three different batches of ID11 chemistry. SMALL: Chemistry for 1 Liter purchased at Pro Photo. OLD: Chem for 1 Gallon purchased about 1.5 to 2 years ago. NEW: Chem for 1 Gallon purchased two weeks ago at Citizens Photo. I debated about using old chemistry, but Ilford technical support said there was no difference and that ID11 should last indefinately. So, it was included.
I exposed 9 sheets of film (all from same box) to Zone 8 level of light. Aperture left untouched, and shutter was "exercised" prior to exposures. Exposed at f8 for 1/8th second assuming ASA of 400. Recent testing of lens shows this speed to be in the interval (126 thousandths, 129 thousandths). Separately tray developed each sheet of film in small glass Pyrex tray for 11 minutes 7 seconds at 72 deg using Zone VI compensating developing timer. Chemistry mixed fresh 1:1 for each sheet of film.
DATA: Small: 0.805, 1.010, 1.045. New: 0.90, 1.00, 1.22. Old: 0.905, 1.025, 1.19.
CONCLUSIONS: No differences found between batches of film chemistry. Average densities (and standard deviation) of three sheets of film for each batch are SMALL: 0.9533 (S.D. 0.1296), OLD: 1.0400 (S.D. 0.1431), NEW: 1.0400 (S.D. 0.1637). Pooled standard deviation is 0.1432, which is too high. (P-Value on one-way anova was 0.717.)
SUMMARY: I was very surprised, given the high variability, that OLD and NEW averages were arithmetically equivalent! This was not a mistake. Likely sources of variability I'm experiencing are shutter (will retest), tray development (probably!), perhaps film (I doubt it), meter (will have checked), or mixing (unlikely). Based on my results, I would not have a problem at this time using different batches of ID11 film developer.
NEXT STEPS: I need to improve this variability. It's too high, as much as an entire zone. It may be tray development, but I was extremely careful to be as consistent as possible. I hate to think that my 8x10 negatives could be subject to this kind of variability. Looks like I have a winter project. I may try this test again, after I improve variability.
1] My meter read dead-on before and after every exposure. My light source is on a track mounted on a tripod, so I can make very small adjustments. If the line on my Pentax Spotmeter V didn't line up EXACTLY to an EV of 10, I adjusted the light source. See previous post for other details on test pattern.
2] Developed each sheet in 20 English fluid ounces, so there was plenty of developer. About 3/4 inch in small tray, so film was always submerged. Chemistry was separately mixed for each development. Average temperature was kept within 1/4 degree, usually less. (Actual temperature within 0.4 degrees, usually much less.) I have an electronic thermometer accurate to tenths. I measured the developer temperature before and after, averaged these values, and compared to target temperature of 72 deg. Used a Zone VI compensating developer timer, so any difference in temperature from target was compensated. Any effect due to temperature was negligible with this device, since previous testing concluded that if I'm within one degree, compensated development time will be accurate to within 3 seconds. Agitation was as exact as I could make it. Lifted each side of tray 5 times at start and twice every 30 seconds. Lifted each side about the same height. All developers "seasoned" for at least 24 hours after mixing.
3] Measured densities using old MacBeth Quantilog densitometer. Stouffer's test strip w/31 steps shows this device accurate to within 0.015 D. Checked calibration before and after each measurement.
-- neil poulsen (firstname.lastname@example.org), November 04, 2001
I don't think I made it very clear that the test was to take 9 photos of test pattern, develop three sheets in each of three batches of ID11, and then measure the densities of each sheet to determine if there was a difference between the three batches of ID11.
-- neil poulsen (email@example.com), November 04, 2001.
I will be a Wienershnitzel Chili Cheese dog with triple onions that your variability comes from tray development. There is a reason JOBO processors are so highly valued by sheet film shooters and TMax users. I tray develop so I can use inspection as I watch the negs come to where I want during the process. It is not as precise as the JOBO negs nor are the negs as even in density across the skies and other single tone areas. But, you live with what you like & take the ups & down with it. Check out earlier statements by John Sexton on his change from trays to JOBO processor and if you don't develop by inspection, and if your film size isn't so big the expert drums aren't made for it, take a good look at the JOBO processors. If you shoot TMax be aware it was developed using these processors as a working development tool. No, I don't sell JOBO processors, just have some experience with them and development anomalies are much less when using them. There are some advantages the the mechanical repeatability & precision we just can't get processing film by hand.
-- Dan Smith (firstname.lastname@example.org), November 05, 2001.
I can see where the Jobo would get improved results. John Sexton's comments were interesting in this regard. But, he also said in the same comments that tray development provided an excellent means of developing sheet film, and that he painstakenly used this approach for several years. Ansel Adams, in his books, cites tray development as his favorite approach. It strikes me that they wouldn't have opted for this approach if it gave inconsistent results, my obvious problems notwithstanding.
Anyway, until such time as I could obtain the Jobo system, I need to work with what I have.
One reason I offered the results of the above testing is the perspective they offer on whether or not there might be significant batch to batch variability in ILford ID11. I suggested this possibility in another post, but this data would imply otherwise.
-- neil poulsen (email@example.com), November 07, 2001.