I believe that Kodak for many years submitted prints to a group of people and had them rate the relative grainess of films in this manner. This whole topic is very subjective and even using RMS granularity is subject to criticism.
A rock pile ceases to be a rock pile the moment a single man contemplates it, bearing within him the image of a cathedral.
~Antoine de Saint-Exupery
We evaluated grain and sharpness both, and if color we evaluated color balance and fidelity. I was a participant in these tests from both sides. I took the test and I gave the test. Usually, we had hundreds of prints and hundreds of participants.
Although I've been posting little, I'm still doing experiments and dialing-in the numbers on a concentrate that gives XTOL-like quality. Speaking of XTOL, the graph below is two identically exposed successive frames on the same strip, developed at the same time in XTOL. These were TMY2 for 6.5 minutes at 20C, with temperature held constant by keeping the tank in a water-bath:
I expected these curves to be nearly identical. Any idea why they diverge so much? Perhaps due to differences in fluid-flow in different parts of the tank?
The shutterspeed should be consistent, as I measured it on both sides of the film-plane with my shutterspeed tester. The OM-1 consistently got 16.6 ms for 1/60th on both sides. So I doubt the differences are due to exposure. Fluid-flow is all I can think of, unless somebody has a better idea.
Anyway, the graph below shows the straighter XTOL curve from above, with my developer (in green). I'd say I need a bit more density, except the curves converge at the right. Odd.
The one-litre formula for the green curve is:
Sodium sulfite ......................... 90 g
Ascorbic acid .......................... 9 g
Sodium metaborate 4 mol ...... 6.2 g
Dimezone S ............................ 0.2 g
Target pH is 8.3 (same as XTOL).
Use same times as XTOL.
Grain looks identical to XTOL in my 22x loupes. Except for the sulfite, this formula can be dissolved into 24 ml of propylene glycol, yielding a concentrate that can be used at 1+29 dilution. I'm still tweaking this formula. By the way, I found that Phenidone can be substituted at half the weight of Dimezone S, so the above formula would use 0.1 g of Phenidone.
Mark, for what it's worth, in comparison to the various other general purpose solvent developers I've tested, in densitometric tests I've consistently found XTOL (at stock and various dilutions) to show both the most variability roll to roll and even frame to frame, and choppier overall density growth (curve shape).
Michael, this makes me ask, "Then why am I trying to imitate the stuff?" Your posting also made me wonder about the choppiness of my developer, so I plotted the last few frames of the prior test-roll. I was tweaking the formula, so density varies some frame-to-frame. Here are five frames:
Originally Posted by Michael R 1974
My developer has little choppiness; certainly less than what I saw with XTOL. The linear areas are reasonably straight. I don't know why choppiness would differ, unless it's because I'm using ascorbic acid instead of ascorbate, and over 50% more borates than XTOL, and thus I probably have better buffering.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
Hi Mark. I just thought my testing results might help explain or support the inconsistencies you've encountered with XTOL. I can't explain it though. My knowledge of photochemistry comes mostly from reading Haist, Crawley, Troop and a few others, as well as my own experience and testing. But I'm no chemist. This thread has made me want to revisit my XTOL tests though. Regarding buffering, I'm not sure how this would affect things since some of the most predictable and consistent/repeatable results I've experienced have been with both buffered and non buffered developers. XTOL has always been somewhat of an anomaly for me, and the only thing unique to it versus all the other developers I've tested is sodium isoascorbate as a developing agent. Actually it is the only developer I've ever tried with C (or variants). Not that this explains things either.