That rule of thumb can work vaguely, but only for ONE developer. If you have a group of them, you can't apply it across the group, nor can you apply it across films. Oh well. I'm ready to give up on this subject.
There was nothing vaguely vague about any of Dr. Collett's rules of thumb. There are two constants in the equation usually applied to film and developer, which is presented along with a simple form of the Arrhenius equation. They allow for different slope and intercept on semilog paper graphs. Over any reasonable range of temperatures I would allow in my darkroom, the simple equation works very well. If you have only one experimental point, you can plot another one at 10 C higher and half as long, or 10 C lower and twice as long and draw the line between those points to get a very good guess at your next experiment.
After 15 years of doing the "real" work on this, I find that the Arrhenius equation is only a vague approximation due to a number of factors. The film, the developers, the ratio, the pH, the temperature ad nauseum....
All of these tend to distort the original assumption. And, they just don't always work. I have applied (or tried to) over 32 years and found exceptions in every case. Sorry.
I am sure you are right! However, I think another important factor here is how accurate people need to be, esp with the use of VC materials. If at 20 degs C my correct dev time is 10 mins and I use 11 at 19 degs (whereas the correct time should in fact be 10.75 or 11.5, it is neither here nor there for VC use.
Can you explain whether making a 10% change for each degree could be more wrong than making no correction. I know this sounds daft, but say at 20 degs C the time should be 10 mins and at 22 degs C I use 8 (whereas the correct factor dictates it should be 9.1, I have made a bigger error by applying correction than none at all.
I guess one needs to use teh provided slopes unless one is using a bespoke developer or one where such things are not supplied, such as pyrocat.
The OP was comparing developer A to developer B and referred to film. If you use one developer and one film, you may see what you refer to as temperature varies.
Consider the attached chart which I am reposting. Note that the slopes of the developers vary greatly and some are even the reverse of others.
This may completely upset things if you move across this set of developers. Note the DK-50 in particular, and how it has a vastly different slope than the other developers.
It's interesting that some people are so enthralled with items like the Zone VI compensating timer. Obviously, it does not know what developer is being used, and I assumed it used the Arrhenius Equation (I like to say that as I know how to spell "Arrhenius"...) to make it's time corrections. There's even a newer version of this type of timer being made that is software along with a stand-alone USB temp probe. Kind of funny.
Kirk
For up from the ashes, up from the ashes, grow the roses of success!
Well, looking at this from a chemistry standpoint, there are a lot of factors that affect rate, however, for each developer, you could conduct a densiometer study and determine two points on a graph of rate of development (measured indirectly as a gamma, in this case) vs. temperature, and then extrapolate other points on the graph from it, you can also get a lot of other chemistry data from this graph (you'll have to excuse me, I can't remember all of the data that this graph can help you with). After the AP's this year, I'm planning on doing a study on developing agents and their associated rate laws. I'll be happy to let you guys know how that all works out when I get there after the second week of may!
According to "The Theory of the Photographic Process", Third edition), the temperature dependence of the rate of development can be expressed roughly in terms of a coefficient, called k, which is the ratio of the rate at a particular temperature to the rate at a temperature ten degrees higher on the Kelvin scale or 18 degrees higher on the Fahrenheit scale. Among the factors affecting k is pH. It was not noted in the reference, but a plot of log(k) vs pH is for practical purposes a straight line between pH values of 7.8 and 10 and values of log(k) from .477 to .362. This region of linearity includes various compositions of Metol, Metol and hydroquinone, and catechol.
Here is a table of the data presented in the reference.
Developer Kbr pH k log(k)
Sease III 1.0 7.8 3.0 0.477
Metol without alkali 0 8.0 2.9 0.462
D-76 0 8.4 2.8 0.447
Gevaert 206 (MQ) 0 8.9 2.6 0.415
Agfa 14a 0.5 9.75 2.4 0.380
Metol-Na2CO3 0.5 9.85 2.3 0.352
The attached graph shows k vs. pH on semilog paper.
We can calculate the k for line C on the graph PE sent to be 2.2. D-19 is on that line. If the log(k) vs pH relationship stands, the pH of D-19 should be about 10.4.
Adv Reply
Originally Posted by gainer
