Notice that I said "at that concentration" it is irrelevant. If you add sulfite to Rodinal it changes the characteristics of it. Did I say it didn't? The OP has store bought Rodinal, not mixed from scratch. Sulfite at low concentrations has very little to do with development in a developer such as Rodinal that relies on pH to attain activity.
If you mix "Rodinal from scratch" you don't have Rodinal now do you? You have something that may be similar in characteristics to Rodinal, but has been modified to suite your requirements by adding additional sulfite. The sulfite is secondary (as a silver solvent) to the Hydroxide. If you doubt this try mixing your rodinal substitute without the hydroxide, just the sulfite, and see how much development you get.
I'll leave it at that.
I think we're talking at cross purposes. I am not talking about adding sulfite to Rodinal.
As part of my own investigations into the "solvent effect" I used gas chromatography-mass spectrometry to analyse three batches of Rodinal to obtain an exact formula for the commercial product. I also measured the specific gravity, pH to four decimal places and a couple of other useful parameters. I obtained most of the components; two I had to synthesize (one I am not even sure I know what it does) because they were not commercially available here. Finding the potassium salts of some common agents which are usually available here as the sodium salts was not as hard as I thought it might be.
I mixed up batches of Rodinal so that the developing components were as they would be in Rodinal 1+25, 1+50 and 1+100. I played with the sulfite so that for each of the three concentrations I had the sulfite of the other two, but with everything else held constant. This was hard, because the hydroxide gets used up a little reducing other components in the developer, but this differs at different dilutions because of shifts in the acid-base eqilibrium and the effect of this on the reducing capacity of the solution. I also mixed up one batch so that there would be no sulfite.
I then developed four short films exposed at half their rated ISO and an Ilford T5 process control test strip using gentle intermittent agitation.
So in the end I had 36 films and 12 test strips developed in Rodinal as follows:
1+25 normal sulfite
1+25 sulfite equivalent to 1+50
1+25 sulfite equivalent to 1+100
1+25 no sulfite
1+50 normal sulfite
1+50 sulfite equivalent to 1+25
1+50 sulfite equivalent to 1+100
1+50 no sulfite
1+100 normal sulfite
1+100 sulfite equivalent to 1+25
1+100 sulfite equivalent to 1+50
1+100 no sulfite
I photographed each frame with a microscope with a scanning back camera at three magnifications representing large, medium and fine image detail and compared the image files using an image analysis program written for this purpose in Zeiss' KS Run format.
I measured the test strips using a densitometer and used ANOVA to compare the curves and the ISO.
I had to immediately remove all the films and test strips developed in Rodinal with no sulfite from the analysis; this formulation caused rapid infectious development that stops just as quickly. The image was thin but very contrasty, way outside any sort of curve shape that could be ISO rated. Further experiments showed that this could not be fixed by extending development. I found that the curves and effective ISO speed had no statistical differences.
I used a Bray-Curtis dissimilarity transformation and generated a non-parametric multidimensional scaling analysis (nMDS - non-metric MDS both finds a non-parametric monotonic relationship between the dissimilarities in the item-item matrix and the Euclidean distance between items, and the location of each item in the low-dimensional space using isotonic regression) and found that the grain assessments grouped out separately from each other. This means that the grain patterns and edge rendition were all statistically different from each other among and between the groups - essentially this means that the grain size and pattern and edge rendition differ both at a given concentration and between dilutions, but not the film speed or curve. The difference at a given sulfite concentration was very interesting, the dilution effect was something I expected.
I then wet printed examples across the samples on Fomaspeed Variant III and developed them all with standard time and agitation in Dektol 1+2. You can see differences in edge rendition and grain patterns.
I then developed films and test strips in commercial Rodinal 1+25, 1+50 and 1+100 and obtained results that did not statistically differ from the previous experiment. This confirmed that my lab-brewed Rodinal is very close to the commercial product.
So what does sulfite do? It preserves the 4-aminophenol as it reduces exposed silver, allowing dilute Rodinal to develop silver effectively. This is reflected, visibly and measurably, in sharper edges and more highly clumped grain as the sulfite concentration is lowered. Why? I guess because local exhaustion occurs more quickly when the 4-aminophenol is less well preserved by sulfite. I have other data that shows that the effect of modifying agitation is about twice as apparent as changing sulfite concentration, but in normal developing, this means you can change the grain and edges more with more dilute Rodinal because the times are so much longer, giving you more opportunity to modify the agitation relative to the total development time.
Sulfite also is largely responsible for Rodinal's very long shelf life.
Next I am moving on to experimenting with ascorbate aminophenol formulae and what happens when you replace sulfite with ascorbate as Patrick Gainer recommends. I am also going to experiment with buffered 4-aminophenol developers that operate at less alkaline pH. I am also going to compare Rodinal with some other developers to see if all that acutance Rodinal is famous for is really there.
I'd like to try to publish this, but I don't think I can legally disseminate the Rodinal composition data. I'm looking into that right now.
Marty
