GVM1 - Discussion of a prototype glycin-phenidone film developer design

[grainyvision]

So I'm currently designing GVM1 and am on my 7th prototype test. It started as an intellectual curiosity since I often see references like "phenidone and glycin are mildly super additive, but there is no developer designed with using both". The results however are incredibly surprising.

Characteristics:

* Speed increasing, observed to likely be by 1-2 stops. Measured to be about 1/2 stop compared to D-76H on paper
* Excels at pushing. Contrast is extremely resistant to running away and becoming too high with over development. Shadows gain density much more quickly than highlights. Too much development actually tends to decrease contrast
* Intolerant to extreme over exposure. It can tolerate some, but over about 2 stops too much exposure tends to block up and turn extremely low contrast
* Abnormally low maximum density. This allows for good rendering of highlights both in scanning and printing with printing often giving great results at grade 2. Extremely compensating
* Extremely fine grain. Sharpness is poor to moderate, though the way the grain is shaped tends to result in the appearance of more resolution, as if you used a film which was slower speed
* In controlled lighting portrait tests, can be an incredibly good portrait developer rendering white skin tones beautifully
* Has unique aging progression. Is brown when freshly mixed, but turns clear and green upon usage or aging.

Questions:
* Has anyone ever messed with this combo of developing agents and just not talked about it? I believe there is FX11 which uses a 3rd agent, but it also contains much more sulfite and had completely different aims. I've not seen anymore messing with this kind of formulation at all in the hobby space.
* How is this so fine grain while containing so little sulfite? I know TEA is a silver solvent, but I've used much more TEA in much coarser grained developers
* What is up with the way this developer oxidizes? It turns green! If I add sulfite, no change in color occurs. If I add ascorbic acid, the color changes to pale yellow. When the developer is green it is still active. I've let it sit overnight in open air once (diluted) and it was very slow but could produce some density on fogged paper.
* Since this is so low sulfite, I expect shelf life to be very poor. However, it at least doesn't die overnight. Any experience with very low sulfite glycin based developers?
* The only thing preventing this from being more concentrated is that the sulfite falls out of solution. Metaborate and phenidone both are much more soluble in glycol than water, hence why I use it. Glycin is also very soluble in TEA, however TEA, glycol, and metaborate make sulfite less soluble. I'm unsure how I would even split this into a 2 part formula to solve this problem due to how glycin is less soluble in less alkali solutions. I'd really like to make this a 1+19 formula if possible. Any ideas? In formulations with similar problems potassium sulfite did not solve the problem. I'd like to make it more concentrated as I've typically seen more concentrated solutions absorb less oxygen leading to better shelf life, and plus 1+14 is an awkward dilution rate.

Prototype formula (note: this is not in mixing order, as that is still somewhat complex) #7:

* To make 150ml of concentrate, which mixes to make 2L of working solution. 75ml of concentrate per 1L of working solution
* 32ml triethanolamine 99% (TEA, likely low freeze grade. Sourced from photographer's formulary)
* 40ml propylene glycol
* 55ml water
* 2.8g sulfite
* 0.16g dimezone-S (it is unknown if phenidone can be substituted)
* 1.6g glycin
* 14g sodium metaborate
* 0.15g potassium bromide
* 1.4g potassium bicarbonate (1.2g of sodium bicarbonate can likely be used instead)
* Add ~3ml glycol to make 150ml solution

Example images from a test run at 10m, 76F.

ultrafine extreme 100@400

high resolution scan detail:

FP4+@200 (blue sky, so not much detail there unfortunately)

FP4+ detail:

Tri-X@800:

Tri-X@200 to show poor over exposure response:

I processed some HP5+ as well, though it was effectively pulled to about ISO 100 and was poor in contrast.

FP4+@200 portrait, this was with GVM1 #5 though which had less ideal contrast and was slightly more grainy due to over development

Appearance of developer after processing paper (no antihalation dye) vs appearance of concentrate:

Plot of curves made using Ilford MGV RC paper and step wedge. (horizontal axis is labeled step number on wedge). Grade 1 filtration. 70F. Prototype #5 is extremely similar to #7

 

[mshchem]

I played around with Glycin for print developer 30 years ago. Glycin has a tendency to revert to tar. Lovely when it's fresh.
My standard print developer is Bromophen, I love the stuff.
I'm afraid I'm not helping out much.
Good luck with your project.
Mike

 

[grainyvision]

I ran a series of experiments to try to figure out the oxidation chemistry involved here. I don't have an analytical lab or anything, so my means are very limited.

The green color can not be reproduced by dimezone alone. The conditions needed:

* pH between 10-12
* Water present
* glycin present
* dimezone-s present
* Low or no sulfite

Thus, this is not a reaction involving borates or TEA. The pH here is extremely important. With a deep green solution, I tried:

* Adding sodium hydroxide -- this caused the solution to go from green to orange (like the starting solution color), but then with stirring and additional water (to add oxygen), the solution eventually became a stable blue
* Adding citric acid -- this caused the solution to revert to the orange starting color. Adding alkali would cause the solution to again turn green
* Adding sulfite caused overall color to lighten and with hydroxide addition would cause the blue color to not appear. Adding a small amount of sulfite would not remove all of the color. A large amount does.
* Adding ascorbic acid (while keeping the solution alkali) causes the solution to go almost completely clear or a very light orange
* Adding glycin would cause the solution to revert to the starting orange color, though would quickly oxidize again becoming green (note: used more alkali to counteract the fact that glycin is acidic, but also used only small amounts of glycin)
* Using potassium bicarbonate as the alkali could not produce the green color

Even when deep green the solution would still produce density on fogged film. When adding bromide, the solution would only barely produce density

I really have no idea what is happening here. I can confirm because glycin additions will cause the solution to revert in color, that glycin is capable of reducing the green phenidone oxidation product. However, even a relatively large amount of glycin will be quickly "spent" by a small amount of phenidone, more than would be explained by dissolved oxygen reacting with the phenidone. However, the reaction seems reversible with either higher or lower pH, or with more glycin, so it is unlikely to be producing peroxide I would think.

The green solution is extremely deeply colored, but seems to wash out of any material without issue.

What is happening here is kind of a complete mystery, but I believe it may be responsible for the unique action of GVM1. It might be that the green oxidation product is actually a developer in itself, maybe dimezone oxidation products react with glycin to form something else and not "just" regenerating dimezone but something else which can explain the unique development action and solvent effect

 

[Alan Johnson]

It would be simpler to replace the metol in FX-2 by 1/10 th its weight of dimezone.
https://web.archive.org/web/20040421084832/http://jackspcs.com/fx2.htm
You have 3 different alkaline buffers in your brew,4 if you count sulfite.

 

[ruilourosa]

Characteristics seem a bit over the top...

Although Crawley FX 11 seem to match some of your claims to your experimental Dev although with a lot more sulfite and with hidroquinone...

Why are you using 3 alcalis? Metaborate should be enough. Why not just diluting on plain water? It wont last long but making a low concentration concentrate right up in the beginning is a bit strange.

What was your dev procedure?

Thanks

 

[Raghu Kuvempunagar]

Is this meant to be both film and paper developer?

 

[grainyvision]

It would be simpler to replace the metol in FX-2 by 1/10 th its weight of dimezone.
https://web.archive.org/web/20040421084832/http://jackspcs.com/fx2.htm
You have 3 different alkaline buffers in your brew,4 if you count sulfite.

Characteristics seem a bit over the top...

Although Crawley FX 11 seem to match some of your claims to your experimental Dev although with a lot more sulfite and with hidroquinone...

Why are you using 3 alcalis? Metaborate should be enough. Why not just diluting on plain water? It wont last long but making a low concentration concentrate right up in the beginning is a bit strange.

What was your dev procedure?

Thanks

I believe glycin and phenidone is a completely different animal than glycin and metol. I actually have a glycin-metol developer, GVK1 which is pretty great https://grainy.vision/blog/gvk1

I use TEA as kind of a signature kind of thing, but also because it does several things. It is an alkali buffer of course, but it does much more:

* Unique silver solvent when kept at alkali pH. Reaction of silver nitrate with TEA will first cause a reaction producing a brown discolored product, which then promptly dissolves. Over several hours this product will fall out of solution producing a silver mirror on glass
* Reducing agent and potential developer accelerator. The most famous example of this is with RA-4 developer. I've personally observed that just matching alkalinity does not produce the accelerant effect that TEA gives in a home mixed RA-4 developer
* Preservative. TEA can be oxidized and so it functions as a weak preservative.

I have a feeling the reactions with TEA are much deeper than would be expected. Regardless, I might try to eliminate it just to see what happens. My expectations would be a significant decrease in development rate, and much coarser grain. Additional sodium metaborate to this developer nor removal of bicarbonate does not cause significant activity increases. In fact, reducing the amount of TEA and increasing metaborate to compensate has already been seen to reduce the speed increasing effect and to decrease shadow density

I used glycol because metaborate is actually much more soluble in glycol than water, as is phenidone/dimezone. I don't really like mixing on demand so I try to create formulations that can be concentrated. I've observed that almost always the higher the concentration the less problems keeping it on the shelf.

My dev procedure is just using a stainless steel tank and I developed 4 films at a time for this initial test. Temperature was measured to be 76F. Room temperature was 70F so the solution may have cooled a bit during processing, but usually this effect is negligible. Time was 10 minutes. I did agitation for 60s initially and then 8x every minute after. I use a running water as the stop bath. TF-4 fixer for 4.5 minutes.

Curves were made using paper in the dakroom and while they should give a feeling of how the developer behaves, I wouldn't say is 100% similar to actual film tests.

Is this meant to be both film and paper developer?

Only film. I only use paper for plotting curves as it's a simple way to get a guess as to how a developer might work on film without running through tons of expensive film and doing tons of subjective tests that can be difficult to judge.

 

[Alan Johnson]

OK here's my peer review re prototype formula 7 that you have now explained.
(1) Leave out propylene glycol, it is not necessary,the solubility of sodium metaborate in water is 16g/100ml 20C
(2) Leave out TEA. It was the basis of Gainer's PC-TEA which is both more grainy and gives lower speed than Xtol, very likely due to lack of sulfite.
(3) Try to get potassium sulfite which is much more water soluble in water than sodium sulfite and use it at high concentration, this is what some proprietary developer concentrates are believed to do.
(4) The oxidation product of phenidone sticks on the grain surface and gives the low contrast that you see. See the Film Developing Cookbook. Likely the contrast can be changed by adjusting the phenidone [dimezone?] concentration.
(5) Leave out the potassium bicarbonate.Instead adjust the pH to give a reasonable development time with metabisulfite.
(6) Get a decent method of evaluating granularity. Fine grain huh, no evidence.

 

[grainyvision]

OK here's my peer review re prototype formula 7 that you have now explained.
(1) Leave out propylene glycol, it is not necessary,the solubility of sodium metaborate in water is 16g/100ml 20C
(2) Leave out TEA. It was the basis of Gainer's PC-TEA which is both more grainy and gives lower speed than Xtol, very likely due to lack of sulfite.
(3) Try to get potassium sulfite which is much more water soluble in water than sodium sulfite and use it at high concentration, this is what some proprietary developer concentrates are believed to do.
(4) The oxidation product of phenidone sticks on the grain surface and gives the low contrast that you see. See the Film Developing Cookbook. Likely the contrast can be changed by adjusting the phenidone [dimezone?] concentration.
(5) Leave out the potassium bicarbonate.Instead adjust the pH to give a reasonable development time with metabisulfite.
(6) Get a decent method of evaluating granularity. Fine grain huh, no evidence.

1) glycol is definitely necessary. If not for the metaborate, then for the glycin
2) I tested this. Although this formula might go on to not include TEA, it functions as a grain smoothing agent in this formula in ways which I can't replicate with other adjustments without radically modifying contrast etc. It smooths grain in ways that I'm unsure how to describe. I don't have instruments capable of measuring it scientifically, but the visual impact is fairly obvious. Also note, with TEA removed, carbonate is needed in order to bring pH high enough to function properly. Additional metaborate and removal of bicarbonate will not give proper pH
3) I have some, though I only rarely use it. I decided to use it in the #9 prototype shown below
4) Increasing dimezone/phenidone concentration will cause shoulder (highlight density) to increase faster and block up. It is not really that trivial of a thing to adjust. Increasing glycin amount has not been tested enough to know what it does to contrast.
5) I ended up removing it and making the solution significantly less buffered
6) Any ideas? Other than very high resolution 100% crops, I don't have precision tools capable of measuring sub um range, nor an electron microscope to define the grain more specifically.

I am not concerned with "actual" grain size or anything like that. I'm talking about the appearance of grain in subjects. Rodinal does not greatly increase the size of grain in reality for instance, but the way microcontrast works with rodinal, it amplifies grain's appearance.

Anyway, here is a 200% (40MP) crop comparison between without TEA vs with TEA:

These are equal magnifications. Although it is difficult to measure grain here, and density is not perfectly matches by nay means... but the difference to me is pretty obvious judging this by eye. The first one is much more randomized and "pokey" and there is evidence of very minor edge adjacency effects. Meanwhile, the second one, though grain does not look smaller is simply smoother in terms of contrast and grain randomness.

Anyway, here is the formula which I used. I'm unsure if I actually like it very much.
The contrast curve looks great on paper, but on actual film the developer is slow (in terms of time), and produces results which are much less attractive. The contrast curve by eye looks softer and with more toe density development and a speed increase, but only with 33% more development time. I'd like to try this developer at 15+500 and 20+500 to see if I can get a return to those striking brilliant results I loved from prototype #7 and if the softer grain look can be reproduced without TEA

Note it has a different dilution and concentration:
GVM1 #9 -- note: these are ingredients, not mixing instructions
* 3.6g potassium sulfite / 8ml of 45% solution
* 0.16g Dimezone-S. 16ml of 1% in propylene glycol. NOTE: the glycol is absolutely essential here to dissolve the glycin
* 1.6g glycin
* 2g sodium metaborate
* 6g potassium carbonate
* water to make 50ml
* Usage: add 10ml to 500ml of water to make 510ml of working solution
* Test usage, 12m at 76F produced about box speed for Tri-X and ISO 200 for Ultrafine Extreme 100

Tri-X@400 with orange filter

Ultrafine Extreme 100 @ 200

200% crop detail:

Ultrafine Exteme 100 @ 200 (no orange filter used, cloudy day)

ultrafine at 200

Edit: Oh yea, also weirdly enough this developer after usage is a stable blue, and when concentrated is very deep orange brown. I left some concentrate of similar design in an open beaker overnight and no color change to indicate oxidation. I believe this has potential despite the low amount of sulfite to be fairly shelf stable. It at least has a very obvious indicator if it does go off of turning blue.

 

[ruilourosa]

You are going through the hardest route...

Dont add to a developer, remove if not necessary...

Why not just use tea instead of having just more alcalis in? Adjusting time..
The seen effect of tea may be just because you will have lower ph...
The silver solvency of tea should be negligible.

I think experimentation is to complement but i would start with the basic things. Glycin, phenidone, in various ratios, and a basic setup of one alcali, tea, metaborate or borax, and one preservative, ss or ps,

You could even try to split it into a divided dev....

 

[Alan Johnson]

(2) I accept that your results show that finer grain is obtained with TEA. But this is not the way it is done commercially, what they do is to use a near concentrated solution of potassium sulfite. OTOH potassium sufite is hard to get.
In FX-2 Crawley chose to go with a crisp sharp grain rather than a fine grain. The concentrate GSD-10 [Jay DeFehr] was similar but less stable. Choose a sulfite or TEA concentration that gives grain you are happy with.
(4) GVM #9 is like the FX-2, GSD-10 but with added dimezone it should work at a lower pH and presumably give finer grain, choose a satisfactory dimezone concentration to give good tonality. You may be aware that dimezone-s replaced phenidone because phenidone is more susceptible to hydrolysis in alkaline solutions.
(6) Agreed, the 200% crop comparison clearly illustrates what is said and is good for the purpose of comparing granularity.
Good work, appears you are already well on the way.

 

[grainyvision]

You are going through the hardest route...

Dont add to a developer, remove if not necessary...

Why not just use tea instead of having just more alcalis in? Adjusting time..
The seen effect of tea may be just because you will have lower ph...
The silver solvency of tea should be negligible.

I think experimentation is to complement but i would start with the basic things. Glycin, phenidone, in various ratios, and a basic setup of one alcali, tea, metaborate or borax, and one preservative, ss or ps,

You could even try to split it into a divided dev....

I basically had to remove TEA to make it more concentrated. Metaborate alone is not capable of getting the developer to high enough pH. I use metaborate for additional buffering which seems beneficial for this developer. I removed the bicarbonate, though I'm unsure if this is a good idea

The solvent effect of TEA is fairly well proven, but it is a subtle thing hard to exactly describe. I've done a lot of chemistry tests with TEA because I had 4 gallons at one point due to most vendors selling low freeze grade and not pure TEA. It also has development moderating effects, it can function as both an accelerator and restrainer depending on the developer due to interactions with radicals

GVM1 is not the first time I've messed with this combination of developing agents, however, previous experiments resulted in poor contrast and high fog. The primary requirement to avoid highlights running away as well as fog is to use a very small amount of phenidone/dimezone as well as a restricted amount of sulfite. Sulfite seems to allow faster recycling of phenidone and less highlight restraining. Contrast can still be hard to increase in this developer design without highlights running away though.

(2) I accept that your results show that finer grain is obtained with TEA. But this is not the way it is done commercially, what they do is to use a near concentrated solution of potassium sulfite. OTOH potassium sufite is hard to get.
In FX-2 Crawley chose to go with a crisp sharp grain rather than a fine grain. The concentrate GSD-10 [Jay DeFehr] was similar but less stable. Choose a sulfite or TEA concentration that gives grain you are happy with.
(4) GVM #9 is like the FX-2, GSD-10 but with added dimezone it should work at a lower pH and presumably give finer grain, choose a satisfactory dimezone concentration to give good tonality. You may be aware that dimezone-s replaced phenidone because phenidone is more susceptible to hydrolysis in alkaline solutions.
(6) Agreed, the 200% crop comparison clearly illustrates what is said and is good for the purpose of comparing granularity.
Good work, appears you are already well on the way.

I typically do not like the appearance of high solvent developers such as D-76. It has a particular grain structure which I find unattractive, even if it is more subtle. This is why I often restrict sulfite in my own formulas. Of course, building a stable developer with low sulfite is always an extra challenge though.

Yes, potassium sulfite is unfortunately expensive and only available from speciality photo chemistry places that I've seen and only in solutions. I buy mine from photographer's formulary. At least the amount I'm using here is very small. It might be possible to substitute sodium sulfite in this formula, but I haven't tried it.

I prefer sharp grain without it being "pokey" if you know what I mean. So while I prefer non-solvent developers, they have to build contrast in a particular way that doesn't make the grain look gritty or obtrusive. Without mixing FX-2 myself though I can't say too confidently how it compares. Unfortunately I've found very few example images of it. Personally I think even an extremely fine grain, or smooth grain, or sharp grain developer is ultimately worthless though if the contrast curve causes it to produce ugly renditions.

The pH of this solution is surprisingly high. On the low end, pH 10.5 and on the high end 11.5, depending on which exact formula. For GVM1 #10, using 20ml of concentrate in 1L of water produces a pH of 11. Using 40ml produces a pH of 11.5.

This is why I chose to use Dimezone-S instead of Phenidone... that and my dry phenidone is 2 years old and is at least suspicious at this point. I've definitely noticed that Dimezone-S solutions stay alive longer without much precaution.

I will post results and formula of #10 once I run it through a film test. The curves are promising and it is more concentrated than I had thought such a developer was capable of being. Specifically I slightly increased glycin, slightly increased metaborate, and significantly increased carbonate. I also added a tiny bit more dimezone solution.

 

[Alan Johnson]

Your results may be compared with those for Jay DeFehr's GSD-10:
https://www.flickr.com/groups/1447120@N21/
This is classified by Flickr as an inactive group, there are not many users of glycin based film developers.

 

[ruilourosa]

Well... I have not used glycin extensively just on fx2 and fx16, and with tabular film and difusion enlargers i much prefer either the revealing clarity of fx1 or the allroundness of diluted d76 or the vestals dd76 with a kodalk b bath...
For grain effects... I learned to use half frame cameras.
Glycin life expectancy is discouraging to long term experimentation... As is its price...

 

[grainyvision]

Your results may be compared with those for Jay DeFehr's GSD-10:
https://www.flickr.com/groups/1447120@N21/
This is classified by Flickr as an inactive group, there are not many users of glycin based film developers.

I've been tempted to make a batch of GSD-10, but ultimately it's characteristics seem a bit undesireable to me. I've formulated a similar glycin-only developer called EXG1 (renamed GVG1 recently) except without sulfite and using TEA as the solvent for glycin, split into 2 parts (A was 10% Glycin in TEA, B was 20% potassium carbonate). It produces excellent negatives for scanning and alt process work including lith printing, but is difficult to wrangle in with regular darkroom printing due to an expanded negative density range. GSD-10 looks to have the same properties, except without the excellent pushing ability and speed increase I get with EXG1. I tend to not prefer stand development either, excluding with Rodinal. I ultimately gave up on EXG1 as a standard "house" developer because of these issues as well as unpredictable shelf life and generally growing very tired of heating up TEA to very high temperatures to use it as a proper solvent. GVM1 has revealed that as long as the glycin is neutralized, then it is very easily soluble in glycol. Definitely a trick I may make use of in future formulations.

 

[grainyvision]

Well... I have not used glycin extensively just on fx2 and fx16, and with tabular film and difusion enlargers i much prefer either the revealing clarity of fx1 or the allroundness of diluted d76 or the vestals dd76 with a kodalk b bath...
For grain effects... I learned to use half frame cameras.
Glycin life expectancy is discouraging to long term experimentation... As is its price...

Literally just posted the previous message lol. I really like how glycin based developers tend to build grain. It's not "fine" grain I'd say, but it's smooth in a very particular way.

I keep a small stock of glycin at room temp (small enough that I'll use it within a year) and otherwise keep my glycin frozen. By the time I run out of room temp glycin and need to refill, the frozen stuff is a noticeably lighter color, so freezing is I believe a reasonable way to keep glycin long term. I've seen no color change thus far with my frozen stock. Yes, it is expensive, but only relative to other developing agents, and often less is required than other developing agents. I priced out a previous prototype which was $0.71 per liter of working solution, so I think even if glycin itself is expensive, developer is still cheap compared to film or lab processing. The biggest problem with glycin is that it is not easily available in many countries

 

[ruilourosa]

I just tend to find the results i want on simpler to reach devs...
I have metol until forever...

 

[grainyvision]

Just for reference, definitely always freeze your glycin! One year at room temp glycin vs opened, but frozen since purchase except when used to refill the small bottle. I'm only keeping 50g at room temp now at a time instead of 100g

 

[ruilourosa]

I have been reading the new edition of the film Dev. Cookbook and Crawleys FX 20 can be a very good start...

 

[JPD]

That's an intersting project. Have you tried Agfa 8 with Glycin as the only developing agent? In my simple test I found that it's sharp as Rodinal and has excellent mid-tones. Looks best if you give the film one stop more exposure.