Combining push processing and compensation?

Helge · Nov 12, 2020

I’m sure I’ve seen something about this somewhere but lost the link.

I know to some, it will sound like pure alchemy and sacrilegious.
But.
The thinking goes like this:
Push processing raises contrast.
It should be possible to control and lower that contrast, while still getting the “speed advantage”(*) of pushing (I’m fully aware that some people want the higher contrast).

This could be done by starting push development, and then at some predetermined point replace the normal developer with a highly diluted one, or otherwise compensating one.

This would of course theoretically continue development of any underdeveloped shadow detail, while leaving highlights alone for the most part.

My questions is: What is this technique called and does anyone have links to any online recourses or know of books where it is explained in detail?

Pre-development is relatively well known in C-41, but “post-development” doesn’t really return anything useful on Google.

Edit. (*) “Speed advantage” in quotation marks. I’m fully aware it’s not real speed and you’re basically just skewing the curve with the foot as a “hinge”. Trick would be to “unhinge” the foot, so to speak.

Paul Howell · Nov 12, 2020

The old saying is to expose for the shadows develop for highlights. When pushing, not the same as expanding or contracting contrast via Ansel Adams, you lose shadow details. as you are exposing for highlights. When I was a working PJ our when pushing we use to day expose for the highlight and let the shadows fall where they may. To control contrast use VC paper.

Lachlan Young · Nov 12, 2020

@Helge I think you're describing the sort of water-bath techniques that Ansel Adams promulgated - and which won't work as well with more modern multi-emulsion films.

What you really want is to be able to affect the emulsion design, such that it behaves like Delta 3200's tone curve in fairly active/ solvent developers, which gives a steeper shadow gradient and a much softer highlight gradient. Xp2 Super does something similar too - without any extended techniques etc.

dpurdy · Nov 12, 2020

I once badly underexposed some 8x10 film using a model with dark hair against a dark background. I had forgotten my flash meter and there was lots of ambient light giving the illusion that the background was not terribly dark. After test processing the first sheet in pyrocat I saw my mistake. I decided to make it a bit more dilute and then do long stand development... like 45 minutes. Theoretically that would compensate some while pushing. It didn't take the density back to what I was originally after but it did make printable negs with a fantastic silvery skin tone (nudes) and a slight separation of her hair from the background.

138S · Nov 12, 2020

Helge said:
My questions is: What is this technique called and does anyone have links to any online recourses or books where it is explained in detail?

There are several techniques for doing that:

> Diafine, it pushes shadows and it compensates highlights from component exhaustion. Of course two bath developer do the same.

> Stand/Semi-stand/EMA (https://www.powerofprocesstips.com/). It is a mith that Extreme Minimal Agitation is low contrast, you may overdevelop/push what you want, but highlights will be compensated to a lower density from higher local bromide concentration working as a local restrainer, compensating effect in the highlights comes well more from non evacuated bromide than from local agent exhaustion.

> SLIMT (Selective Latent Image Manipulation Techniques) http://www.davidkachel.com/assets/cont_pt3.htm. You may perform a contrastwise bleaching prior a pushing development, in that case you limit the contrast from the pushing development while conserving the "pushed" speed".

I've been experimenting with those 3 techniques for night photography, in that situation we have more LIRF in the shadows and a risk of excessive contrast...

Donald Qualls · Nov 12, 2020

Simplest way I know to gain speed in the shadows without running the contrast too high for a #1 contrast filter: take Rodinal (or equivalent -- I learned to do it with Parodinal), dilute 1:50 to get a nice long process.

Multiply the Massive Dev Chart time by 1.5 -- that is, if it should be 14 minutes, you'll process for 21. Agitate continuously first minute, as usual, then five inversions every third minute.

The combination of high dilution and low agitation will give compensation, reining in the highlights, while the long process time will get everything possible out of the shadows. You aren't leaving the film standing long enough to develop edge effects that people talk about from leaving the film to stand for an hour with no agitation; the negatives (in my experience) will look completely normal. You'll gain from 1/3 to 2/3 stop of true speed.

If you don't like Rodinal, you can do the same with HC-110 1+119, D-76/ID-11/Xtol 1+3, most likely Dektol 1+14 or 1+19 (you may have to extrapolate times for Dektol) -- pretty much anything where you can find times for high dilution. If you use a developer with phenidone (Xtol, for instance) you may gain more than 2/3 stop.

Bantam RF, Ultra 100 (828 cut from 120), Parodinal 1:50.

ic-racer · Nov 12, 2020

My tests show about one stop difference between a SLOW and a FAST developer formula. So, depending on your starting point on the spectrum of developer formulas, one could gain up to a whole step moving to a 'high-speed' developer. Otherwise, if one already uses a high speed formula (for example I use T-Max Developer) there is not much to be gained.
BTW: There us no standard developer for ISO speed tests. Manufactures can choose any developer they want for the test.

Helge · Nov 12, 2020

Thank you! Love you guys! ;-P
Don’t let this post mark the end of the thread though.
Some very useful pointers here so far.

Clarification:
As said in the edit of the OP, I’m aware pushing is not gaining real speed, but only pushing back the top of the curve.
Getting the foot to move back too and/or up is the whole point of many development gymnastics.

~Zone III Pre-flashing will help the film in recording “something” in the deep shadow areas.
That would be my “affecting the emulsion” that @Lachlan Young mentions.
So the hope could be to use some kind of compensation to give emphasis to those areas, hopefully avoiding just making mush of the negative.

Helge · Nov 12, 2020

Donald Qualls said:
Simplest way I know to gain speed in the shadows without running the contrast too high for a #1 contrast filter: take Rodinal (or equivalent -- I learned to do it with Parodinal), dilute 1:50 to get a nice long process.

Multiply the Massive Dev Chart time by 1.5 -- that is, if it should be 14 minutes, you'll process for 21. Agitate continuously first minute, as usual, then five inversions every third minute.

The combination of high dilution and low agitation will give compensation, reining in the highlights, while the long process time will get everything possible out of the shadows. You aren't leaving the film standing long enough to develop edge effects that people talk about from leaving the film to stand for an hour with no agitation; the negatives (in my experience) will look completely normal. You'll gain from 1/3 to 2/3 stop of true speed.

If you don't like Rodinal, you can do the same with HC-110 1+119, D-76/ID-11/Xtol 1+3, most likely Dektol 1+14 or 1+19 (you may have to extrapolate times for Dektol) -- pretty much anything where you can find times for high dilution. If you use a developer with phenidone (Xtol, for instance) you may gain more than 2/3 stop.

View attachment 259278
Bantam RF, Ultra 100 (828 cut from 120), Parodinal 1:50.

What you are describing is AFAICS classic semi stand. I’ve done that with great effect in Rodinal.

But it’s not really push development, in that you are actually loosing a bit of real speed.

One way I could think of doing it, would be to push develop in XTOL say, 75 percent of the time or even let development finish to the normal pushed time.
Then wash and replace with 1:100 or even higher diluted Rodinal and let that sit in stand for an hour or two.

Before I ruin a roll of film and the time used taking it in that experiment, would that be a very bad idea?

Lachlan Young · Nov 12, 2020

@Helge I'd strongly suggest trying a roll of Delta 3200 run at Ilford's recommended 'box speed' time in ID-11/ Microphen/ Xtol (g-bar 0.62) because it'll very clearly show the effect of pushed shadows and pulled highlights (essentially its curve gives an effective shadow gradient pushed to 0.7 + highlight gradient pulled to 0.5 = average gradient of 0.6) - you may be surprised at how strong the effect is. Otherwise, I'd suggest that rather than getting too deep into developer manipulation, fairly straightforward masking techniques can be used to very strongly adjust shadow and highlight contrast at the printing stage.

Helge · Nov 12, 2020

Lachlan Young said:
@Helge I'd strongly suggest trying a roll of Delta 3200 run at Ilford's recommended 'box speed' time in ID-11/ Microphen/ Xtol (g-bar 0.62) because it'll very clearly show the effect of pushed shadows and pulled highlights (essentially its curve gives an effective shadow gradient pushed to 0.7 + highlight gradient pulled to 0.5 = average gradient of 0.6) - you may be surprised at how strong the effect is. Otherwise, I'd suggest that rather than getting too deep into developer manipulation, fairly straightforward masking techniques can be used to very strongly adjust shadow and highlight contrast at the printing stage.

By printing time, the latent information in the shadows is lost.

Will try Delta 3200 at that time/contrast. Thanks.
Highly underrated film BTW.
P3200 gets all the praise. But I’d say D3200 has almost a stop of real speed on P3200.
No wonder it’s a bit grainier.

Developer being the weird and fickle amplifier it is, should probably be one of the main areas of research until/if RnD money gets back to film (I know this is old as the hills pussycat stuff compared to what the really heroic people have done).

I still have a fantasy about some kind of bimat developing with highly localized development done by mechanical means.
Sort of like chemical dodge and burn.

Donald Qualls · Nov 13, 2020

Helge said:
But it’s not really push development, in that you are actually loosing a bit of real speed.

Where do you get that idea? Yes, Rodinal is a speed losing developer (compared to the standard ones, selected to give the highest possible speed at standard contrast), but speed is measured at the toe -- minimum light exposure to get (IIRC) 0.2 above base + fog. Further, this isn't really semi-stand -- that would be 30 minutes at 1:100 with agitation at start and at 15 minutes. I think of the 3 minute cycle as "reduced agitation" and based on toe speed, I'm pretty sure this is the sweet spot; it surely does gain shadow detail over standard (one minute cycle) agitation, and gives less contrast in a "push" as well.

Helge said:
Sort of like chemical dodge and burn.

The problem with that is would pretty much have to be applied by inspection, because you'd have to partially develop the film to know where to "chemical dodge and burn" in the first place. So you want to give partial development, inspect (under that super-dim green safelight, or via infrared, preferably with desensitized film in either case), then decide where to make irreversible changes to the negative. Seems like a bad idea to me.

Now partial development in one developer, followed up by partial development in another, is legitimate, if probably a case of extra work for little or no gain...

138S · Nov 13, 2020

Helge said:
As said in the edit of the OP, I’m aware pushing is not gaining real speed, but only pushing back the top of the curve.

pushing back the top of the curve is usually an undesired effect, what we aim is making the toe more printable

we gain some "speed" when pushing, if we consider speed point moves a little to the left in the curve, still that real speed gain is quite small compred to the "pushed stops".

Ian Grant · Nov 13, 2020

It's very rare that I push process these days, if it was something important I'd push process XP2 Super which is something I did for many years starting with XP1. I did a lot of work for a small record company initially push processing HP5 in ID68 (Microphen) but switched to XP1 as soon as it was released.

When Ilford released XP1 they gave times for push processing in C41 developer, however the recommended normal development time was non standard and labs didn't like processing it or push processing. So when Ilford released XP2 they made no mention of push processing and it used the standard C41 development time but it can still be push processed and it doesn't build up the contrast like HP5 or Tri-X and is much finer grained.

More recently I've used Pyrocat HD for push processing LF film, the advantage here is the more development in the highlight area the greater the stain, it's something I need to fine tune I was pushing to close to 3200, it hadn't been planned.

Ian

138S · Nov 13, 2020

Ian Grant said:
More recently I've used Pyrocat HD for push processing LF film, the advantage here is the more development in the highlight area the greater the stain,

...and the greater the stain the more the blue light is selectively blocked in the highlights, compared to lower densities, ending if a selectively lower contrast grade in the highlights if using a VC paper, so highlights are easier to print.

The green/yellow/brown tint of the stain blocks more the blue... so proportional stain is in fact a selective contrast filter when printing with VC paper. This effect is not there with graded paper, and if scanning we don't need it much.

Helge · Nov 13, 2020

138S said:
pushing back the top of the curve is usually an undesired effect, what we aim is making the toe more printable

we gain some "speed" when pushing, if we consider speed point moves a little to the left in the curve, still that real speed gain is quite small compred to the "pushed stops".

Exactly. That's what I'm going for. Guess I didn't make myself clear enough.

Helge · Nov 13, 2020

Donald Qualls said:
Where do you get that idea? Yes, Rodinal is a speed losing developer (compared to the standard ones, selected to give the highest possible speed at standard contrast), but speed is measured at the toe -- minimum light exposure to get (IIRC) 0.2 above base + fog. Further, this isn't really semi-stand -- that would be 30 minutes at 1:100 with agitation at start and at 15 minutes. I think of the 3 minute cycle as "reduced agitation" and based on toe speed, I'm pretty sure this is the sweet spot; it surely does gain shadow detail over standard (one minute cycle) agitation, and gives less contrast in a "push" as well.

The problem with that is would pretty much have to be applied by inspection, because you'd have to partially develop the film to know where to "chemical dodge and burn" in the first place. So you want to give partial development, inspect (under that super-dim green safelight, or via infrared, preferably with desensitized film in either case), then decide where to make irreversible changes to the negative. Seems like a bad idea to me.

Now partial development in one developer, followed up by partial development in another, is legitimate, if probably a case of extra work for little or no gain...

Last thing first: Inspection would be one way to go with bimat type selective development. IR goggles would probably be the only sensible option. You could probably paint the developer on with a brush.
It would be real hand-crafty, time consuming and would probably look it.
The other option would be to make some kind of mask, through a simultaneous exposure. either digital or film.
It wouldn't need to be that high resolution to get the same basic effect you get with simple dodge and burn. But would actually develop shadow detail not possible with paper enlargement.
But this would of course, as always with these simple to describe things, take an enormous amount of work to get to work well and smoothly.
Might be worth it though, to get very high speed.

Second point: I might very well be mistaken that it's not technically push processing and that it's not semi stand. But you normally do a push to gain shutter speed or DoF, which would be less the case with the Rodinal push.
Your points are taken to heart. I will test them out soonish.

Donald Qualls · Nov 13, 2020

Helge said:
But you normally do a push to gain shutter speed or DoF, which would be less the case with the Rodinal push.
Your points are taken to heart. I will test them out soonish.

Nothing says you can't start with a "Push +1" or "Push +2" time before the 1.5x increase -- you'll just be standing there watching the tank for even longer. I've done this (once) with Fomapan 100, but Fomapan isn't the film I'd usually choose to push. In the past, when I needed the most I could get out of Tri-X, I'd use Super Soup, which effectively amounted to "develop to completion". With 320TXP, I'd get about EI 5000, and with 400TX (early oughties version) I could get EI 6400 with fairly normal looking contrast.

138S · Nov 13, 2020

Helge said:
Exactly. That's what I'm going for. Guess I didn't make myself clear enough.

One interesting thing of SLIMT is that it conserves the result in the toe but it has the strong effect in the more exposed areas. Large crystals responsible for the sensitivity in the shadows are difficult to dissolve/bleach, while small crystals that add density in heavily exposed areas are quite easy to dissolve, hence the selective action.

Of course it is debatable if with other techniques we get something equivalent. A drawback is that some bleaching chem has to be handled with extra care.

Using Farmer's reducer after development won't be as effective in the compensation, as metallic silver is in clumps and we don't have those small/weak crystals anymore.

Donald Qualls · Nov 13, 2020

That's interesting. I can see calibrating a SLIMT process needing at the least a lot of testing (film cost, time) and probably wanting a densitometer to see what you're actually getting, but it could be a good way to get another stop out of a film/developer combination that's already at its limit -- for things like available-dark work in bars or poorly lit night streets...

ic-racer · Nov 13, 2020

If faster film is available, I'd go that route. One reason I look at methods to increase shadow detail (speed) is that all the inexpensive 8x10 B&W film I have used has very poor shadow detail, and frequently is a stop slower than the box indicates. And this is with T-max developer.
Obviously if 8x10 T-max 400 film were priced better, I'd use that and not mess with inexpensive 8x10 film.

On my 'to do' list is to try SLIMT with Shanghai 8x10. I was hoping someone else would do it. We are on our third COVID surge and I'm wishing for it to ease or go away so I have some time to do these things.

Lachlan Young · Nov 13, 2020

Helge said:
Developer being the weird and fickle amplifier it is, should probably be one of the main areas of research until/if RnD money gets back to film (I know this is old as the hills pussycat stuff compared to what the really heroic people have done).

Developer technology was much more thoroughly researched than people seem to assume in the post WW-2 era - and the effects produced by the fairly low-level technology discussed in these sort of threads was found to be able to be enacted more effectively (to much more dramatic levels) at the manufacturing stage. The quantities and placement of iodide within the grain structure are arguably much more important than what developer you use - other than the developer being able to access and release that iodide in specific ways. The real trick is balancing solvency against shadow speed without runaway fog.

Helge said:
I still have a fantasy about some kind of bimat developing with highly localized development done by mechanical means.
Sort of like chemical dodge and burn.

In all seriousness, pin-reg masks will do what you want, much more easily. You could even split highlight and shadow values via litho film blocking masks - and print those values at much more extreme variance in grades - which will get you to the same end point - as would an unsharp mask on the neg to soften highlights, followed up by a separately exposed litho mask to burn the shadows down (Radeka et al call it a 'Shadow Contrast Increase Mask').

Combining push processing and compensation?

Privacy & Transparency

Privacy & Transparency