Contrast and Reciprocity Failure

[Vlad Soare]

Hi, guys,

I posted this on Photo.net, too. I apologize to those of you who might read it twice.

I recently took a picture of a church interior. The shadows (which I wanted to place in Zone III) and highlights were four stops apart. So far, so good. But in order to place the shadows in Zone III, I needed an exposure of 1 minute, which due to reciprocity correction became 12 minutes.
Lengthening the exposure to compensate for reciprocity failure also raises the contrast. I should probably pull the development, but by how much? Is there a formula? At least an approximate one? Should I try N-1? Or N-2?
Unfortunately, due to the long exposure time and the fact that I had already pushed my luck by mounting the tripod in a church, I couldn't take a backup exposure. I only have one sheet, so I only get one chance to develop it right.
Should I err on the side of underdeveloping, relying on the fact that adding contrast during printing will be easier than subtracting it? I'm aiming for wet darkroom printing.
The film in question is Fomapan 100.

Thank you.

[Vaughn]

Tri-X needs only about 9 minutes for a 1 minute exposure, Tri-X exposed for 9 minutes to compinsate for a metered 1 minute exposure should be developed about 30% less (according to Kodak's tech sheet on the new version of Tri-X). Might give you a starting point.

You could easily duplicate the lighting in your living room, take a couple shots and do your tests on them.

Vaughn

PS

http://www.foma.cz/Upload/foma/prilohy/F_pan_100_en.pdf

Interestingly, the tech sheet for Fomapan gives a recommended increase of exposure for long exposure times...but gives no recommendation for reducing the development. So I repeat my recommendation of duplicating the lighting in your living room and doing a couple tests.

[Doremus Scudder]

Vlad,

It very much depends on which film you are using, as Vaughn indicated above.

That said, I routinely use a "rule of thumb" that reduces development time by 10% for each doubling of the exposure time. However, it depends heavily on arriving at the correct exposure compensation for your film It seems to me that you have over-compensated for the reciprocity failure. I know of no film that takes so much compensation for a one-minute exposure.

For example, Tri-X is about 9 minutes (the new version slightly less)

Bergger BPF 200 (and the other labels it it marketed under) is only 4 minutes

T-Max films are even less.

The overexposure is probably not a problem, but you don't want to develop too little (you lose a lot of film speed too with reduced development in addition to the contrast recuction).

I would recommend re-checking the reciprocity-failure data for the film you used. You will probably find you overexposed. Don't worry about the overexposure, simply take what should have been the proper exposure time and apply the rule above to arrive at a figure for reducing your development. For example:

IF your film data calls for an 8-minute exposure...


1*2 = 4 (one doubling, - 10%)
2*2 = 4 (two doublings, - 10%)
4*2 = 8 (three doublings, - 10%)

Total development reduction = minus 30 % (what Vaughn recommended)

You should, of course, base your actual development reduction on the published exposure compensation data for the film you were using (NOT on your actual exposure).


Hope this helps some,

Best

Doremus Scudder

[Vlad Soare]

Doremus, Fomapan is one of the worst films when it comes to reciprocity failure. I personally don't mind that, firstly because I love it for other reasons, and secondly because in this particular case the long exposure was advantageous to me (people could wander in front of the camera without being recorded).
The reciprocity correction for Fomapan 100, according to its official data sheet, goes as follows:
1s - 2x
10s - 8x
100s - 16x
My inference was that 60s is halfway between 10s and 100s, so the compensation must be (at least approximately) halfway between 8x and 16x. Thus 1 minute became 12 minutes. Well, I know it's not a linear function, but the error cannot be that great.
Unfortunately, as Vaughn noticed, the Foma data sheet gives no indication regarding any development correction.

Trying to duplicate the lighting at home and to make some tests seems like a great idea. I think I'll try that.

Thanks. :)

[Nathan Potter]

Vaughn is spot on. If you have your metering data from the single photo, rig up reflection objects in your home area and duplicate the metering of the original scene. Take several photos and try N, N-1 and N-2 development tests. I've done this with both interior and exterior scenes where I've realized that I've mis-exposed, especially if the image is particularly valuable.

Nate Potter, Austin TX.

[Brian Ellis]

Going from 1 minute to twelve minutes is a gigantic reciprocity correction, I've never heard of anything even close to that kind of a correction when the metered exposure is only a minute. For comparison, with T Max a metered 1 minute exposure becomes 2 1/2 minutes, with all other films (except Fomopan apparently) a 1 minute metered exposure normally becomes 5 minutes. With T Max the contrast increase is N+1 at 2 1/2 minutes, with all other films it normally is N+2. So if I wanted normal contrast with a film other than T Max I'd process for N - 2. I don't know if that holds for Fomopan since it apparently is a very different film than all others when it comes to reciprocity failure but that's what I'd try.

[CG]

The various contradictory sites I've checked out say Tri-X at a metered minute expands to anywhere between 4 minutes and 9 1/2 minutes.

C

[Vaughn]

The various contradictory sites I've checked out say Tri-X at a metered minute expands to anywhere between 4 minutes and 9 1/2 minutes.

C

Off the Kodak website I found their tech publication for Tri-X...the graph says for 60 second exposure, increase to 550 seconds (9.2 minutes).

Like anything else, one's mileage may differ.

Vaughn

[Bruce Watson]

Tough situation. Just remember that reciprocity failure typically only happens in the shadow areas -- not enough photons hitting the film to form a stable latent image. The highlights however, are doing fine with plenty of photos and normal latent image formation. Then when you increase exposure to handle the reciprocity failure in the shadows you way overexpose the highlight areas. In your case, you literally cook them with 10x the exposure they should get. It shouldn't be any wonder then that your highlights might be very dense indeed.

For this piece of film I would probably give it N-2 or N-3 processing. And hope that Foma shoulders strongly. But I would also expect that it will be very difficult to print in the darkroom because of the highlight density.

In the future I suggest that you use a neutral density filter. Sounds crazy doesn't it? But what I think you should do is reduce the light hitting the film until the entire sheet is in reciprocity failure, shadows to highlights. Your exposure will be measured in hours. But your film should have a much nicer density range and should be easier to print.

Since every film will respond differently to such treatment you'll have to experiment a bit to find out what really works best for you. But at least I've given you something to think about maybe.

[Nathan Potter]

Bruce, that's an interesting approach - using a neutral density filter to effect uniform reciprocity failure everywhere. Never thought about that. But if this failure is highly nonlinear as a function of incoming photon density - I think it's an avalanche multiplication effect failure that yields the nonlinearity - are you sure that the application of a neutral density filter which will not change the brightness range of the scene - only reduce it - would help. I'm writing this without thinking it through clearly. I need to get my French Press going!

Nate Potter, Austin TX.

[Bruce Watson]

Bruce, that's an interesting approach - using a neutral density filter to effect uniform reciprocity failure everywhere. Never thought about that. But if this failure is highly nonlinear as a function of incoming photon density - I think it's an avalanche multiplication effect failure that yields the nonlinearity - are you sure that the application of a neutral density filter which will not change the brightness range of the scene - only reduce it - would help.

Do I know it will work? Actually, no, I don't. It's not my kind of shot, and it's not a film I've ever used. My personal approach is to use TMY-2 and drag the film the other way -- out of reciprocity failure entirely.

But... the full reciprocity thing has worked in the past with older thicker emulsions such as Super-XX (Foma is supposedly closer to Super-XX than it is to more modern "traditional" films like Tri-X). Remember those wonderful old (1930s?) photographs of railroad stations completely devoid of people? Taken from the mezzanine with the light streaming in from the windows? These were usually six+ hour exposures resulting in beautiful prints with good shadow detail and excellent highlights. I figure if it worked for photographers back then it might be worth a try today.

[Vlad Soare]

Thanks, guys.

I followed Vaughn's suggestion and shot a test sheet. The original scene was EV 2 in the shadows and EV 6 in the highlights, which on an ASA 100 film gave an exposure of 1 minute at f/32. So, I found something textured and almost black, I put it next to a white paper towel and adjusted their position and the lighting until I got a reading of EV 2 and EV 6 respectively. Then I exposed for 12 minutes at f/32.
I developed it for N-1 (about 15% shorter than normal), and surprise! It came out perfectly. Admittedly, I haven't tried to print it, but on the negative I can clearly see the texture of the paper towel, and its density is clearly different from the maximum black the film is capable of. At the same time, the black object shows full, strong details. At first glance it looks like an easily printable negative.
At the same time I also developed another sheet, which needed N-1 (no doubts there - there was no reciprocity failure involved). It turned out exactly as I expected. That's why I believe that my N-1 dilution and time were correct.

Anyway, whether I call it N-1 or otherwise is not important. The important thing is that now I know how to develop the real picture. I'll do it tonight and let you know how it turns out. :)

Thanks.

[Doremus Scudder]

Vlad,

Glad you solved your problem. I took a look at the Foma data sheet, and, indeed, the times for reciprocity failure (Schwarzschild effect) are as you say. I would have a tendency to do some tests to see if they are really accurate. Patrick Gainer's article on reciprocity failure (LIRF here: http://www.unblinkingeye.com/Articles/LIRF/lirf.html) maintains that the log graph angle for all films is essentially the same, and the Foma data seems to contradict this.

Bruce,

I think using a neutral density filter as you suggest will simply increase the contrast. A ND filter reduces all light equally, both high and low intensity, however the film reacts non-linearly to low light: i.e. the lower the light level after a certain point, the more you need to increase exposure. Using an ND filter and longer times increases the effective difference in density on the film, not decreases.

FYI I often us an ND filter and the technique you describe to move the film into reciprocity. However, I do it to INCREASE the contrast. Sometimes it has a nicer look and less grain than extended development. Kodak's data, and other manufacturers also, recommend reducing the development more in direct proportion to the length of the compensated exposure. That shows that contrast increases with the time of exposures compensated for reciprocity.

Best,

Doremus Scudder

[CG]

Tough situation. Just remember that reciprocity failure typically only happens in the shadow areas -- not enough photons hitting the film to form a stable latent image. The highlights however, are doing fine with plenty of photos and normal latent image formation. Then when you increase exposure to handle the reciprocity failure in the shadows you way overexpose the highlight areas. In your case, you literally cook them with 10x the exposure they should get. It shouldn't be any wonder then that your highlights might be very dense indeed.

For this piece of film I would probably give it N-2 or N-3 processing. And hope that Foma shoulders strongly. But I would also expect that it will be very difficult to print in the darkroom because of the highlight density.

In the future I suggest that you use a neutral density filter. Sounds crazy doesn't it? But what I think you should do is reduce the light hitting the film until the entire sheet is in reciprocity failure, shadows to highlights. Your exposure will be measured in hours. But your film should have a much nicer density range and should be easier to print.

Since every film will respond differently to such treatment you'll have to experiment a bit to find out what really works best for you. But at least I've given you something to think about maybe.

Won't work. It won't make it easier to deal with. Each different zone of an image exposed for hours or days will be facing differing levels of reciprocity failure. By increasing exposure time you will increase the contrast as well as the pain in the backsides issue.

C

[Vaughn]

Vald...it always makes me nervous when someone actually takes my advice.LOL! Glad it seems to be working out!

Vaughn

[Vlad Soare]

I developed the "real" sheet last night, using the N-1 dilution and time that worked so fine with the test sheet.
At first glance it seems to be a success. Both shadows and highlights are rich in details. I have some small completely black areas, but those were black to begin with - they were below EV1, outside the range of my meter.
I can't wait to print it. :)
Thanks, guys.

[Vaughn]

That's great, Vald!

FWIW, I use the reprocity failure of a film to bump up the contrast of the scene, as I use negatives of very high contrast (but full tonal range) for carbon printing. Nothing nicer than a scene with 7 stops of light in it and giving the film about a stop or so less than normally required to compensate for the RF (30+ second exposure) and then giving the film (usually FP4+) 150 to 200% of "normal" development.

Vaughn

[Andrew O'Neill]

In the future I suggest that you use a neutral density filter. Sounds crazy doesn't it? But what I think you should do is reduce the light hitting the film until the entire sheet is in reciprocity failure, shadows to highlights. Your exposure will be measured in hours. But your film should have a much nicer density range and should be easier to print.

Interesting but won't work....It would only work if the ND filter lowered the intensity of the highlights down to the same level of intensity as the shadows...You would then end up with a very flat negative and would have to give N+CRAZY to get a "normal/printable" negative...

I use FP4+ and HP5+, as well as Efke 25, and occasionally TMY. My testing has indicated (using the Phil Davis method of testing reciprocity) that these films do not require development compensation for long exposures. The increase in contrast is so small, I don't worry about it. I've been using my data for 10 years so I know it's solid.
So, as someone previously mentioned, every film is different and that includes its response to reciprocity effect or failure or whatever you want to call it...so, you will have to test your film with your gear with your developer and with your working methods...

[Patrik Roseen]

Hmm, is this one of those scenes where it would be good to pre-light the total sheet of film to decrease the contrast. Would this actually mean lifting the shadows out of the zone of reciprocity?

[willwilson]

Great Article by Howard Bond, includes his times for tri-x, hp-5, delta 100, Tmax400/100:

http://www.phototechmag.com/articles/articles/200705/0403Bond_Reciprocity2.pdf

I use his times for Delta 100 and they are great.

[Vlad Soare]

This is the final photo. Though it required split contrast printing and some mild dodging in the upper (inner) parts of the arches, overall it was an easy print.
This is the scan of the print, not the negative. The print actually looks better than the jpeg. The stained glass window seems a little washed out here (at least on my lousy laptop monitor), but it looks just fine on paper. The columns also look a bit brighter on paper, and the blacks between the pews a little blacker.

It may not be a great image from an artistic point of view, but it was the first time I took a picture under such circumstances, and I'm glad that I got the technique right. :)

Thanks, guys.

http://img396.imageshack.us/img396/8105/image0256micrn4.jpg
Votivkirche, Vienna

[Vaughn]

Nice, Vlad! it is great you kept the feeling of light coming through the stain glass windows, yet have a great feeling of light on the organ pipes!

Will...Thanks for the article!

Vaughn