I was using Fuji Acros for all my B&W photography, but since moving to LF, I can't find this film in sheets locally. So I'm considering moving to Ilford Delta 100 (a few less coins too). Now here is my question: I can't find table for reciprocity failure of this film (mathematic formula would be nice). The graph that Ilford is showing on their PDF is a bit on a smallish side, especially on the beginning of the graph (1/2-10 sec). Is there any kind soul out there, that can help me with their experience? Thank you Marko

Sandy King published a formula for calculating the increased exposure. Many of us made charts Here is mine http://home.pacbell.net/mkirwan/reciprocity_failure.htm. Click on the workbook link and you can down load the tables. It includes Delta 100 Enjoy Mike

How about Agfa APX? How does this formula work? I'd like to get better data on Agfa APX 100 and 400 and the Efke films. Bruce

I think you might get a better fit if you fit an equation to just the amount you must add to the indicated exposure time. Thus, (tadj - tmeas) = a * (tmeas ^ b) You can linearize it by the log transform to: log(tadj - tmeas) = log a + b log tmeas Just call log a = c and solve for b and c. Then a = 10^c. On log-log paper, this is a straight line with intercept a at tmeas =1 and it goes through 100^b at tmeas=100. I have found that for the films tested and reported by Howard Bond in Photo Techniques, the b coefficient is 1.6. The only significant difference among those films is the a coefficient. I found these values to fit the data quite well: 400TX.....0.17 400TMX...0.06 100TMX...0.07 HP5+.......0.10 100Delta..0.05

Thank you, folks! Mike, are you sure your courve is right one? I mean, on Ilford's pdf, the curve is much steeper. Marko

The chart in the Ilford Delta 100 data PDF data sheet uses a linear scale on both axes. The chart I show on my site uses a logarithmic scale on both axes which will give a linear and less steep curve. Hope this helps Mike

Simply plotting the Ilford data on log-log paper will not give a straight line. The Ilford curve plots measured exposure plus additional exposure against measured exposure. If you first subtract out the measured time from the total time, you will get a linear plot on log log paper. All the curves for the films I listed above are parallel lines on log-log paper. You see the problem: the log of a sum is not the sum of the logs. The log of a sum plotted against the log of either of the addends is therefore not a straight line.

By the way, when I plot the Ilford data, subtracting the measured time from the total time, the result is a straight line exactly parallel to the data provided by Howard Bond, but shifted upward. On log-log paper, this means that the data from Ilford differs from that from Bond by a constant multiplier.

Gotcha. This is really strecthing my poor old brain. The sum of logs is multiplying, subtracting is dividing. OK now off to take a well earned nap Mike

Mixing Long Exposures and Short Exposures on the same roll I'm shooting this film in a Hasselblad 501C, and developing it in Kodak XTOL. My question: As long as I apply the reciprocity conversion to long exposures (e.g., 15 minutes), I can shoot other shots on the same roll at normal shutter speeds (e.g., 1/125th sec.), right? Basically, what I'm asking is if it's okay to mix long exposures and short exposures on the same roll, and then develop the roll according to my normal procedures. Much thanks!

The only possible problem with mixing long and short exposures is that very long exposures may involve unusual contrast ratios that you would otherwise want to compensate for by adjusting your development. If that isn't the case, go right ahead. Matt

Yes, but . . .any metering error (I don't know if error is the right word) will be amplified by orders of magnitude. These reciprocity curves are great, if your metering is spot-on. At normal shutter speeds, a quarter-stop slow on this one and a quarter-stop fast on that one is no big deal. But when you go several stops out on the published curve . . .extrapolating five stops out turns a 1/4 stop error into an 8 stop error. Some of these things will not be like the others.

As an example, if you are shooting something like a night scene illuminated by a street lamp, the contrast may be quite harsh. If you try to compensate for that by "pulling" your development, you will reduce the contrast for all frames on the film, and that reduced contrast may not work well for those other frames. It isn't really a problem with the longer exposures, but rather a problem with the type of lighting conditions that often lead to the need to do longer exposures in the first place. Matt

Howard Bond, in the article from which Gainer derived his equations, found that there was no increase in contrast with Delta 100 with exposures adjusted from a base exposure of up to 240 seconds. So there's no need to worry about this problem with Delta 100. Lee

Lee: Thanks for your post. If I understand it correctly, you (and Howard Bond) are referring to the behaviour of the film when exposed for longer times. My observations were related more to the lighting conditions one often encounters when longer exposures are required. Matt

I was editing the Bond quote (from a messy multi-column .pdf "cut") when your post #14 to this thread came in, which makes your meaning clearer. Yes, my post is in reference to the contrast increase in the film caused by longer exposures, which isn't present in many modern films, including Delta 100. I'm not referring to the greater dynamic range lighting conditions found in many low light settings that you mention. Lee

Thanks, everyone. To tell you the truth, the long exposures I'll be shooting probably won't be shot in contrasty lighting, so I should be fine on that aspect. Stuff like this is good to know; I'd just hate to waste frames or delay on developing a roll simply due to exposure [time] variations from frame to frame. I think I should be fine, but I'll test a few rolls before I try this for real.

Search www.unblinkingeye.com for my article "LIRF is Lurking at Your f-stop". LIRF stands for "Low Intensity Reciprocity Failure." There you will see what the logarithmic curves should look like. The straight line results when one plots the amount of exposure time to be added to the time measured with the ordinary exposure meter against the measured time on log-log graph paper, or the log of one against the log of the other other on ordinary graph paper.