Reprocity failure for FP4+ film Question

[Willie Jan]

Hi,
i use fp4+ on 4x5 (landscapes and fineart) and when using times of +1 second i add time for the reprocity failure.

But my feeling when i look at the negs afterwards is that it is not that high as mentioned on paper.

Yesterday i did a test with a fuji acros, shot it at 8 seconds, took a fp4+ and shot it also at 8 secs. After development the density of the gray background looks the same ( i tried rodinal and pyrocat HD).
ilford says when 8 seconds is found by the meter, use 25 seconds instead.
that is 1.5 stop overexposed. So my ilford neg should be 1.5 stops underexposed, but the shadow details are the same on both negs....

What I do is I use a digital spotmeter and gray card to get the right time in my fine-art studio. the light I use is daylight.

Is the reprocity failure a phantasy story when you use a proper exposure and development????

 

[Lee L]

Robert Reeves' tests ( http://www.robertreeves.com/b&w.htm ) indicate a Schwarzschild factor of 0.79 for Ilford FP4+. You can plug this into the formula:

Corrected time = (metered time+1)^(1/p)-1

where p = the Schwarzschild factor and time is in seconds.

Manufacturers have a general tendency to publish generic reciprocity adjustments unless the film is exceptional in that regard.

Here are times in seconds for FP4+ using Robert Reeves' numbers and the Schwarzschild/Covington formula (overkill on precision from my spreadsheet):

metered adjusted
1 1.4
2 3.02
4 6.67
8 15.14
15 32.44
30 76.23
60 180.93
120 431.95
240 1034.65
480 2482.82

Here's a good thread on reciprocity failure with lots of perspectives: (there was a url link here which no longer exists)

Lee

 

[gainer]

Here is another one:
http://www.unblinkingeye.com. Look for "LIRF is Lurking at Your F-stop". My numbers were obtained from experimental data of Howard Bond as published in Photo Techniques. Neither my equations nor Howards data agree with the Swarzschild equation.

If the time added at 1 second indicated exposure is 0.4 seconds, The corrected time Tc for any other indicated time Ti will be:

Tc = Ti ^ 1.62 * 0.4 + Ti.

Thus, at 8 seconds indicated, the corrected time should be:

Tc = 8 ^ 1.62 * 0.4 + 8 = 19.6 seconds. Oh heck! Call it 20.

The shape factor 1.62 agrees with every reciprocity curve of real data I have seen. It does not agree with Swarzchild. A test is in order.

 

[Willie Jan]

Thanks for the replies.

I haven't used a scientific approach, but my results show me that even doubling the time from 8 to 16 is much to much.
I will do a test in the future for ilford fp4+ rollfilm where I expose each frame doubled in time and normal to see the difference and see what the density is...

 

[Lee L]

Bond and Reeves used different methods of determining adjustments.

Reeves uses this method http://www.robertreeves.com/filmtest.htm which involves a "normal" 1/8 second exposure and then a single data point with known neutral density at 128 seconds to test the amount of reciprocity failure. Reeves tests for a mid-tone and for astrophotography.

Bond uses a much more comprehensive method across fewer films, with speed loss and contrast tests at 1, 2, 4, 8, 15, 30, 60, 120, and 240 seconds. FP4+ is not one of the films Bond tested. Bond tests for shadow detail speed and for "normal" photographic subjects.

I have run the numbers on both methods using data on a couple of films from Bond and Reeves, comparing Gainer's method to the Schwarzschild formula. As Patrick says, the curve shapes are different. However, I found that if you tweak the Schwarzschild numbers within the 1/3 stop experimental error in Reeves' results, I could match the curves from both the Gainer/Bond and Reeves/Schwarzschild curves within 1/6 stop out to several thousand seconds.

The real point here is that the manufacturers numbers are very often generic, outdated, or inapplicable, as you found for yourself, and as Reeves and Bond both found. Pick either method mentioned so far in this thread and you'll get better results than with some of the manufacturers' generic adjustments. Gainer's method is easier for calculations in the field. As Patrick says, even better to test for yourself, pretty easy with one film/developer combination.

Lee

 

[gainer]

As a matter of fact, the manufacturer's data for HP5+ fit closely the equation:

Tc = Ti ^ 1.62 * 0.11 + Ti.

The adjusted exposure time for HP5+ at 8 seconds indicated will be only 11.2 seconds. I was unable to find the manufacturer's curve for FP4+.

Although the constant 0.11 should be the correction to be applied at 1 second indicated, that is not a good way to estimate the coefficient. The difference at 1 second is often negligible. If you have reliable experimental data at 8 seconds indicated that tell you the corrected exposure should be, say, 10 seconds, you may calculate the coefficient, call it f, as:

f = (Tc - Ti) / Ti ^ 1.62.
Thus:

f = (10 - 8) / (8 ^ 1.62) = 2 / (8 ^ 1.62) = 2/29.04 = 0.07

Now you can calculate the estimated adjusted exposure for any other indicated time as Tc = Ti ^ 1.62 * 0.07 + Ti.

Just about any pocket scientific calculator will be able to do the calculation in exactly the above sequence of key strokes in a couple of seconds, so all you have to remember is the constant f and the sequence of strokes.

My calculator is the Casio fx-991s. The key that does the ^ operation is labelled Xsuperscript. Others will have the ^ symbol.

I do not think the factor for FP4+ will be much different from 0.11, so make allowance for that as a test point in your experiments.

 

[gainer]

It is possible that Ilford changed their reciprocity data since Bond did his experiments. The data I found at an Ilford site agreed with the Bond data which gave a coefficient of 0.11 and the 1.62 shape factor.

From what I remember of Ilford data when Bond published his data, the shape of the curve for FP4+ was the same as his but shifted so as to produce greater adjustments.

 

[Willie Jan]

Thanks for your help.

appriciated!

 

[John Bond]

Two questions:
First, when comparing all of the different ways of calcualting reciprocity adjustment, are development times similar between the various techniques? If not, would different development times account for some of the discrepancies?

Second, it seems useful to bracket long exposures. By how many stops (or how much time) do these different methods differ and what would be a reasonable strategy for bracketing?

Thanks,

John Bond (no relation)

 

[Lee L]

Bond found that development changes (reduction in time) were not really necessary with the films he tried, as contrast didn't build in the way it did with older emulsions. BTW, this was in the July/Aug 2003 Photo Techniques.

Bracketing with really long exposures can become tedious and time consuming. Apart from that, I'd suggest that you just try moderate bracketing and see if there's any pay off for your specific conditions.

The only long exposures I've bracketed consistently are astrophotography shots using regular camera prime lenses, and I do that mainly to get maximum exposure time before light pollution overcomes the background skies. I do that by adjusting times using Fibonacci series numbers for the minutes, which works out to a factor of 1.62 between steps. I've also done the same using seconds for terrestrial night scenes where the dynamic range is great and I want to get variations from which to choose. It also helps me remember my exposure times (especially when I'm back out shooting under similar conditions), and is a reasonable adjustment for getting a significant difference without going too far.

Lee

 

[gainer]

Bond found that development changes (reduction in time) were not really necessary with the films he tried, as contrast didn't build in the way it did with older emulsions. BTW, this was in the July/Aug 2003 Photo Techniques.

Bracketing with really long exposures can become tedious and time consuming. Apart from that, I'd suggest that you just try moderate bracketing and see if there's any pay off for your specific conditions.

The only long exposures I've bracketed consistently are astrophotography shots using regular camera prime lenses, and I do that mainly to get maximum exposure time before light pollution overcomes the background skies. I do that by adjusting times using Fibonacci series numbers for the minutes, which works out to a factor of 1.62 between steps. I've also done the same using seconds for terrestrial night scenes where the dynamic range is great and I want to get variations from which to choose. It also helps me remember my exposure times (especially when I'm back out shooting under similar conditions), and is a reasonable adjustment for getting a significant difference without going too far.

Lee

And my article analyzing his results was in the following issue.

You could move to my part of West Virginia to avoid light pollution.

Bracketting f-stops is OK when you are not concerned about depth of field. Bracketting time is hard to predict and sometimes hard to tolerate.

I knew there was something magic about 1.62 besides its geometric perfection.

 

[Lee L]

And my article analyzing his results was in the following issue.

You could move to my part of West Virginia to avoid light pollution.

Bracketting f-stops is OK when you are not concerned about depth of field. Bracketting time is hard to predict and sometimes hard to tolerate.

I knew there was something magic about 1.62 besides its geometric perfection.

Patrick,

I haven't ordered the back issue with your reciprocity article yet, although I did fill in around my lapsed subscription with most of your developer articles. And you've described your method well here on APUG and Unblinkingeye. I did happen to stumble across Bond's article when it was on the magazine website briefly as a publicly available .pdf, so I snagged it. For those looking for your follow on article, it's titled "Reciprocal Trade Agreement". They appear to have taken down all the sample articles they used to have on the Photo Techniques web site, and I don't feel free to distribute the .pdf of the Bond article due to copyright.

As you know, 1.618... is the Fibonacci ratio, the golden mean ratio known as phi after the sculptor Phidias, etc... lots of good natural resonance right around that spot. I used Fibonacci numbered seconds to hand print B&W in a lab, about 2000 prints a month with 2 D5's and a roller transport, judging exposure by eye. People were surprized when they asked for repeat prints and got essentially identical reprints, but I knew it was always at 3, 5, 8, or 13 seconds, and that spacing works great for meaningful but not drastic changes in density. You can also make a test strip at that spacing by just moving a card as you count seconds, and letting the exposures stack up, like f-stop printing at just over 1/2 stop intervals.

I used to live south of you near the Jefferson National Forest and the Blue Ridge Parkway, where I first started shooting night skies under good conditions, but a job change forced a move. I do visit the "wild and wonderful" when I get the chance.

Lee