hp5 Reciprocity Effect Question

Hi Folks. Yesterday at the library I looked through a 1992 edition of

Amphoto Book of Film by George Schaub. It gave exposure info as well

as developing reductions for many films. For hp5 it suggested to add

1/2 stop for 1 second exposure, add 2 stops for 10 second exposure,

however, it did not give a % of developing reduction for these

figures. Can anyone give me advice. I want to take some rainy night

street shots with this film. I will soup in either Nacco Super 76 or

Clayton F76 Plus.

PS - When the author writes to add 2 stops, that means to open up the

lens two more f stops wider correct? I kind of new to night

photography as you can guess. Thanks a bunch!

Regards,

Marc

Yes, opening up two stops means to open the lens two stops wider. I don't think you would want to reduce development in this case.

There is a fascinating and informative discussion of reciprocity going on here:

 

http://www.apug.org/forums/showthread.php?t=11566

 

 

complete with calculations, tables and charts.

You would develop the film as normal. Reciprocity means a relationship between shutter speeed and aperture. 1/60 @ f/1.4 is the same exposure as 1/30 @ f/2 for example, the same amount of light is recorded on the film, which really means that in both cases, the chemical reactions in the emulsion are the same. This means that for a given exposure, you have a certain amount of scope to use shutter and aperture creatively. If you want a nice motion blur, for example, you can stop down and the overall exposure of the scene will be the same. If you want to freeze motion, you can open up the aperture, which will make the shutter faster (at the cost of some DOF, but the less DOF the better as far as I am concerned).

 

However film is not perfect and this relationship breaks down outside of the shutter speed range 1/10000 to 1. The film simply becomes less chemically responsive the longer it is exposed. To record the same amount of light as 1 @ f/1.4 you might need 3 @ f/2, to overcome this chemical inertia. However, this doesn't mean that you are exposing the film at ISO 200 - it is still an ISO 400 film and should be developed as such - you are merely compensating for reciprocity failure, not pulling.

 

Digital does not suffer from reciprocity failure, but it does suffer from noise, whereas film grain is not really affected by long exposures, so provided you are willing to live with reciprocity failure (and it's not difficult to compensate) film still has the advantage.

Marc,

 

Reducing development is often recommended for long exposures because reciprocity failure has the effect of increasing contrast on the negative. This is because dimmer parts of the scene require more reciprocity correction than brighter parts. As we increase exposure to pull out more detail from the shadows, we run the risk of blowing out the highlights which are not as affected by reciprocity.

 

Opening up the lens has the effect of shortening both the highlight and shadow exposures so they are not as affected by reciprocity. In this case, your contrast probably won't be affected too much and you can leave the development time alone.

 

If you're doing rainy street photos, you may have a contrast problem built into the scene already, if there are streetlamps reflecting off of wet asphalt. I hesitate to make any recommendations since I don't know your printing setup or the type of negatives you normally get from your development process. Try it and if it's too contrasty, reduce development by 10-15% and try again.

 

- Robert<div></div>

Reducing development depends on the contrast of the scene. If it's really contrasting - the only light sources are point sources, like single street lamps, etc, rather than any kind of diffuse lighting - then you'll need to reduce development. Usually, night scenes do have such contrast. I almost always end up needing to reduce by 10% if not more like 20%.

 

For what it's worth, I've made up a reciprocity chart _based_ on Pat Gainer's formula from the apug thread. The tricky part is finding out whether it's accurate or not (since it does deviate from manufacturer charts quite a bit in some instances).

 

allan

Guys;

 

Just FYI, reciprocity failure can affect up to 4 characteristics of every film or paper and is different for every emulsion.

 

It can change the slope of the characteristic curve thereby changing contrast.

 

It can change the shape of the curve in the toe.

 

It can change the shape of the curve in the shoulder.

 

It can change the threshold (ISO) speed.

 

In color, it can change these differently for all 9 emulsions and in B&W it can change them differently for the 3 emulsion blends.

 

This leads to latitude changes, contrast changes, bumps in the curve, and in color - crossover and color shifts.

 

Of course, I am speaking of times longer than about 1 - 2 seconds here, not less than that. Typically, when I ran reciprocity failure tests at EK, I plotted at least 3 factors for every emulsion present.

 

Considering anything less than what I describe above (using one point for example to generalize) is a dangerous oversimplification, but it can be made to work as seen in the thread on APUG referenced above. The manufacturers do everything possible to minimize these effects, and a great deal of improvement has been made in recent years as seen in the fact that there are no longer "S" and "L" color films. Nevertheless, reciprocity in films remains with us as a potential pitfall with long exposures.

 

I would like to remind you as well that reciprocity failure rears its ugly head at very short exposure times as well, such as those less than 1/1000th of a second, and therefore can be seen from time to time in electronic flash exposures.

 

Ron Mowrey

Ilford is not so picky. They print the same reciprocity chart for all their films. They present corrections only for exposure times longer than ... well, 0. The equations I presented in Photo Techniques and again recently in APUG fit the Ilford graph to within 1 second all the way from 5 to 35 seconds indicated time, or 12 to 200 seconds adjusted time. I'll try to submit a graph to show what I mean.

 

My equation also fits the old Kodak data for Tri-X. Have they presented any of the data from the tests you did while working there, Ron?

I forgot to say the numbers on the right side of the curve are from my formula. The numbers on the left side are the best I could get from an measurements on an enlarged copy of the Ilford chart.

Patrick;

 

Fully 100% of my reciprocity tests were with emulsions not then used in products and virtually all of that was intended for color. Of course, I used checks and balances, but again - those were all color. The experiments were in color and B&W coatings but cannot be related to products today.

 

I was looking for some graph paper the other day, and I found a folder of it in a filing cabinet. Along with the graph paper I wanted was a packet of EK reciprocity test paper, chromaticity graph paper, and a host of other specialized graph papers that we used to use. Just a coincidence but it did rather surprise me especially in view of this and the other thread.

 

Ron Mowrey

I'm not positive what the author meant when stating "add two stops". Normally, this can be done by adjusting either f/stop or the time, but with long exposures, the results will not be the same (hence reciprocity FAILURE).

The recommendations published by ILFORD specified no change in the f/stop, all changes were made with the exposure time.

 

Until getting to extreme exposures, HP5+ does not require the change in development that was common years ago. However, many situations that require long exposures are often very high in contrast; this may call for modification of development.

 

There was a good article by Howard Bond in Photo Techniques Magazine a while back; I think about May or June 2004. He tested HP5+, and other films, extensively.

 

The ILFORD chart works pretty well for HP5+ and FP4+, although it's not perfect. It doesn't work very well for the Delta films, despite being published in the Delta data sheets. The Delta films do not experience as great amount of speed loss.

 

David Carper

In the next issue after Howard's report, I analyzed his data and found that an equation of the form:

 

Tc = Tm + Ff ( Tm ^ 1.62 )

where Tc is the adjusted exposure time, Tm is the time indicated by your meter, Ff is a factor that is constant but different for each film, and 1.62 is an exponent that works for all the films Howard tested. It fits the Ilford curve just about exactly when the Ff is 0.51. The coefficient that fits Howard's measurements of HP5+ to the equation is 0.11.

 

The term (Tm ^ 1.62) is the same for all filmes I have analyzed. It is easily solved on any pocket calculator that will do exponents, such as the TI-30. You can plot the basic corrections vs indicated time on a graph or in a table. Plotted on log-log paper they are a straight line. All you need is this chart or table and a list of film factors and you can take pictures in cathedrals or caves, as long as you can see to read your meter.

Patrick;

 

If it was that simple, I wonder why the 'great' figures in photography never 'discovered' this equation and published it. Haist, James, Hamilton and a host of others are silent about this.

 

If it helped photographers, I wonder why EK or Fuji never published it?

 

Maybe it is because it is a recent development? Maybe it does not work all the time? IDK. I'm not suggesting anything or suspecting anything. But it does seem to me that they would have discovered such a simple relationship if it has existed all along.

 

Anyone have any thoughts? (no conspiracy theories please UFOs and LGMs have nothing to do with this)

 

Ron Mowrey

James and Hamilton would not come up with such an equation because they are very well aware of other factors having influence on low intensity reciprocity failure or sensitivity in general... humidity, temperature, etc... And don't forget to include Tani to that list :-)

 

(I know, there are at least a couple of Tani's, and there is Tan, it's confusing :-)

We don't all think alike, especially after 30 years at NACA-NASA, 1952 - 1982. And I didn't always think like the other guys there. I didn't produce a large number of reports, but each one was on a different subject. The equation I gave with the exponent 1.62 and a film factor of 1 fits pretty well what Kodak used to publish for Tri-X. There's not a whole lot new about that.

Ryuji;

 

I'm familiar with Tadeki Tani's work. I've met him and had dinner with him and Paul Gilman to talk over some of our mutual work IIRC. We were joined by several EK and Fuji people. I may have a photo of that around here somewhere.

 

We had a separate dinner with the Konishiroku people.

 

Regards.

 

Ron Mowrey

I don't know all that you guys do. Maybe that's a bad thing, maybe it's good. If you don't know that you are not supposed to be able to do something, maybe you try it and find that you can. If you don't know that zero sulfite in a developer is a bad thing, maybe you find out that it really is not a bad thing. Sometimes if you start with a complicated system and try to simplify it, you hit a peak of some kind that makes you think you shouldn't go any farther in that direction. This is something like the sonic wall where an airplane reaches infinite pressure at the speed of sound..or does it? That was linearized theory, not fact. It got in our way for many years, that sonic barrier.

 

Humidity, temperature or the phase of the moon may affect our photography, but if I test film after film and find the same exponent at work (or play) maybe the system isn't so complicated after all. If I analyze the work of others who have no vested interest in my ideas and find a common thread that men of genius have overlooked, should I put my tail between my legs and say I must have made a mistake? If someone shows by analysis of data that I have , it won't bother me too much to admit it, but so far I have not seen any evidence of it. I must say that I found a different exponent for the data for printing paper in a Kodak publication, but the equation was of the same form: a*T^b, and only the constant a varies with film, even when comparing old style and tabular grain types.

 

Remember, I am not throwing black and white and color films into the same bucket. I don't get much fun out of color film: just pictures to send to relatives.

Patrick;

 

If you found your data agreed with a published equation except for the exponent or the constant, then where I come from it is correct science to give a reference to the prior art and a footnote at least with an attribution that you are extending the work.

 

This way, the originator is acknowledged properly, and your work is also supported on a firmer foundation. Remember too that many times these formulas are known but not published, rather just the graphs of the data are published.

 

It should be clear to everyone out there that if you can graph data, then a formula exists to achieve that graph. Anyone can then work backwards from the data to achieve the formula. You have now made clear that you have verified your work by doing just that, but you have not cited any resources which would be very helpful.

 

I should add that if you can graph data and have a formula, a generalization often cannot be made to other sets of data or other formulae related to your original graph. If graphs matching or even approximating your derived formula exist, then it would be correct to state "Existing data show the following.... which yields the following equation.... and by extension I have found that the constants for these films are .... by application of my tests and the general formula derived by .... in ..... etc etc". If you discover this fact after your work is published, then a follow up publication is warranted which gives full attribution to the prior art.

 

To use an analogy in your field for consideration, the dynamics of supersonic vehicles can be described by similar aerodynamic equations, but the SST and the Mercury Capsule certainly marched to different drums during their supersonic careers. I would not want to predict the exact behavior of one from the other based on general principles. And, anything I did would have copious footnotes citing all of the work of others as well as attributing their work properly at all time. I would also point out all of the potential pitfalls in my predictions comparing the two vehicles just as should be done when comparing films based on different emulsion types, dopants and etc.

 

Ron Mowrey

You did not see the full extent of my report. I showed you an equation form and a particular example of a fit to the work of others. Ilford and Kodak do not always tell us who did the research for the foundation of the curves they present. The research is company property and the results are only published as an aid to the consumer. I did cite Ilford and Kodak. Howard Bond's experimental research as presented in Photo Techniques, July/August of 2003, is also pertinent and will be a reference in whatever I report to the public.

Sir Isaac Newton hesitated to present his equations of gravity because he could not explain action at a distance with no transmitting media. Einstein considered him great. Am I great? My children and grandchildren think so. Who else matters?

My theory is that reciprocity was never linear. It is an exponential function right from zero. Perhaps this exponential relationship is related to the relationship between current through and EMF across a silicon diode. That also depends on two factors, the reverse voltage threshold which varies from diode to diode, and the forward voltage with an exponent containing Boltzmann's constant and absolute temperature. We see similar relationships when film is subjected to extremely low temperature to sensitize it for astronomical use. In short, I think there is a theoretical basis for an exponential rather than linear reciprocity relationship. I am satisfied for the present with the empirical one.

 

There are two coefficients that nature provides and that we must determine in order to test the theory. Even Einstein left room for an arbitrary constant. Why does the equation fit what data I have seen? That is, as you say, a two part question. The "why" part has been asked from the beginning of mankind and has not yet been answered. The answer to the question "Does it fit?" is "Yes."

To my knowledge, no one else has suggested that form of equation, let alone a universal exponent of 1.62, so I have no one else to accredit. If you know of such a reference that should be made, let me know before I embarrass myself in public.

Patrick;

 

Our reciprocity test paper at EK was on a log scale, and the EK plots were also on a log scale thereby telling us that the data was exponential.

 

As for references, try reading Hamilton or James as cited by Ryuji above, or read Mees or Mees and James. They have something to say about it. I have Mees, so I will read it up tonight to refresh my memory and post more information.

 

Ron Mowrey

Patrick;

 

Here is an answer to you from Mees, "The Theory of the Photographic Process" Revised Edition P201, Reciprocity and Intermittency.

 

Curves corresponding to various suggested equations for reciprocity failure include Catenary, Hyperbolic (Kron), and others by Bunsen and Roscoe. The hyperbolic function includes both HIRF and LIRF, but when calculating LIRF the simplified equation Log It = K - a Log I which in linear form becomes the Schwartzschild equation. There are constants that must be adjusted in these equations for each emulsion and therefore a generalized curve cannot be drawn, nor can a generalized equation be derived other than the one above or the expanded version for LIRF and HIRF.

 

The work by Kron goes back to the Ann. Physik (4), 1913, 41:751.

 

The equation: Log(tc) = log(tc,1) + 1.62 log

 

which you posted previously on APUG therefore appears to come directly from this previously published work which I had read years ago but had entirely forgotten.

 

The most important fact of this work is illustrated in figure 71, and the statement that there have been efforts to derive a general equation for all emulsions, but this has proven impossible. Figure 71 shows the range of equations that corrrespond to various equations and films.

 

Bottom line: Use it but be careful - and it was invented in 1913 by Kron and verified over nearly 100 years by others over the years as being only a rough approximation.

 

I hope that these references help you in your future work.

 

Ron Mowrey

That is very interesting. The fact of the matter is that I had not seen the equations of others, only experimental data and published corrections like thos of Ilford that I referred to. I used this method to analyze the experimental data of Howard Bond that was published in Photo Techniques, July/August isue, 2003 and published the results in the September/October issue. I heard no accusations of plagiarism at that time. I see now why Ilford's data fit the equation so well. Perhaps it is not all data, after all.

 

So you see, when you wondered why the great minds of photography had not come up with such a simple analog, it was because you did not know they had. Now it is one thing to demand scientific accuracy of experiments, and it is quite another to know what that is or means. If there is a dependency of the relationship on temperature, it will very likely be absolute temperature which does not change much where we live.

 

What you are saying in effect is that we cannot take pictures because there are too many factors that effect exposure and development. Forget the fact that there is an exponential relationship or some more esoteric combination of fudge factors. We do take good pictures if we have a good eye for it. Anything we can do mathematically to aid that is worth doing. I would still publish what I have found out about reciprocity even if I were thoroughly immersed in the previous work of others. The title would be different, something like "An Application of Joe Blow's Theory of Reciprocity to Experimental data " It would still be worth knowing if every practical film we have does or does not fit that equation.

 

I did not claim to find data that fit someone's equation that I did not think was my own. I claimed to have an equation that fit someone's data that was in the public domain. I can certainly acknowledge that another had the equation before I discovered it.

 

The polite thing to do would have been to tell me gently that someone else got there before I did.

Maybe I got a little too incensed. I'm not a plagiarist, but I may be more ignorant than the rest of you when it comes to who did what. I have certainly not seen anywhere the treatment I gave the data that I have. If you will tell me whose equation I unwittingly pilfered, I will simply insert "The following equation has been proposed by ..... as more representative of the reciprocal relationship between illumination and exposure time." I have never claimed to have invented it. I used a common ploy among research engineers who are looking for some kind of sense in nature which is to plot things on all kinds of available graph papers, and there are many of those. Each kind has an equation that defines where the lines go.

 

You say use it, but don't depend on it. Does that go as well for Kodak's charts and tables?

"I have certainly not seen anywhere the treatment I gave the data that I have."

 

Did you read the photocopies of review articles I mailed to you years ago?

 

"If you will tell me whose equation I unwittingly pilfered, I will simply insert "The following equation has been proposed by ..... as more representative of the reciprocal relationship between illumination and exposure time.""

 

Well, that's actually *your* homework since you have been posting those things. Ron already gave you where to look, and I'll give you a couple more.

 

"I have never claimed to have invented it. I used a common ploy among research engineers who are looking for some kind of sense in nature which is to plot things on all kinds of available graph papers, and there are many of those. Each kind has an equation that defines where the lines go."

 

But those superficial, phenomenological fit to a certain model equation does not constitute an evidence that the nature obeys that model. These are two different problems, do you see? So,

 

"You say use it, but don't depend on it. Does that go as well for Kodak's charts and tables?"

 

Kodak doesn't say "there is an universal law about reciproicity failure" or "the mechanism of reciprocity failure obeys these tables" or anything like that. They are just giving correction factors that work well enough in actual applications. They keep science and applications rather separate. (Of course, it's easy to sell applications but very difficult to sell science.)

 

Actual scientific understanding is way beyond that level and they are well documented. Ron already gave a pointer for empirical fits, so here are refs for mechanisms.

 

Eachus, R. S. and Olm, M. T. 1989. Aspects of applied inorganic photochemistry: the photographic process. Annual Reports on the Progress of Chemistry, Section C, 86, 3-48.

 

Hamilton, J. F. 1988. The silver halide photographic process. Advances in Physics, 37, 359-441.

 

James, T. H. 1986. Chemical sensitization, spectral sensitization, and latent image formation in silver halide photography. In Volman, D. H., Gollnick, K. and Hammond, G. S. eds. Advances in Photochemistry, vol. 13. 329-425.

 

Sturmer, D. M. and Marchetti, A. P. 1989. Silver halide imaging. In J. Sturge, V. Walworth, and A. Shepp, eds., Imaging processes and materials: Neblette's eighth edition, New York: Van Nostrand Reinhold.

 

Tani, T. 1995. Photographic sensitivity. (Oxford series in optical and imaging sciences, vol 8.) New York: Oxford University Press.

 

These are a book, book chapters and review papers. There are shorter reviews in journals as well. But if you grasp the basics of latent image formation, you'll appreciate that those simple models are only a aggregate phenomenon observed from a much more comlex stochastic system and there are a lot of factors overlooked by such an emprical fit.

 

I only know those authors as smart guys who did great work, but Ron probably knows most if not all of them personally. (I think all but Hamilton and Tani are with EKC.)

 

I think Mitchell also wrote a lot on this topic but can't think of a good review article right now...

"Now it is one thing to demand scientific accuracy of experiments, and it is quite another to know what that is or means. If there is a dependency of the relationship on temperature, it will very likely be absolute temperature which does not change much where we live."

 

I suggest you say that after seeing the data. James did a lot of work on environmental effects on sensitivity. Some of those variables studied are used as means of hypersensitization.

 

"We do take good pictures if we have a good eye for it. Anything we can do mathematically to aid that is worth doing. I would still publish what I have ���found out about reciprocity even if I were thoroughly immersed in the previous work of others."

 

The problem is that you have taken your empirical observation and stated them as if they represent the universal mechanism underlying reciprocity failure.

Other than that, I wouldn't say much about restatement of what's known.

Being unaware of previous work that is printed in well known textbooks like Mees is embarrasing, but not as bad as illogical arguments.

 

And one can certainly take pictures without knowing all these. So can he take pictures without fallacious statements.