Short duration underexposure reciprocity failure explanation please

[Christopher Walrath]

Wirelessly posted (BlackBerry9000/4.6.0.167 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/102 UP.Link/6.3.0.0.0)

Ian brought something I had NEVER before considered. Could someoe please explain to me how and why there would be reciprocity on underexposures of short duration? Would this occur with small apertures and s/s less than one second as well with general films?

 

[Mike Wilde]

My sort of non sceintific short duration exposure reciprocity explanation- you need something like 4 electrons to hit the film spot and transfer their energy to turn a halide to a pending elemental state (latent image formation). When exposires are really short, there is not a good cahnce of 4 electrons to make it though while the shutter is very briefly open.

My long duration reciprocity failure explanation - some of the electron charges recieved fade away before enough come together to give the 4 electron equivalent and make the speck suitable to turn metallic once developed.

 

[Photo Engineer]

Chris;

There is both LIRF (Low Intensity Reciprocity Failure) and HIRF (High Intensity Reciprocity Failure). Any film can fail under low intensity light with long exposures or at high intensity light with short exposures.

PE

 

[Q.G.]

Ian brought something I had NEVER before considered. Could someoe please explain to me how and why there would be reciprocity on underexposures of short duration? Would this occur with small apertures and s/s less than one second as well with general films?

It's called the "intermittency effect", and occurs when exposure times get very short. You may find info on the net about that if you search using that as key words.
Not very noticeable when you add exposures in the "will our shutters do" range.

It's not a case of "reciprocity", by the way.
Rather of "reciprocity (law) failure". Yet that only indirectly.
So it would perhaps be better to say that it has nothing to do with reciprocity at all.

(It's neither the low nor high intensity effect PE mentioned, but the phenomenon that when exposures get short, the sum of such short exposures is not equal to one single exposure of the same total duration.)

 

[MattKing]

Chris:

I assume you are asking about reciprocity failure when using shorter (not shirter ) exposures.

You probably know all or most of this, but here goes:

The effect that light has on light sensitive materials is generally a function of exposure, which is actually made up of two components:

1) the intensity of the light striking the materials, and
2) the length of time that the light strikes the material.

For most materials we work with, in the ranges we most commonly work with them, to a very great extent the light intensity and duration work in a reciprocal relationship - i.e. if you double the duration and halve the intensity, the exposure is the same, and the resulting density on the material is the same as well.

This reciprocal relationship only applies however over a range of light intensities. If you go outside that range, and the intensity is either so much greater, or so much less than the more commonly encountered intensities, then the corresponding change in the duration of the exposure won't have sufficient effect to result in the same density on the light sensitive materials - i.e. we have reciprocity failure.

We generally tend to speak in terms of duration of exposure when we discuss reciprocity failure, because we are usually faced with a given light level (which we cannot control) and we are attempting to deal with it by adjusting the duration of the exposure (which we can control).

All this is pretty well understood by most here, but we sometimes overlook the following:

a) no light sensitive material evidences perfect reciprocity over any range - it is always at best close; and
b) there is no such thing as a perfect shutter or iris aperture - all exposures involve at least some change in intensity over the range of an exposure.

In the example Ian referred to in the other thread, wherein there were several exposures of 1/5 of a second, the difference due to reciprocity failure between the exposure at 1/5 of a second at one f/stop and the exposure at 1/10 of a second at the next f/stop might be too small to be obvious, but it would be real. If one used 5 such exposures, then the differences would add together and may as a result become quite significant.

There are other factors to keep in mind however. In that example, the (in)accuracy of the shutter or the iris diaphragm of the lens may very well have a larger effect.

Matt

EDIT: As indicated above in Mike Wilde's and Q.G.'s posts, there is another effect that one encounters with very short exposures - not the ones in the 1/5 of a second range that I assume Chris was asking about

 

[Q.G.]

Good explanation of reciprocity.
Also a good explanation of why the intermittency effect (the thing alluded to in the thread about multiple exposures) has only very indirectly to do with the assumptions that led to the reciprocity law.

In reciprocity law issues, both Matt's #1 and #2 are supposed to be in balance, and problems/failure occurs when they no longer are.

In the intermittency thingy, intensity is not a variable. The #1 and #2 here are total duration and frequency, or rather 'helpings'.

 

[richard ide]

IIRC some Literature (I believe from Kodak) gave minimum exposure information. Exposures of very short duration were not sufficient to generate a latent image in the emulsion.

 

[Christopher Walrath]

Wirelessly posted (BlackBerry9000/4.6.0.167 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/102 UP.Link/6.3.0.0.0)

Well, I am indeed referring to multiple short exposures. For instance, I have 140mm lens on my 4x5 that has s/s's of 1/100, 1/50, 1/25, B and O for focusing. If my metering requires a 1/6 second exposure. In order to acheive this, I would need to expose the scene at 1/25 about four times plus a fraction more to get the same exposure. But by doing so I am making four exposures that were two stops underexposed each. What if I need to acheive 1/3 of a second? That would be eight exposures underexposed by three stops each.

By my understanding this repeated underexposure might add up to deeper shadows due to the inability, or rather inaccuracy, of the photons to repeatedly strike all of the halides in a local area of the negative in a uniform manner, thus providing increased contrast in the higher negative density values and evident in the lower print values as well.

Did I get it right? More or less?

 

[Photo Engineer]

I did not consider that you meant the intermittency effect.

In this latter effect, 10 exposures of 1/100 second does not equal 1 exposure 0f 0.1 second. This means that several short exposures are not additive and do not equal an exposure equal to the sum of all of the exposures. It is due to two effects, one rather esoteric involving the formation of the latent image and the other due to straightforward reciprocity failure. So Intermittency is the approximate "sum" of two effects.

PE

 

[ic-racer]

I don't have anything else to add to what Matt King wrote but sometimes a picture is worth a thousand words. This 3d graph shows the effects of both very short exposures and the effects of very dim light (This is an older diagram, the Log E axis would be Lux in SI units). The areas where useful "reciprocity" exists are on the two slopes A and B. The rest of the points on the surface show little or no reciprocity and are not useful for most photographic purposes.

 

[Photo Engineer]

Very nice reference.

I would like to point out that the curvature n the density axis as a function of exposure also leads to contrast reciprocity, a subject seldom mentioned.

We know that reciprocity can cause us to lose speed as a function of light intensity, but it can also cause us to observe moderate to severe changes in contrast over the same range. This is known as contrast reciprocity failure.

In a single emulsion film or paper, this may not be a severe problem, but in a product with emulsion blends and in which the emulsions don't respond equally, you begint to get bumps in the characteristic curve.

PE

 

[Christopher Walrath]

Wirelessly posted (BlackBerry9000/4.6.0.167 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/102 UP.Link/6.3.0.0.0)

Thanks a lot, folks. I have always said that, though I may have a certain 'mastery', I will never know it all. But I now know a little more than I did. And that is what it is all about. You guys rock!

 

[naeroscatu]

Chris, thank you for digging deeper into this subject and thanks to all who contributed. Very instructive thread.

 

[smieglitz]

Well now I'm curious. First, by "underexposures of short duration" I would think more in terms of electronic flash exposures like 1/10,000 sec rather than 1/3 sec or longer (which would be closer to the threshold of long duration exposures -whatever they are). Second, it seems several people have described the results of the reciprocity failure effect (i.e., less effective exposure at certain exposures outside a given range) , but haven't really explained why it happens.

Does it have something to do absorption or loss of photons in the latent image (as Mike Wilde has suggested) or is something else going on? I thought it had something to do with the wavelength of light, surface area of silver halides, and probability of photons being in the right place at the right time. Isn't that why T-grain films and higher-speed emulsions have reduced reciprocity effects compared to other films i.e., a higher probability that larger grains or flatter grains oriented a certain direction will be struck by enough requisite photons in a given amount of time?

What was done to make, for example, Vericolor type-L have a different reciprocity range than other standard negative films?

 

[Photo Engineer]

Smieglitz;

Well, I'm not the person to give a detailed answer on your question. There are others better qualified.

I do know that several things are going on based on grain size variations within an emulsion. The more polydisperse it is, the more non-uniform the added chemicals are in distribution and therefore the more non-uniform their action on the grains. Most of the chemistry in emulsion making is surface related in one way or another and the finer grains have more surface to volume area and adsorb more chemistry based on how many are present.

Therefore, the effect varies from toe to shoulder and that is a function of grain size.

As for Vericolor-S and L, yes, they used different chemical addenda and emulsions to adjust the ranges of reciprocity for short and long exposures. I'm not sure what they used as they had been discontinued when I worked on consumer negative films in the 80s.

Today, the monodisperse emulsions tend to even out these problems by adsorbing chemicals more evenly in a given emulsion, and modern dopants such as Iridium and Osmium tend to minimize or completely eliminate reciprocity failure in most cases. In fact, it is even very difficult to get a modern emulsion to solarize for much the same reason.

PE

 

[JBrunner]

Very nice reference.

I would like to point out that the curvature n the density axis as a function of exposure also leads to contrast reciprocity, a subject seldom mentioned.

We know that reciprocity can cause us to lose speed as a function of light intensity, but it can also cause us to observe moderate to severe changes in contrast over the same range. This is known as contrast reciprocity failure.

In a single emulsion film or paper, this may not be a severe problem, but in a product with emulsion blends and in which the emulsions don't respond equally, you begint to get bumps in the characteristic curve.

PE

I have noticed something similar to this effect in very long exposures with my 8x10, where I calculate the reciprocity failure into the exposure, and well lit parts gain more density compared to darker areas which expose normally. In essence, I'm getting expansion without processing for it.

 

[Photo Engineer]

I have noticed something similar to this effect in very long exposures with my 8x10, where I calculate the reciprocity failure into the exposure, and well lit parts gain more density compared to darker areas which expose normally. In essence, I'm getting expansion without processing for it.

This is contrast reciprocity failure in action.

PE

 

[greybeard]

As a point of reference, the older Kodak literature stressed that development should be lengthened when using electronic flash of very short duration, such as that offered by the original Kodatron units. The flash duration was described as being in the 1/25,000 second range, much shorter than a typical studio flash today. Some of the "thyristor" flashes may do this today, but not a 1200 watt-second power pack.

For what it is worth, reliably measuring a quarter of a millisecond with the instrumentation of the 1940s was not a trivial task, so the duration figures are somewhat open to question unless someone has confirmed them with modern gear. At least one citation that I found from 1941 gives a duration of 1/3000 second, which is improbably low given the design of the unit (2000V oil-filled storage capacitor).

 

[Photo Engineer]

There was even a journal for that topic alone. "The Journal of High Speed Photography".

And, the old oil-filled capacitors used to go dry leading to failure.

PE

 

[Stephen Benskin]

According to Photographic Materials and Processes, "When exposures are made at high illuminance, with corresponding short exposure times, the electrons are released so rapidly that the relatively slower-moving silver ions cannot neutralize them fast enough for the centers to grow to the predicted size. A greater number of sub-image centers are spread over the crystal surface, and sometimes into the crystal's interior. There is more competition for the additional silver atoms formed, and thus a smaller number of them grow large enough to become permanent developable latent-image centers."

The above is related to the Gurney-Mott hypothesis of latent image formation. Also, high energy reciprocity failure decreases contrast as opposed to low intensity reciprocity which increases the contrast.

 

[John Koehrer]

The discussion is enlightening, but from a practical/perception view will 5-125th sec exposures be perceptibly different than 1-1/5th sec. exposure. Enough to say, require an additional 125th sec exposure?

 

[Q.G.]

The discussion is enlightening, but from a practical/perception view will 5-125th sec exposures be perceptibly different than 1-1/5th sec. exposure. Enough to say, require an additional 125th sec exposure?

Yep. Quite different.

5x 1/125 will be about 2.3 stops underexposed, compared to 1x 1/5.

5/125 = 1/25