Multiple exposures / reciprocity with Velvia 100F

[DrPablo]

I'd like to reproduce a very interesting photo that I took digitally.

It requires 5-7 separate exposures of about 30 seconds each. I quite laboriously combined them in PS when I shot it with my DSLR. Now I'd like to do the same thing with my 4x5 using multiple exposures on a single sheet film.

I have oodles of Velvia 100F lying around, and I think it would work very well for this kind of shot. But we're talking about a total exposure time of 3-5 minutes. So calculating exposure will be tricky, and I worry that I'll run into color shifts.

Do you think this is reasonable to try with Velvia 100F, or is there a better film for this purpose? And how will I calculate the exposure -- if I plan on, say, 8 exposures, do I just meter in Tv mode for 4 minutes and divide it into 8 separate shots?

[Ron Marshall]

Astia needs no correction for reciprocity up to two minutes.

[Kirk Gittings]

My experience from years of doing multiple exposures seems to indicate that reciprocity is not cumulative. For instance the reciprocity correction for 4 one second exposures is not the correction for a four second exposure.

[DrPablo]

But if even a 30-second exposure can have reciprocity failure, then what about several 30-second exposures?

[Kirk Gittings]

Say 4/30 second exposures. Each 30 second exposure has a calculated RF. But it is different than calculating the RF of a 120 second exposure.


The real difference comes into play with the marginal exposures (where RF just starts to come into play with a given film). A film may have no reciprocity failure problems at 10 seconds say, but big problems at 60 seconds. In this scenario 6/10 second exposures will still have NO RF, because RF is NOT cumulative, even though the total exposure is 60 seconds. Does this make any sense? This has been my experience that has guided me for many years in this sitution with many films.

[DrPablo]

Yeah, that makes sense.

So say I calculate exposure for 8 30-second shots; i.e. 4 minutes of exposure time will be a proper exposure, aggregated over 8 separate clicks of the shutter.

It would then be fair to assume that reciprocity would only be an issue insofar as it might happen in 30 seconds, rather than 4 minutes?

[Jorge Gasteazoro]

I think what Kirk is trying to say is that RF does not have a linear behavior, if I am correct then he is right. In simple terms light needs to have enough energy to remove electrons from the silver crystal, when we find ourselves in the recirpcity area what is happening is that the light does not have enough energy to "push" the electron. The electron gets exited, but immediately falls back to it's energy level, thus we need to give the light more time to keep "pushing" the electron out of the energy level. The lower energy the light has, the more time we need to give it to push the electrons so that they dont fall back, as I said this is not a linear relationship.

In these situations what needs to be done is to calculate the RF for the TOTAL exposure time and then if you wish you can break it up in parts. For example if you have an exposure that requires 4 minutes, and then when you add RF it turns out it needs 15 minutes, you can break the exposure in 3 five minute exposures. You cannot say I will make 4 one minute exposures and calculate the RF for a 1 minute exposure, you will get an underexposed sheet.

[roteague]

Here is the datasheet for Velvia 100F: http://www.fujifilmusa.com/JSP/fuji/epartners/bin/RVP100FAF3-148E_1.pdf

No compensation is needed to 1 minute, with less than one stop loss to 8 minutes.

[Kirk Gittings]

Jorge, sort of. Lets say at 4 sec there is no RF on a film. 4 exposures at 4 sec will still not have RF.

[Kirk Gittings]

what is happening is that the light does not have enough energy to "push" the electron. The electron gets exited, but immediately falls back to it's energy level, thus we need to give the light more time to keep "pushing" the electron out of the energy level.

I believe this phenomena repeats itself anew with each separate multiple exposure with no memory or cumulative RF affect. I have used this approach successfully for years on daylight chromes of all types (EPP EPN Velvia Velvia 100F etc.).

[Jorge Gasteazoro]

Jorge, sort of. Lets say at 4 sec there is no RF on a film. 4 exposures at 4 sec will still not have RF.
I think you mean 4 exposures of 1 sec..... of course, if the calculated 4 sec exposure did not need RF correction then 4 one second exposures should not need them either.

The problem comes when you are in the 40 to 50 min exposures and you need to close the shutter. What do you do? In this case two 25 min exposures will work just as well as one 50 min exposure or better illustrated, five 10 min exposures will wok just as well as one 50 min exposure.

Of course, color film is much more demanding than B&W which is where I have most of my experience, but the theory should hold. And then, we have not discussed color crossover..but I guess that is a topic for another thread.

[keithwms]

In my understanding of things, the RF effect actually is cumulative but when one does multiple exposures, the usual tendency is not to overlap areas of high redundant exposure, so the cumulative effect is usually not an issue. As a case in point I would cite star trails; notice that there is no colour shift along the trails themselves, but there is a colour shift in the background of the whole sky which can overlap the trails, and the colour shift is certainly due to RF. Velvia shifts green and 64T shifts blue, as I recall.

Exposure is, of course, quite generally a cumulative effect in terms of how the film reacts photochemically. If it were not, we wouldn't be able to rely on the "stop" system of apertures and shutter speeds to prescribe correct exposures. RF is the nonlinear failure of the usual relationship between exposure time and (developed) film density upon which we intuitively rely.

My guess concerning the actual mechanism of RF is that there is depth-dependent photochemistry going on; incoming photons have a finite penetration depth into the film and so the surface of a particular film layer is dosed faster than the bulk of the film. Hence the topmost film sublayer doses up quicker and starts to behave like a non-neutral density filter after long exposures. This causes the backside of the film (or underlayers) to get less exposure per time. Probably there is more to it than that but that's my simple view.

So I think it is all a matter of light attenuation. I think colour shifts result because the coloured layers are stacked and, very generally speaking, reds will transmit further than blues. This must be why the red layer is usually the lowest one in the layer stack. Of course if you look at a cross sectional view of film it is actually very complicated and there is separate photochemistry in each sublayer so there are many effects that can go into RF.

The situation is entirely different in the case of digital, in which "wells are emptied," so to speak, after each exposure. A digital sensor's long-term memory is in the cache and the card, not in the photosite. So Paul, if your particular shot requires that there be no reciprocity failure due to cumulative exposure in a certain part of your frame, then I think you are going to have to get creative, i.e. employ GNDs, or use colour filters, or in the worst case expose separate sheets and then combine later after scanning. But I can't imagine reciprocity failure being such a dominating effect for most films. I mean, the newest fuji 64T is really great in terms of RF, it's way better than tri-x, for example:

http://www.mkaz.com/photo/tools/reciprocity.html

Most b&w films are horrendous in terms of RF but that is another topic.

So that's my two cents on the matter!

Keith Williams

[Kirk Gittings]

the usual tendency is not to overlap areas of high redundant exposure,

This doesn't hold up to my experience. Multiple exposures are stock in trade for architectural photographers shooting interiors with strobe to "multiply" the flash power. When I am doing this (and I do it nearly every day) eveything in the frame is overlapped every time.

[keithwms]

This doesn't hold up to my experience. Multiple exposures are stock in trade for architectural photographers shooting interiors with strobe to "multiply" the flash power. When I am doing this (and I do it nearly every day) eveything in the frame is overlapped every time.


For a week I thought I brought the whole site down with my last rambling response!

Anyway, Kirk, I think we have to distinguish between flash illumination and cw lighting. The former is of much higher peak intensity spanning a much shorter time than one gets with cw lighting.

To relate the two would require a little math, you'd have to compute the equivalent average exposure for cw lighting. And then of course reciprocity only enters the picture after minutes of equivalent, average cw exposure. How quickly you reach "minutes of equivalent exposure" depends on the power of your flash- it could happen after one flash or after 100, that all depends on the power of the flash, the flash duration, and the reflectivity of the subject. What most assuredly does not matter is how much dark time your film experiences, i.e. time with the flash off or with the shutter closed or with a darkslide in. Time only enters these reciprocity discussions because people need to take mW/cm^2 and multiple by exposure time to convert to joules.

Now, if you look up the fuji info sheet for provia 100F on the topic of flash compensation, all you get is some vague comment:

"No exposure correction or color balance compensation
is required for up to eight consecutive multiple exposures
using an electronic flash."

Hmmm... thanks Fuji, but at what power? And what pulse duration?! To do a proper reciprocity correction for multiple flashing and try to reconcile that with usual the cw corrections, one would need to compute actual joules delivered to the film at a particular wavelength. Watts alone won't get you to the right answer- it is the time-integrated exposure that controls the photochemistry... how many molecules are hit by how many photons. That holds true until you get to extremely high fluences, which you do not reach with a flash. I reach that limit quite often in my lab but it isn't something that you should encounter in ordinary photography. So, total energy delivered to the film is the important thing.

Keith

[400d]

I think what Kirk is trying to say is that RF does not have a linear behavior, if I am correct then he is right. In simple terms light needs to have enough energy to remove electrons from the silver crystal, when we find ourselves in the recirpcity area what is happening is that the light does not have enough energy to "push" the electron. The electron gets exited, but immediately falls back to it's energy level, thus we need to give the light more time to keep "pushing" the electron out of the energy level. The lower energy the light has, the more time we need to give it to push the electrons so that they dont fall back, as I said this is not a linear relationship.
OK, since you mention the energy level, but from a physics perspective, how long it will take for the energy to drops back to the ambient level??

[keithwms]

OK, since you mention the energy level, but from a physics perspective, how long it will take for the energy to drops back to the ambient level??

The cycle of electronic excitation and de-excitation is over in an extremely small time compared to exposure... i.e. probably nanoseconds. The light absorption takes place in the sensitizer and the de-excitation goes through the halide, as I recall. The reduction of the halide to the noble form is what leaves us with a recording on the film that can then be 'developed.'

[Kirk Gittings]

Keith, I am not sure how to answer you. You are way over my head but does this matter......?

My work always combines flash with a long ambient light exposure. In a given room we may have 8000 watts of power in 5 heads firing 4 times each with a 1 to 4 second ambient light exosure. How might the combination of flash and a long ambient light exposure explain the phenomena that I experience that reciprocity is not cumulative?

[keithwms]

Keith, I am not sure how to answer you. You are way over my head but does this matter......?

My work always combines flash with a long ambient light exposure. In a given room we may have 8000 watts of power in 5 heads firing 4 times each with a 1 to 4 second ambient light exosure. How might the combination of flash and a long ambient light exposure explain the phenomena that I experience that reciprocity is not cumulative?

Kirk, I am most assuredly not over your head ;)

Perhaps the way to reconcile this is to recall the nonlinear nature of the way the halide grains are stabilized. The effect is cumulative but also nonlinear.

It is not at all clear to me that the intense pulses from a flash and the less intense, more protracted exposure from ambient light both lead to the same photochemical reduction rates. Sure, the photons have roughly the same energy, that is just derived from the wavelength components. But the number of joules of energy delivered to the film per time is totally different. That may be important because with silver halide, it's not that one photon comes in and switches one halide grain to one silver grain and you have a single "pixel." On the contrary, it takes several photon hits to stabilize the silver particle, the process is nonlinear, as previously noted. So if you have higher intensity then you can have totally different reduction rates in the halide than you may have with less intense, ambient light. So I am abit leery to apply the same reciprocity logic to your situation that I would apply to, say, long exposures with ambient light or star trails or whatever.

Bottom line: experience trumps theory, no doubt about that! But I do think there is a difference between long, ambient light exposure and building up an exposure with a sequence of shorter, more intense pulses. My first response really had the latter in mind.

Keith

[Andrew O'Neill]

I'm with Kirk on this one. Once I made 9 exposures on one sheet of film. Each exposure was 1 second long. RF compensation was +1/3 stop for the 9 exposures. Each exposure was 1 second at f/16 2/3.
If I make one, 9 second exposure then to compensate for RF I would have to actually give an exposure of 16 seconds.
So I think it is safe to say that 9, 1 second exposures isn't the same as 1, 9 second exposure as far as RF is concerned...in my experience, anyway.

[keithwms]

I'm with Kirk on this one. Once I made 9 exposures on one sheet of film. Each exposure was 1 second long. RF compensation was +1/3 stop for the 9 exposures. Each exposure was 1 second at f/16 2/3.
If I make one, 9 second exposure then to compensate for RF I would have to actually give an exposure of 16 seconds.
So I think it is safe to say that 9, 1 second exposures isn't the same as 1, 9 second exposure as far as RF is concerned...in my experience, anyway.

Ah, I am not disagreeing with your experience!

If you just look again at the very first link that I provided, your experience is immediately explained:

http://www.mkaz.com/photo/tools/reciprocity.html

Look at the first film, fuji 64T. If you meter a 4 sec exposure, your actual exposure should be 5 sec.

But if you want to do two 4 sec exposures (to make an 8 sec total exposure), then the reciprocity corection is not just 1*2=2 sec. According to the chart it is 4 sec, twice as much as you'd expect just by adding the individual reciprocity corrections for each 4 sec exposure.

My point about the effect being cumulative is that you have to consider the net exposure to the film, that is what matters when you compute the correction. You have to add up the total exposure time and then tack on the recip. correction for that. It makes no sense to work out reciprocity corrections for each subexposure and then add them, because the addition is nonlinear.

I tried to explain why it is nonlinear and cumulative but I guess that is too much in the details of the photochemistry. Sorry if that was too tangential.

Keith