EI and Reciprocity.

[Athiril]

I have a simple question, want to know if Im doing it right.

I expose my film at night at a certain EI, and do not compensate for reciprocity, lets assume that exposure is 'perfect' for that EI apart from reciprocity, lets assume that EI is faster than the box speed of the film and I am pushing.

So if my box speed is 100, my EI I chose is 400, my exposure @ 400 is 1 minute.

Then I'd use the 4 minute mark for calculating number of stops to add for reciprocity? Lets say 4 minute mark is 2 stops, and 1 minute mark is 1 stop, I'd add 2 stop and expose for 4 minutes EI 400, and not 2 minutes EI 400?

Reason I am asking, is because I pushed some Tri-X quite far to see how it would handle... one of the shots I did not compensate for reciprocity, and the mark for reciprocity based off converting the exposure time back box speed and use the difference in stops on top of my EI, puts it at a bit over 3 stops.

The mark for the time without converting down was 3 stops flat anyway in this case.. so either way would have been similar..


So would I say I underexposed the film 3 and a bit stops from my chosen EI? I still got a picture out of it, albeit grainy..



The reason I am asking is because I pushed the Tri-X in Rodinal to 6400, shot at EI 6400, one of the exposures that counted for reciprocity looks pretty good, the other I did not (exposure was 2 min, with reciprocity counting it the way I assumed, then it would be ~25 min) and has rough as guts grain (6x7cm neg), if I count the number of stops of reciprocity I did not compensate for as the number of stops I underexposed under 6400... then it puts the EI for that shot @ 80,000, which I find absolutely ridiculous figure, and ridiculous that its giving me a picture at all.


Basically, what I want to know is this:

If I fail to account for x stops in reciprocity compensation, do I count that as being x stops underexposed?


I want to work out how much underexposure in stops this shot has had, it was 2 minutes @ EI 6400 - that being the 'perfect' exposure for the scene for EI 6400 without compensating for reciprocity (fresh Tri-X 400)

 

[photomem]

I thought I understood reciprocity, but you just gave me a math migraine. Ow ow ow.

 

[Joe VanCleave]

I think reciprocity correction is based on a fractional exponential function, one derived from curve-fitting actual data collected from many standard types of films. I don't remember where I've seen the formula, but hopefully someone will chime in with it.

I don't think it's based solely on "X" number of stops. The amount of correction depends on the initial, uncorrected time; the large the uncorrected time, the corrected time grows exponentially larger.

For example, someone recently posted an 8x10 pinhole camera image captured on Tri-X; their camera's f/stop was somewhere around f/300, and resulted in an exposure time of almost a minute. By comparison, I shoot paper negatives in pinhole cameras of the same f/stop ratio, which I rate at EI=12, and my exposure times are about the same as the (supposed) ISO320 film. This is because the reciprocity failure is so severe in the case of the Tri-X that the photo paper (which is very linear) ends up being faster than Tri-X and many similar films in extremely low light levels. Sorry I don't have any quantitative numbers to share, but that's all I can manage right now.

I understand that Fuji Acros is very much more linear than more traditional films.

~Joe

 

[Ian David]

The amount of correction depends on the initial, uncorrected time

Right. So, Athiril, if your light meter says you need 1 minute exposure, you will need, due to the length of that exposure, to add on further time to account for reciprocity failure. If you fail to add on that extra time, your frame will be underexposed accordingly.

So, if you are shooting ISO100 film at 400, then you are already underexposing by 2 stops. If you then ignore reciprocity failure, there will be further underexposure. Depending on the figures for your particular film, you may be able roughly to approximate that further underexposure as 1 extra stop, and then try to compensate in processing as if you had underexposed by a total of 3 stops.

But I would normally think of reciprocity failure in terms of a specific necessary time correction, rather than a number of stops, because reciprocity failure is really the breakdown of the usual relationship of doubling/halving that stops describe.

Ian

 

[MPandolfo]

If I fail to account for x stops in reciprocity compensation, do I count that as being x stops underexposed?

:confused: Good question! Don't know the answer, but my guess is yes. I think you need to experiment yourself...

 

[Shadowtracker]

And here comes the sacrificial sheet film now....

I don't have an answer, but I know I have read this; Probably in Adams' The Negative if memory serves me, which, lately, it hasn't been.

 

[henpe]

Maybe this article can help!?
http://www.unblinkingeye.com/Articles/LIRF/lirf.html

 

[Ross Chambers]

I've been thinking about this reciprocity thing lately, Fomapan being rather extreme in that quality. However I've never seen an example of "hey mate, look at that reciprocity, it's off the scale" unlike under/overexposure, or over/under development.

Can someone tell me what reciprocity failure looks like? I pretty much guess the EI (usually as time rather than aperture) when I'm hitting 2 or 3 seconds, and for my purposes--salt printing--it works OK.

Regards - Ross

 

[Sparky]

Reciprocity law failure has ONLY to do with time of exposure. If you're in reciprocity law failure territory - then it means you have to compensate. If your film loses reciprocity at one second - then it's completely irrelevant what the box speed is. HOWEVER - where things get messy - is that any time you're compensating for RLF - you're ALREADY pushing your film significantly. Because the highlights will record to the film more efficiently than the shadows - your contrast goes up. It's a VERY slippery slope. If you want to TRULY compensate for reciprocity effects to a more accurate degree - you need to drop your dev time a whole bunch - decrease more the longer your exposure is. But this, in turn, requires dropping your effective film speed again! So what was a five second exposure becomes, say THIRTY seconds just for reciprocity law failure exposure compensation but to get the same kind of contrast you'd get at 1/8 sec, you may need to drop your development by, say 25% - which would require you to again compensate the exposure to something like two minutes... which again would require further development compensation, albeit less. All you can do is target a development time/exposure time pair that you know works, and stop down to the aperture that will give you a good result. With a hypothetical film - that means dropping the box speed from, say, 125 down to something like 20 or even less. I know it's kind of confusing - but you'll get the hang of it...!

 

[Athiril]

Hmmm... so by exposing at 2 minutes... I would have to look at what reciprocity would correct to 2 minutes for Tri-X in that setting/amount of light, count that number of stops.. and find that as the effective EI.. etc.. maybe...

I just want to work out how many extra stops it was under exposed not including my pushing it to 6400, in any case, its phenomenal that it gives a picture... my 6400 corrected for reciprocity is very fine grain compared to the uncorrected one.

EG, if I were to test if I were to get a picture from EI 80,000 to prove that you use the stops difference from uncorrected towards corrected time as the amount that its underexposed, I'd need to shoot EI 80,000 in a non-reciprocity scenario... in Sunny-16 with like 6 stops of ND (which I have) and 1/400th and f/32 should do it.. by 1/400th EI 80000 converted to shutter speed in EI 400 should be -just- away from reciprocity... I'll test that at mid day tomorrow

Even though I shot 6x7, it was Tri-X @ 6400 pushed in Rodinal.. Im surprised at the grain.. and the further underexposed (how ever many stops it works out to be) to come out at all

I'll read the articles and see if i can get a copy of The Negative later guys, Im going out to watching cartoons at my friends house

 

[Rick A]

Different films have different rates of failure. Some film manufacturers list that information somewhere in their literature, others just let you guess. Multiplying your exposure time to compensate for reciprosity can be anything from (1.2x) on up to (2x) depending on the specific emulsion, also it can kick in at relatively fast times, say around .5 second. In the old days, we would guess at it and off-hand start with (1.5x)at one second exposures or even doubling just to be safe, then bracketing. Take notes of your exposures for future reference, then you wont have to guess any more.

 

[Marco B]

Athiril,

If you want to give yourself a headache calculating EI or trying to push negatives shot in reciprocity failure range, that's fine with me... but if your intend, like many, is to get good night time negatives having both shadow detail and controlled highlights without to much hassle, than pushing is the last thing you should do...

Like Sparky commented, when you get in reciprocity failure exposure times (over 1 second for most films, except a few films like Neopan Acros 100), film tends to give higher contrast results than normal. By pushing the film, you are only going to make things worse...

Pushing is only a suitable thing to do when in non-reciprocity failure zone, e.g. to make a 1/10 of a second exposure, that would require a tripod, one that can be shot hand held at 1/40 of a second by a two stop push.

The right thing to do for night time photography is to pull develop in combination with overexposure.

Now to make matters more easy for you, artificial light sources like lamp poles, generally give of a quite uniform amount of light, which means that only a couple of exposures are "valid" at night time, the range is far smaller than at daytime.

Generally speaking, I think for TriX at night time in an illuminated town, exposure times of 10 seconds - 30 seconds at F8 - F11 will cover most situations. Combine this with a pull development, and you should get some nicely printing negatives.

To illustrate this, here are some examples of my recent night shots of Haarlem with full info:

SOFT LIGHT:
Film: Kodak TriX 400
Exposure: 30 seconds, F8
Development: D76, dilution 1:1 (stock to water), 20C, 6 3/4 minute
Paper: Kentmere Fineprint VC Glossy, printed at grade 1 (condenser enlarger)

MEDIUM LIGHT:
Film: Kodak TriX 400
Exposure: 10 seconds, F9.5
Development: D76, dilution 1:1 (stock to water), 20C, 7 1/2 minute
Paper: Kentmere Fineprint VC Glossy, printed at grade 1 1/2 (condenser enlarger)

STRONG LIGHT:
Film: Kodak TriX 400
Exposure: 30 sec, F11
Development: D76, dilution 1:1 (stock to water), 20C, 6 3/4 minute
Paper: Kentmere Fineprint VC Glossy, printed at grade 1 (condenser enlarger)

Marco

 

[Rick A]

Athiril,

If you want to give yourself a headache calculating EI or trying to push negatives shot in reciprocity failure range, that's fine with me... but if your intend, like many, is to get good night time negatives having both shadow detail and controlled highlights without to much hassle, than pushing is the last thing you should do...

Like Sparky commented, when you get in reciprocity failure exposure times (over 1 second for most films, except a few films like Neopan Acros 100), film tends to give higher contrast results than normal. By pushing the film, you are only going to make things worse...

Pushing is only a suitable thing to do when in non-reciprocity failure zone, e.g. to make a 1/10 of a second exposure, that would require a tripod, one that can be shot hand held at 1/40 of a second by a two stop push.

The right thing to do for night time photography is to pull develop in combination with overexposure.

Now to make matters more easy for you, artificial light sources like lamp poles, generally give of a quite uniform amount of light, which means that only a couple of exposures are "valid" at night time, the range is far smaller than at daytime.

Generally speaking, I think for TriX at night time in an illuminated town, exposure times of 10 seconds - 30 seconds at F8 - F11 will cover most situations. Combine this with a pull development, and you should get some nicely printing negatives.

To illustrate this, here are some examples of my recent night shots of Haarlem with full info:

SOFT LIGHT:
Film: Kodak TriX 400
Exposure: 30 seconds, F8
Development: D76, dilution 1:1 (stock to water), 20C, 6 3/4 minute
Paper: Kentmere Fineprint VC Glossy, printed at grade 1 (condenser enlarger)

MEDIUM LIGHT:
Film: Kodak TriX 400
Exposure: 10 seconds, F9.5
Development: D76, dilution 1:1 (stock to water), 20C, 7 1/2 minute
Paper: Kentmere Fineprint VC Glossy, printed at grade 1 1/2 (condenser enlarger)

STRONG LIGHT:
Film: Kodak TriX 400
Exposure: 30 sec, F11
Development: D76, dilution 1:1 (stock to water), 20C, 6 3/4 minute
Paper: Kentmere Fineprint VC Glossy, printed at grade 1 (condenser enlarger)

Marco

Excellent work Marco. Your work shows that you have done your homework, seems logic and good notes prevail.

 

[Athiril]

I've already shot it.. I just wanted to know what the effective EI is of the non-corrected shot, as I want to know at what EI I can abuse Tri-X with to still give me a picture, no matter if its thin and sandpaper country.

My EI 6400 shots (corrected) came out very nice.

I know how to correct for reciprocity, and I do often overexpose and pull, or add restrainer or dilute etc, highest I've done is a 14 stop pull on colour film (I bracketed a massive range on the film whe I was doing custom processing tests to see what the developer was doing), I simply chose not to correct for reciprocity as I heard Tri-X is like a french working girl and wanted to see what would happen.

So now that I did actually get a visible picture (with grain), I just wanted to figure what the effective EI for it would be, its higher than 6400 thats for sure.

I should point out: Both those below shots are from the same roll, they were developed in Rodinal 1+100, for 2 hours, with initial agitation for 30 sec, then 2 inversions at 40 minutes and 2 inversions at 1 hour 20 min.

Here is the shot its 2 minutes (EI 6400), uncorrected for reciprocity, 6x7cm - the corrected time would have been ~25 minutes, if I use the 30 min mark for reciprocity (32 min EI 400 = 2 min EI 6400) etc it's 3 and 2/3 stops correction, if I simply look at the 2 minute time without conversion, it's 3 and 2/10th's of a stop correction, which gives 18 minutes. The difference between the two times is approximately half a stop.

Here is a crop:

Here is another shot @ EI 6400, street, corrected for reciprocity, there is detail in both the dark and white areas on the neg, this is just how I set the contrast for my preferred taste:

Here is a crop of an out of focus area (focus is in foreground) to show grain - this image has also had sharpening applied unlike the previous image.

 

[Marco B]

OK, whatever you want to do to get the look YOU want is of course fine...

But, as I understand it now, what you are really looking for is a reliable reciprocity table or graph for TriX. I found Dead Link Removed, and the "convential film" column therein, of use. Scroll down on the page linked to find it. But I am sure someone can point something out for you maybe also specific for TriX, although in the end, it also depends on your personal workflow... But a "guesstimate" should be in reach...

One last question:

What were the real exposures for these shots? You say 2 minutes exposure time, but what F-stop did you use?

Marco

 

[Athiril]

It was f/11. But it doesn't give any real info or change anything, unless I decide to open up and use a shorter time which would change the amount of light hitting the film per millisecond and thus change the effiency at which the light is absorbed (reciprocity).

 

[Marco B]

So now that I did actually get a visible picture (with grain), I just wanted to figure what the effective EI for it would be, its higher than 6400 thats for sure.

....

Here is the shot its 2 minutes (EI 6400), uncorrected for reciprocity, 6x7cm - the corrected time would have been ~25 minutes, if I use the 30 min mark for reciprocity (32 min EI 400 = 2 min EI 6400) etc it's 3 and 2/3 stops correction, if I simply look at the 2 minute time without conversion, it's 3 and 2/10th's of a stop correction, which gives 18 minutes. The difference between the two times is approximately half a stop.

Athiril,

I am afraid I have to disappoint you a bit. The one huge error in thinking you make is to start doing calculation on IE's using reciprocity values for timing. E.g. 32 min EI 400 = 2 min EI 6400 IS NOT TRUE for reciprocity corrected timings.

That is why it is called "reciprocity failure", you can no longer say: 2s F8 equals 4s F11. Nor can you say "32 min EI 400 = 2 min EI 6400" if that 32 minutes was a reciprocity corrected time.

What you have to do is calculate back first to measured or indicated times.

If I understand you right, your reciprocity table is telling you you should expose Trix for 25 minutes if the measured or indicated time is 2 minutes.

Now instead of exposing your film for 25 minutes, you simply exposed the film for the duration of the metered time of 2 minutes.

What you now should do, is regard the 2 minutes as the reciprocity corrected time, as that is how you actually treated it and how much time you exposed, and look up the equivalent measured time of a 2 minutes reciprocity corrected time.

If I look at Ralph Lambrechts table (which differs from yours, as my corrected time for 2 minutes measured is 12 minutes instead of 25), I get a measured time of 30 seconds.

So we have 2 minutes versus 30 seconds, the difference between the measured times is two stops... Taking into account a likely difference between our reciprocity tables, at most 3 stops or so.

So now we come to IE. I think you have treated your film as the equivalent of a 1600 or at most 3200 ISO film max.

This also seems in line with some of my own results on TriX, as my previous posts here in this thread. The situations you photographed (especially the beach), are darker, but not that much.

Marco

(hope I didn't screw this up, my head gets dizzy too sometime when thinking about reciprocity )

 

[Athiril]

You misunderstand my intention, I didn't explain clearly, I have no trouble with exposure and reciprocity on a normal basis, I'm not saying its the same, I was converting times for the lookup table as reciprocity charts are for box speed of film, not my speed.

EI 6400 2 min, Im converting to the exposure time of EI 400, which is 32 min uncorrected, then I look up the corrected time for ~30 min, which was 6 or 7 hours etc, took the difference between that and the uncorrected time, which was about 3 and 2/3 stops, and add about 3 and 2/3 stops to the 2 min time to use as the corrected time, which gave 25 minutes (factor was 12.5 iirc, so log(12.5)/log(2) = stops).


So even if I do not convert back down to EI 400 for the lookup table, and simply look up 2 minutes instead of 30, the corrected time was 18 minutes, which was a factor of 9, very close to a factor of 12.5 (the factor for 30 minutes, or the time I'd use for EI 400 if I were shooting EI 400 f/11 at the beach), so I use the 12.5 factor on my EI 6400 time, not 9 if I was going to correct for it (which I chose not to), either way it's just a touch under half a stop difference in time. (log((log(12.5)/log(2))/(log(9)/log(2))^2))/log(2)

It'd be the same as metering @ 400 and getting 32 minute exposure time uncorrected, looking that up, and getting the 6-7 hour time, then saying "well on top of that im going to push 4 stops so I reduce that by 4 stops", which gives ~25 minutes.


I metered correctly, the exact correct exposure for my beach was f/1.4 (which I know the t-stop is f/1.8 for the lens I used) 4 seconds 3200, I took my digital along as my preferred night light meter and made test exposures, every Sekonic both spot, and incident are completely useless well before night this dark, the spot meter in my digital at least works, lets me measure contrast range and I can verify exposure data with easily.

Now f/11 2 min 6400 = the same exposure as f/1.4 4 seconds 3200, without accounting for reciprocity. If its 2 stops difference you say by working backwards, then the effective EI becomes 25600.


My street photo metering and correct exposure is f/1.4 1 second 3200, I can assure you I did not treat the film at 1600 or 3200, it was 6400.


Looking it up backwards as the corrected time I can agree with, but 2 minutes is the correct time for 6400 (without reciprocity correction), which means there is 16 times (4 stops) less light hitting the film per x period of time as there would be for a correct time of 2 minutes for EI 400 or box speed, and the charts are for box speed (and reciprocity is based upon x number of photons or x amount of light hitting the film per wavelength or point in time etc).

So logically, I would say I would take 2 min + 4 stops as the corrected time since the chart is for box speed, then look at the stops difference, which is what I was trying to get at earlier, which gives 3 stops, not 2.

IE: If reciprocity starts at 1 second for Tri-X, it means when your correct exposure is 1 second at EI 400, the amount of light hitting the film falls to a certain level where the efficiency of it being absorbed to form the latent image is reduced, if its 1 stop its reduced to 50% of the effiency of normal shooting conditions.

If it starts at 1 second, then it starts at 1/16th of a second for 6400 etc, as there is the same amount of light hitting the film per x period of time even if the exposure duration is different, so the effiency between the 2 are the same (50%, 1 stop).

Both shots were treated as 6400, one was corrected, the other was not as I were curious.


Depending upon the moon, the beach and general areas can be basically pitch black, or they could be lit up well enough to shoot video with one of the new dSLRs that shoot video. We dont have ambient light spill here, no light pollution, I was able to shoot a nebula on fixed tripod from my backyard with a 135/3.5 lens on my digital at 1600 (8 sec), single exposure you can see it.

 

[MattKing]

Pet peeve (redux).

Reciprocity failure doesn't arise because of abnormally long exposures (or for that matter because of abnormally short exposures).

It arises because the intensity of the light to which the film is exposed is too low (or too high), and that intensity cannot be compensated for by a reciprocal change in the length of the exposure.

If possible, the best way to respond to the problem is by increasing the intensity of the light hitting the film so that reciprocity is restored. Unfortunately, this is often impossible. When that is the case, we are forced to extend the length of the exposure, but due to the reciprocity failure, the amount of that extension is not linear.

The OP talks about combining increase in the length of the exposure, and a development "push", as a method of dealing with the low intensity of the light. With respect, these two techniques do different things.

The increase in the length of the exposure is intended to increase the shadow density. The development "push" will have little effect on the shadow density, but will instead increase the highlight density and apparent contrast. As the increase in the length of the exposure may very well have the effect of bringing the highlight density up, independent of the "push", the result of combining the two techniques is likely to be unpredictable at best, and deleterious at worst.

 

[Athiril]

"and reciprocity is based upon x number of photons or x amount of light hitting the film per wavelength or point in time etc" I've stated it enough times. I even infer in my opening post that it's not exposure duration based by the way I account for reciprocity.

I'm shooting the Tri-X @ 6400 as you can in non-reciprocity situations, I'll put it this way: in the street photo, the amount of light the film has absorbed to form an image is the equivalent amount of light for a good EI 6400 exposure in non-reciprocity conditions. In the beach photo, it is not, it has absorbed much less than the amount of light for a good EI 6400 exposure in non-reciprocity conditions, because I chose not to correct for reciprocity on this particular image.

 

[Marco B]

"and reciprocity is based upon x number of photons or x amount of light hitting the film per wavelength or point in time etc" I've stated it enough times. I even infer in my opening post that it's not exposure duration based by the way I account for reciprocity.

Athiril. I think you do understand the process of reciprocity failure well, especially considering your correct remark about the efficiency of photons creating the latent image as reposted here below, but your way of describing things is to say the least a bit difficult to understand... In addition, your original post missed some vital info.

IE: the amount of light hitting the film falls to a certain level where the efficiency of it being absorbed to form the latent image is reduced, if its 1 stop its reduced to 50% of the effiency of normal shooting conditions.

In addition, since I am now also convinced you do understand those basics very well, I am really wondering why you initially posted the thread, as I think you don't really need our approval for your calculations.

Anyway, lets keep it at this.

Marco

 

[Athiril]

It was a curiosity, as I wanted to know if I dont account for reciprocity, what my actual underexpose or EI I shot the film at was, but I can agree with using the chart in reverse as it makes more sense

And I want to the know the EI, as I'd like to find the point at which Tri-X just stops making a picture period.

The fact I can also overexpose the hell out of it by several stops and just pull it, or expose it as fast as that and still get a picture just makes it so versatile and wonderous, as long as I keep the EI consistent on the one roll you'll get pictures if you match development.

I could just carry the one film... if I didn't like the benefits of others.