Degradation of Colour Film over Time

[DREW WILEY]

Wilhelm deserves a lot of credit for putting the ball into motion, so to speak. The movie archiving trade was especially interested in his work. Some of his conclusions were simply based on examining older material and seeing how it had fared in time under various storage, display, and projection circumstances. Newer media were subjected to accelerated aging tests, with theoretical extrapolation of the results, a methodology with inherent flaws because it can't take into account all the potential variables, but is nonetheless a starting point. Wilhelm was mistaken in several instances. In the interim, others have stepped in and improved Wilhelm's methodology, and taken into account new products as well.

Kodak had their own testing methods relative to their own marketing concerns. There is plenty of presumption attached to Arrhenius tests, although that too might be a useful starting point. The whole problem is exactly what I just hinted at - all of these testing methods make certain assumptions which are essentially predictive, which simply cannot account for every real-world condition. Today, we see a lot of claims of so and so hundreds of years of lightfastness to certain inkjet prints, for example. Well, how many years have they even been around to observe them? So far, Aardenburg seems to have the best job improving the testing and prediction of that category of media.

None of this should be taken as absolute. When I was highly involved with industrial coatings, pigments, and architectural color consultation, I had to take in all kinds of climate, lighting, and environmental factors to predict how certain colors would fade in relation to one another, and what the best re-paint cycle might be. And most of those pigments were more permanent than anything found in an inkjet print. They had to be, due to direct sunlight exposure years on end. All the actual manufacturer could do, besides their own accelerated aging torture chambers, is representative testing in two or three specific climates, like humid Florida versus cold winter Minnesota. I did my own supplementary tests based on the dramatically different microclimates found here on the West coast among my own clients (foggy coast versus hot inland versus the freeze-thaw cycles characteristic of the Sierra mountains). It's just plain complicated.

I have also farmed out my own color prints to widely varying display environments, some of them deliberately abusive. One test involved mounting two matching Cibachrome prints using the same adhesive which pros who mount mirror-surface Formica use. The result is remarkably smooth (but hazardous to apply due to the nasty solvents). One of these was framed and put in a residence using ordinary household tungsten bulb illumination; the last time I checked it, it was still doing fine after 20 years. The other one was set in a window and received direct sunlight - it totally faded in a week! That told me how that kind of harsh solvent exposure pre-conditioned the degree of UV sensitivity. Then there were tests for oxygen exposure versus hermetic sealing - on and on. If I shipped a print, I took into account the climate it was going to when I framed it - how many galleries do that? There are just soooo many variables!

But we're obviously off on a tangent different from the initial thread. And Fuji chromogenic papers in particular have significantly improved since Wilhelm's day. As far as color films go, I haven't noticed any fading in any of my old chromes which needed reprinting, only in a few old Vericolor negs.

 

[ChrisGalway]

OK, so we have discussed the lifetime of unexposed (and unprocessed) colour film (my original question), and of (exposed and) processed film.

So what is the "lifetime" of exposed, but unprocessed, colour film?

Like many here on this forum, I process my films (mainly E6, not colour neg) in batches using a commercial kit. Generally I process most films within 2-3 weeks of exposing the roll of film, but occasionally I have to keep the films for up to 8-10 weeks before I have a viable batch of films to process in the kit. I just keep the exposed but unprocessed films in a drawer until I'm ready to process them (not in the fridge).

Any comments? I've never noticed any problems but honestly I'm probably not the most critical person, and of course I've never processed test strips and done proper densitometry (I'd love to do this, but I have no test strips and no densitometer!). By the way, I nearly always expose films within their use-by date, and the unexposed and unprocessed films are stored in the freezer.

 

[MattKing]

If you were doing narrow tolerance, specialized work for publication - e.g. catalog photography back in the day - then it would be inadvisable to mix "promptly after exposure" processed film with "wait 10 weeks after exposure" processed film, because you might very well see subtle differences that would not be good for a high volume, low tolerance for variation workflow.
That issue would be of increased importance if the drawer you keep the film in is subject to a range of temperatures and humidity levels.
It is best to keep the film in a cool and dry environment, where cool means room temperature or lower. One of the big reasons to a consider a fridge is that, in addition to providing a low temperature environment, it also provides a temperature controlled environment.

 

[ChrisGalway]

If you were doing narrow tolerance, specialized work for publication - e.g. catalog photography back in the day - then it would be inadvisable to mix "promptly after exposure" processed film with "wait 10 weeks after exposure" processed film, because you might very well see subtle differences that would not be good for a high volume, low tolerance for variation workflow.
That issue would be of increased importance if the drawer you keep the film in is subject to a range of temperatures and humidity levels.
It is best to keep the film in a cool and dry environment, where cool means room temperature or lower. One of the big reasons to a consider a fridge is that, in addition to providing a low temperature environment, it also provides a temperature controlled environment.

So we should keep exposed but unprocessed film in the fridge? Makes sense, I guess, but I did not realise that.

 

[koraks]

So what is the "lifetime" of exposed, but unprocessed, colour film?

I don't know, but I'd expect it to be on the scale of months, probably.

I vaguely recall having read test results on latent image retention, with the general curve of time (horizontal) vs. latent image retention (vertical) being something like this:

Note that the horizontal axis would be logarithmic, not linear, meaning that much of the latent image loss would happen in the first few seconds. Then there would be a relatively long, stable plateau, and eventually further loss/degradation would occur.
How exactly the curve shape is and what the scale of the horizontal axis would be - I don't know. I suppose it would vary from one emulsion to the next (and even between the different emulsions of a color film pack - i.e. there would be a crossover effect), and I bet it's influences by environmental conditions at well.

Mind you, the above is synthesizing based on vague recollection of having read stuff. Don't take it as gospel.

 

[ChrisGalway]

I don't know, but I'd expect it to be on the scale of months, probably.

I vaguely recall having read test results on latent image retention, with the general curve of time (horizontal) vs. latent image retention (vertical) being something like this:
View attachment 376542
Note that the horizontal axis would be logarithmic, not linear, meaning that much of the latent image loss would happen in the first few seconds. Then there would be a relatively long, stable plateau, and eventually further loss/degradation would occur.
How exactly the curve shape is and what the scale of the horizontal axis would be - I don't know. I suppose it would vary from one emulsion to the next (and even between the different emulsions of a color film pack - i.e. there would be a crossover effect), and I bet it's influences by environmental conditions at well.

Mind you, the above is synthesizing based on vague recollection of having read stuff. Don't take it as gospel.

Can anyone put a SCALE on both axes of this graph?

 

[koraks]

Not me in any case.
Edit: rooting around quite extensively on Google turns up....a Photrio post. https://www.photrio.com/forum/threads/latent-image-stability.76909/
If I chart the values in that post against the time in seconds, this is what I get (horizontal = log scale, seconds):

The timescale 'only' extended out to 1 month (31 days). Maybe the plateau in the long run is stable and doesn't taper off down as I supposed; IDK. There must be some data on this, somewhere.

 

[MattKing]

So we should keep exposed but unprocessed film in the fridge? Makes sense, I guess, but I did not realise that.

The advice differs slightly between films with backing paper, and films without.
I don't put opened (removed from foil wrappers) rolls of 120 film back into the refrigerator due to issues with humidity/condensation and how that interacts with backing paper.
I do sometimes put exposed colour 35mm film back in there, in the plastic storage containers.
I live in a very temperate climate, so room temperature storage is usually fine.
The Kodak Storage and Care Information Summary: https://kodakprofessional.com/sites/default/files/wysiwyg/pro/CIS_E30.pdf

 

[Mr Bill]

So what is the "lifetime" of exposed, but unprocessed, colour film?

We actually DID do latent image shift tests on all of the films we looked at. But... they would not actually answer the questions you have. Our purpose was to determine a minimum holding time before developing the film. We wanted to make sure that the latent image shifting had become relatively stable before doing critical image evaluation testing. So we did pretty short term tests - only a week or two.

As koraks indicates the greatest changes happen immediately after exposure, and then slow down. So if you process immediately, say within a half hour after exposure, the results may be different than after a one-day delay.

Now, our business in those days was to shoot studio portraits for a half-week to a week, then ship that film to our lab(s) for processing and printing. So a typical exposure might be a week or so old when it was processed. When we did critical evaluations of a new film we wanted to make sure that the testing was representative of the actual studio exposures. So the first question is, what are the aging characteristics of the latent image?

Very roughly, going from a distant memory, we found measurable differences between a 2-hour holding vs 4-hour vs 8-hour. After 8 hours there was essentially no change up to the limit of what we tested, which was maybe 2 weeks. (Longer than that, we didn't care ... it didn't matter to our business.) From fuzzy memory I'm thinking that from a 4-hour hold up to overnight we might see mid-tone densities on a sensitometric wedges change by possibly 0.02 density units. So essentially next to nothing but possibly pertinent in critical color evaluation of optically-printed portraits. (To put this 0.02 density change into perspective I doubt that 90% of pro photographers would have been able to even detect it, much less ascribe it to latent image; it would likely be eaten up in "process noise.")

Let me clarify the conditions - these were for the slower-speed Kodak professional portrait/wedding films, and the holding conditions were about 72 F and 50% RH. I wouldn't try to extrapolate to other situations.

FWIW you probably don't see many people doing latent image shift tests on the internet cuz it's not straighforward how to do it while keeping good "controls." Stephen's test, linked by koraks in post #32 is about the only one I've ever seen. I'm sorta presuming that Stephen has periodically exposed a sensi wedges on his sensitometer (he says he processed them all at the same time). Which brings up the question, how do you know that the sensitometer remained stable the entire time? (Unless you had an instrument to verify it, or perhaps exposed dual samples - one for the test and one as a control.) Then he reports results as "EFS" which I'm guessing means "effective film speed" which is a bit indirect. (In our case we looked at full sensi curves to see exactly where they differed.) Those things aside, process all at the same time is generally a good "control" for consistency of processing.

Short answer is that I don't really know for sure, and am skeptical about what most other people think they know (unless they have specifically tested). The common wisdom is that lower temperatures help preserve the latent image, which I'm pretty sure is true. And I'm pretty sure that the top-tier manufacturers know how to make film with good latent image stability.

 

[halfaman]

So we should keep exposed but unprocessed film in the fridge? Makes sense, I guess, but I did not realise that.

I developed a Polaroid HD 100 roll (made by Agfa) for a friend that was exposed and then kept in a frigde for 15 years. The result was totally normal to my eyes, and no issues at all scanning the roll.

Develop CN film exposed in 2006

A friend of mine has given me a roll 8 rolls of different CN films (Superia, Gold, Polaroid) he shot at least 15 years ago to develop it because I have a big mouth... :angel: At least he kept the rolls in the freezer all this time. I don't think C41 is going to work, I developed once a roll...

www.photrio.com

We actually DID do latent image shift tests on all of the films we looked at. But... they would not actually answer the questions you have. Our purpose was to determine a minimum holding time before developing the film. We wanted to make sure that the latent image shifting had become relatively stable before doing critical image evaluation testing. So we did pretty short term tests - only a week or two.

As koraks indicates the greatest changes happen immediately after exposure, and then slow down. So if you process immediately, say within a half hour after exposure, the results may be different than after a one-day delay.

Now, our business in those days was to shoot studio portraits for a half-week to a week, then ship that film to our lab(s) for processing and printing. So a typical exposure might be a week or so old when it was processed. When we did critical evaluations of a new film we wanted to make sure that the testing was representative of the actual studio exposures. So the first question is, what are the aging characteristics of the latent image?

Very roughly, going from a distant memory, we found measurable differences between a 2-hour holding vs 4-hour vs 8-hour. After 8 hours there was essentially no change up to the limit of what we tested, which was maybe 2 weeks. (Longer than that, we didn't care ... it didn't matter to our business.) From fuzzy memory I'm thinking that from a 4-hour hold up to overnight we might see mid-tone densities on a sensitometric wedges change by possibly 0.02 density units. So essentially next to nothing but possibly pertinent in critical color evaluation of optically-printed portraits. (To put this 0.02 density change into perspective I doubt that 90% of pro photographers would have been able to even detect it, much less ascribe it to latent image; it would likely be eaten up in "process noise.")

Let me clarify the conditions - these were for the slower-speed Kodak professional portrait/wedding films, and the holding conditions were about 72 F and 50% RH. I wouldn't try to extrapolate to other situations.

FWIW you probably don't see many people doing latent image shift tests on the internet cuz it's not straighforward how to do it while keeping good "controls." Stephen's test, linked by koraks in post #32 is about the only one I've ever seen. I'm sorta presuming that Stephen has periodically exposed a sensi wedges on his sensitometer (he says he processed them all at the same time). Which brings up the question, how do you know that the sensitometer remained stable the entire time? (Unless you had an instrument to verify it, or perhaps exposed dual samples - one for the test and one as a control.) Then he reports results as "EFS" which I'm guessing means "effective film speed" which is a bit indirect. (In our case we looked at full sensi curves to see exactly where they differed.) Those things aside, process all at the same time is generally a good "control" for consistency of processing.

Short answer is that I don't really know for sure, and am skeptical about what most other people think they know (unless they have specifically tested). The common wisdom is that lower temperatures help preserve the latent image, which I'm pretty sure is true. And I'm pretty sure that the top-tier manufacturers know how to make film with good latent image stability.

Very curious compared to cine stock. In the "old days" where "dailies" were done, footage was developed within hours.

 

[koraks]

I developed a Polaroid HD 100 roll (made by Agfa) for a friend that was exposed and then kept in a frigde for 15 years.

I've developed Kodak Gold that was expired 10-15 years previous to me processing it; this was not stored in the fridge, but was left inside the camera the roll was exposed in. The colors were heavily shifted and crossed over, but the images were recognizable and could be digitally restored.

 

[Prest_400]

There was a fellow forumer that shared his findings as he left exposed 120 film for a good part of a decade and recently developed it.
Ektar was the most affected compared to the Fujicolors and Portra. Just now am not able to find that particular post.

 

[DREW WILEY]

This whole latent image topic is so specific that it's probably best to defer to the manufacturer's own studies and recommendations. Kodak specifically states that the latent image characteristics of their most recent Porta films have been deliberately improved. But even that is subject to many variables of climate, storage, etc.

 

[Agulliver]

It varies a lot between different films and with storage conditions. About 6 years ago I came across an old camera with a Kodacolor II 120 film in it....had it developed by my local lab and they got a good set of images, though only a few had sufficient usable colour to obtain colour photos. From the subjects it strongly looked like the mid 1970s so 40 years between shooting and processing.

More common I've found old part shot rolls in cameras or at at my parents' house that I've processed maybe 10 years after and they're usually fine....sometimes a bit of colour shift especially with Fuji Superia but always possibly to correct the colour shifts.

I'd not be worrying about 10 weeks unless the films were in 90+% humidity and very hot. As others say, there might be a very slight measurable difference but you're unlikely to be able to see it. Film is remarkably stable in most cases. Especially negative film.

 

[Romanko]

Here are some empirical results:
- all summer in a car rendered Kodak Gold mostly unusable

- after 10 days you start noticing the difference (provided it was not the scanner doing strange things)

 

[koraks]

provided it was not the scanner doing strange things

No guarantees whatsoever. She even mentioned it was acting up.
I have no trouble believing that her film was cooked. But the examples shown don't really tell the story.

 

[Romanko]

In the first video (by srad) the film is beyond rescue without any doubt. It looks 30 years expired (at least). And that lady does horrible things to her film. Please be warned that watching her other videos can have a disturbing effect on many young film photographers. Parental guidance is recommended.

The fact that scans from the "cooked" film in the second video look more natural than from fresh film raises a lot of questions as to what is actually being compared. In any case the results between the two rolls were different (for whatever reason) so his experiment was successful.

 

[Tel]

This is a really broad question, probably too broad to answer.. There are (have been) many different iterations of color emulsions over the years, and each different formulation likely shows a different response to aging. I’ve shot a lot of expired film, mostly for fun and hardly as any sort of scientific study, but i’ve noticed that some stocks tended to shift toward the greens, some blues and others magenta. Agfa XPS, by contrast, seems to have shifted little, though the negs do seem to be getting thinner. (This from some 46mm stock I’m currently shooting, expired in 2003.) So I’d venture to say that we can’t lump all color films into one category and reach general conclusions about them. The different stocks do seem to respond to aging differently. And then, as noted above, there are all the atmospheric variables: humidity, temperature, contact with other chemicals (in backing paper, for example), etc.