Old Film Evaluation

[Viscommbozak]

I have many rolls of outdated 6x6 cm color negative (print) film. To try and establish a functional ASA, I shot a series of test exposures with my Bronica S2 of a three-stripe card (black/18% (middle) grey/white) - first at metered value of f/8 using the factory ASA and then three stops above and below. I had the film processed (C-22) and commercially scanned to jpg files. In Photoshop, I desatuated the image to eliminate color balance issues so I could measure only the luminance. I assumed tthe "correct" exposure would be the one that yielded a measured 18%/middle grey value of approximately 119/119/119. If my approach makes sense, then I have to conclude that the film has not significantly lost sensitivity and the metered exposure is pretty close - but the white value - instead of being 255 - is depressed on al test exposures. Apparently, there is a gamma issue and the old film has lost contrast capabilities.

First, I need a professional opinion as to whether my evaluation process is valid. Second if it is possible, once I settle on a specific scanned image, I need a step-by-step Photoshop procedure to correct the contrast range so the readings for the stripes on the test card come out to around 4 for black, middle grey at around 119 and whites to around 250.

Please do not hesitate to be critical if my logic is faulty but don't beat me up too bad since I am really old 🙂

[D Fosse]

The problem here is the scanning. That introduces a new and totally unpredictable variable into the equation. Establishing a calibration procedure for the scanning again requires a reference negative. The depressed highlights may just as well come from the scanning (and most likely do).

So this rapidly becomes an equation with a lot of unknowns, and I'm not sure it's worth the effort. I think I'd just make some test exposures at nominal ISO (ASA), and see how it behaves.

In any case, with old color film, I'd be less concerned with overall sensitivity, and more concerned with curve inconsistencies per dye - in other words, color casts in shadows, midtones or highlights. I sometimes used old and poorly stored film myself back in the day, and they often turned out to have very weird color casts. I've later tried to fix some of them up in Photoshop with varying success.

EDIT: To be a little more precise: yes, there is a gamma issue, but not because the film is old. Negative film didn't have the same tone response curves as digital files, and they can't be directly translated.

[Viscommbozak]

Based on your observation, I pulled out my Canon 9000F flatbed scanner and ran a series of 4400 ppi tests using 1) Totally unaltered settings 2) Adding what the CanoScan sofware describes a PHOTO correction and 3) Adding to PHOTO a low level of FADE correction. I desaturated each and measured the luminance value in Photoshop. Attached are the results of each compared to the initial "commercial" scans. Nothing really definitive and, in fact, sometimes values that one would expect - such as a decrease of black values with a decrease in exposure - actually go up. The experiment has been fun but I  am running out of patience given all the variables. Beyond the results of these tests, no one has commented on the validity of my test process, ie: assuming a technically correct scan, should the value of an 18% grey card in a properly exposed color negative yield Photoshop RGB values of 117/117/117 or an sRGB values of 125/125/125? Do you think I should go back amd start all over with fresh film?

[D Fosse]

The point is that the whole tone response curve is different. Yes, you can probably get an 18% card to reproduce as a given value in Photoshop, using any given color profile - but that tells you precisely nothing about the rest of the tone range.

Sorry, but this isn't going to get you anywhere.

[Viscommbozak]

Thank you so much. This was supposed to a be a simple, objective academic exercise. Apparently I am not as logically brilliant as I thought. There is an article on the web by a Daniel J. Schneider who recommends just adding a stop of exposure for each decade of expiration. In my case, since the film exired in 2000, ASA 400 becomes ASA 100. OK, I am good with that. However, where I can then get a"proper" scan (regardless of all the hex values ... yadayadayada) sounds like a whole another adventure. Do you have any recommendations for a top-notch scanning service?

[D Fosse]

You can get a scanning service as top notch as you like, the argument is still the same. A scan is not a valid reference.

Even if you could get a scanner rigourously calibrated, the question is still - calibrated to what, exactly? What would be the reference curve and where would you get it?

You're trying to solve an equation with two unknowns, by introducing a third unknown. You can see where that's going: deeper into the rabbit hole.

[Viscommbozak]

Thanks. Obviously I do not know enough to discuss curves and, since I solicited your opinions nor do I want to appear argumentative. But I do have a bit of historic reference on which to base my "project." Years before digital cameras and photoshop, I produced photographs for a high-end, glossy annual corporate report. We shot both color transparency and color negative film. Especially for the staged pictures, we always took a referencere photo which included a Kodak 18% gray card. The selected images along with the reference shot were sent to be scanned, digitally retouched and eventually ouput as a master transparency. That was over 30 years ago but I have to believe that the gray card reference had some value. Additionally, I have a Gretag Macbeth 10-step reflective grayscale calibration target and my Canon scanner has a custom profile that returns a digital image where the Photoshop RGB values "closely" match the RGB values of the target. I do not have a tool to develop the same profile for film media - especially color negatives - but I have to believe that professionals who still love and shoot negative color film would insist on scanning services that have similarly calibrated equipment. There has to be a way to ensure that when X goes in you get X back.

[Per Berntsen]

From my point of view, ISO speed is more about capturing shadow detail than getting a perfect 18% gray.

I suggest that you take regular photos of scenes with a high dynamic range, with deep shadows, midtones and highlights at various ISO speeds. Then have them scanned with the aim of getting a low contrast image that contains all the information in the negative. Then evaluate the level of shadow detail in the scans; the ones that provide sufficient shadow detail should give you a rough indication of the film speed.

This might to some extent eliminate, or at least mitigate some of the variables introduced by the scanning process.

And a lot will depend on the quality of the scanner and the scanning software, as well as the skills of the scanner operator.

If your scanning service uses auto adjustments in the scanning software, it is quite likely that some shadow and highlight detail will be lost in the scanning process.

Color negative film has a huge dynamic range, and blown highlights are generally not a problem in my experience.

The midtones will take care of themselves, and can easily be manipulated in post processing. 

Like @Chuck Uebele I am also curious as to why you have the film processed in C-22.

This is an obsolete process that was replaced by C-41 in the early 1970's.

[Viscommbozak]

First of all, you are most likely correct. Process was probably C-41, not C-22. I did not specify the process to the lab and I have to assune they used the correct one. The C-22 was just stuck in my 79 year-old memory 😞

I am going to take your syggestion. I will buy some, new decent color negaive film in 35mm. (Suggestions?) I will shoot it in my Nikon F, have it processed "normal" and send it out to a couple or three different scanning services and see where that rabbit leads me.

Not surprisingly, your comments of  "And a lot will depend on the quality of the scanner and the scanning software, as well as the skills of the scanner operator. If your scanning service uses auto adjustments in the scanning software, it is quite likely that some shadow and highlight detail will be lost in the scanning process" are kind of the reason for my request for a recommendation. Do you or any one on the Forum know of a scanning service or services that would provide Hi-res scans that meet your suggested criteria and provide a nice, "neutral" image?

Thanks,

Robert

[Chuck Uebele]

Your use of a gray card make sense. I remember doing that also. I can relate about mixing up C-22 and C-41. I used to process both. I used to work in a pro lab, decades ago, and used both reflective and transmission densitometer. You might be able to get a relatively cheap used one, if you still want to really measure the film's values. Haven't kept up with current films, as after at least 30 years of working in dark rooms, I've fully embraced digital. 

[Chuck Uebele]

As D Fosse mentioed, scanning introduces a whole new variable. Frankly, with color negs, you can pretty much just eyeball the negative to see if it was properly exposed. If you really wanted to do it to the numbers, you would need a densitometer to read the negative. 

That film must be old, if you had it processed in C-22!