Dealing with under/over-exposed & -developed negatives; scanning or enlarging

[dcy]

{Moderator note: this thread was split off of another one here: https://www.photrio.com/forum/threads/harman-kentmere-200-officially-released-2025-05-08.213596
It's placed in the B&W Film, Paper & Chemistry forum despite the fact that scanning plays a role; the rationale is that it's not purely about scanning, but rather about the question how to deal with negatives that didn't come out optimally.}

In my experience, the only circumstances where there are clear differences are at the periphery:
- with clearly under-exposed or under-developed negatives, it may be easier to obtain an acceptable final result from scanning and post processing; and
- with clearly over-exposed or over-developed negatives, it may be easier to obtain an acceptable final result from darkroom printing.
Which is why the preference for push processing and under-exposure among those who exclusively develop and scan is so surprising to me.

Hey! I'm responding to a month-old comment, but can you explain to me why under-exposed/developed negatives may be easier to process digitally and over-exposed/developed negs may be easier to process in the darkroom?

I'm not coming at this with any expectation. Before reading your comment my guess would have guess that if the negative has lost information then it's lost, regardless of what you do with the negative afterward.

 

[MattKing]

Hey! I'm responding to a month-old comment, but can you explain to me why under-exposed/developed negatives may be easier to process digitally and over-exposed/developed negs may be easier to process in the darkroom?

I'm not coming at this with any expectation. Before reading your comment my guess would have guess that if the negative has lost information then it's lost, regardless of what you do with the negative afterward.

The comment you quote is based mostly on practical experience, not any consideration of theory. And it is probably best to understand that I have way more experience - ~ close to 50 years - in the darkroom than I do scanning and digital post processing. Although my experience with the latter is creeping up - I've now got around two decades of it.
But I expect that it has a lot to do with the end result - with darkroom printing your goal is a reflected light medium - darkroom printing paper - while with digital processing your goal is a file which will at least initially be optimized for viewing on a transmitted light medium - a computer or other monitor.
With respect to the dense negatives, there is usually a lot more information embedded in a too dense negative than there is in a too thin negative, and in the darkroom, you can often get at that densely embedded information by manipulating contrast and greatly extending exposure. A lot of that is due to the nature of photographic paper, which responds in significantly different ways than digital sensors do.

 

[dcy]

But I expect that it has a lot to do with the end result - with darkroom printing your goal is a reflected light medium, while digital processing your goal is a file whitransmitted light medium.

Oh. I wonder if that changes if the purpose of scanning is printing.

My transition from digital to analog started when I realized that I never looked at photos on my computer, but I frequently stopped to admire the few prints I had around the house. I decided that society lost something when it transitioned to digital. That realization led me to re-discover analog. Today my goal is always to print at least a few shots because what I print is what I'll remember.

 

[MattKing]

Oh. I wonder if that changes if the purpose of scanning is printing.

Not really.
Film and darkroom photographic paper are designed as a part of a system - the behaviors of one are complementary to the behaviors of the other.
Everybody who scans film still has to first deal with the behavior of a digital sensor, rather than the behavior of a photographic paper emulsion.
And scans are post processed using monitors, not printers, even if the intended end result is a print.

 

[koraks]

why under-exposed/developed negatives may be easier to process digitally and over-exposed/developed negs may be easier to process in the darkroom?

1: In the darkroom you're basically limited to the contrast of grade 5. A very flat and thin negative may result in a dark and/or flat print even at that grade. There's virtually no way to really fix that in the darkroom. However, in digital space, you just apply a dramatic contrast curve and Bob's your uncle. It may not yield the best image, but it will often be better than what you can wet print.

2: On the other hand, very dense negatives can sometimes produce problems with scanners, which tend to become more noisy as density is very high. In these cases, you may get a slightly better print in the darkroom, where you can just expose through the density.

In my experience, (1) is more 'true' than (2), because scanners for the most part actually can handle dense negatives quite well (unless they're really ridiculously dense, but you don't get there just by accident). Of course, in both cases, it's a marginal situation and the results will be sub-par.

 

[albireo]

my guess would have guess that if the negative has lost information then it's lost, regardless of what you do with the negative afterward.

That's a correct guess. In 25 years of experience in scanning my negatives, I've never been able to 'rescue' via scanning and/or post-processing a very thin negative. A poor negative always, without exception, yields a poor scan.

I'll go even further: after a bit of experience, it even becomes possible to identify from the scan, whether the negative was too thin or not. The negative is the final word, of course, but the scan often offers important hints. If you do this for years you can pick up, from a scan, whether photo-shopped or not, whether intensely 'curved' or not, if the negative wasn't ideal for scanning.

I've never, without fail, seen a perfect scan from a very thin (or very thick) negative, even if the person producing it was a Photoshop wizard. In fact, Photoshop wizardry doesn't exist - it's an excuse wet printers often use to justify their poor understanding of the hybrid process. You can't polish a turd. Of course, everything is now changing with AI-produced or doctored images.

But to go back to your question above, in my own experience, and my own workflow, the best scans are obtained for a negative exposed and developed for a target gamma in the interval .55-.6. This is basically a target gamma in the region of what was considered 'standard' in the days of wet printing with a condenser enlarger head.

Sounds honestly not so surprising to me, as the engineers who designed film scanners had to start somewhere and my unsupported assumption is that they would have started designing the device to perform well in a reasonable distribution of densities centred at gammas of negatives exposed and developed according the accepted standard at the time.

 

[koraks]

Sounds honestly not so surprising to me, as the engineers who designed film scanners had to start somewhere and my unsupported assumption is that they would have started designing the device to perform well in a reasonable distribution of densities centred at gammas of negatives exposed and developed according the accepted standard at the time.

Film scanners were made with color positive/slide film in mind. The density range of any negative (color or B&W) is a fraction of the density range of a color slide.

You can't polish a turd.

Actually, you can. It'll still be a turd, but a prettier turd than before. And when it comes to rescuing, that's what people may sometimes be looking for.

 

[albireo]

Actually, you can. It'll still be a turd, but a prettier turd than before. And when it comes to rescuing, that's what people may sometimes be looking for.

If I read you correctly, you're suggesting that a thin negative is akin to a 'technical impossibility' if the purpose is wet printing, but merely a 'creative choice' if the purpose is scanning.

If true, then the question becomes why does it become a 'creative choice'? Is it because people are actually 'looking for' those results, as you say, or because they don't know any better? Is it a chicken, or egg issue?

The exposure and development of negatives for wet printing benefits from often pretty rigorous discussions on this and other forums.

Not so the exposure and development of negatives for scanning: we often read stuff like 'are you wet printing or scanning? Ah you're scanning. Expose and develop like a$$ and just play with the curve anyway '.

I don't think this is fair, you see. This forum is a precious teaching and learning resource for many people out there, and I believe it should focus on giving strong foundations whatever the output sought.

 

[koraks]

If I read you correctly, you're suggesting that a thin negative is akin to a 'technical impossibility' if the purpose is wet printing, but merely a 'creative choice' if the purpose is scanning.

If you read correctly, you would have noticed that I did not say the above, and instead said this:

Of course, in both cases, it's a marginal situation and the results will be sub-par.

You ignore or reject (for reasons unclear to me) the possibility that someone may sometimes end up with a sub-par negative, unintentionally, by accident, and may then want to salvage it as best as they can. That's what this is about. Not about the question whether people should aim for thin negatives or whatever.

 

[albireo]

If you read correctly, you would have noticed that I did not say the above, and instead said this:


You ignore or reject (for reasons unclear to me) the possibility that someone may sometimes end up with a sub-par negative, unintentionally, by accident, and may then want to salvage it as best as they can. That's what this is about. Not about the question whether people should aim for thin negatives or whatever.

I don't think anyone mentioned "salvaging" anything, but I haven't gone very far back in the thread.

 

[Don_ih]

There are negatives so thin you will find it next to impossible to enlarge but can get a picture from by scanning. It should be noted that it works better if the negative is very thin from underdevelopment rather than underexposure. A very thin negative that has detail over the entire surface can sometimes scan and almost look normal but be pointless in an enlarger.

Very thin negatives due to underexposure - the scanner can't find what's not there.

 

[koraks]

I don't think anyone mentioned "salvaging" anything, but I haven't gone very far back in the thread.

The original thread this was in (I've now split it off) was really about lots of things (primarily about the introduction of Harman 200), and not so much about the question we're discussing here. So I've separated out this discussion from it, making things perhaps a little easier to oversee. I think from what is now #1 it's clear that we're discussing a situation of negatives that didn't come out as intended, as implied by the mention of under- and overexposure and -development as well as 'lost' information. Hence, the term 'salvaging' seems appropriate to me.

What you seem to worry about is an alleged tendency of people (I'm not sure who, or how many, or where they may be) to recommend just winging it when it comes to film exposure and development, because after all, anything can be fixed in Photoshop. Frankly, I'm not aware of this being a substantial problem, nor do I see many (or even any) people suggesting this here on Photrio (maybe on Reddit or YouTube?) So for all I know, the problem you're worrying about may not even exist in the first place. If it does, it may be helpful to be specific (links, quotes etc.) in highlighting it so all of us are on the same page.

What I do acknowledge is that there's a stream of thought that's fairly popular and often voiced, which contends that negatives produced specifically for scanning should ideally be of a lower gamma than those intended for optical enlargement. Personally, I disagree with that and would argue the exact opposite on the basis of signal-noise ratio of the sensor + ADC system, which benefits from being fed a signal (i.e. density/luminous flux range) that occupies as much of the dynamic range of the sensor/ADC combination as possible. After all, the noise floor is at least partly constant, and thus, the bigger the (unclipped) signal, the higher the signal-to-noise ratio. However, I've rarely (or even never) seen this argument being worked out on forums. I'm not exactly sure why, as to me, it makes perfect sense, but for some reason, there's a substantial group of hybrid photographers who seem to be leaning into the opposite direction.

Even so, the 'negatives for scanning should be a little lower in contrast' argument still does not constitute the same as 'ah, just do whatever and the scanner will sort it.' For all intents and purposes, methodologies that aim for a higher vs. a lower gamma can still be as rigorous or lax, depending on the habits and preferences person(s) involved. There's nothing inherent to the aim for a lower gamma that implies a lax attitude.

Coming back to the topic of salvaging thin negatives - I think there's practical merit to it and I know from experience (as many people do) that scanning and digital processing are very useful in such a context. No, scanning and curve adjustments will not create a stellar (from a purely technical viewpoint) image from a significantly underexposed and underdeveloped negative. But it can be the ticket to a presentable or (for the purpose) acceptable image that the photographer prefers over...well, no image at all, or just a very muddy darkroom print that's barely recognizable.

Case in point - my niece regularly feeds me a roll of poorly (variably) exposed Fomapan 400. She enjoys (for whatever reason) shooting film from time to time, but she does not have a solid background in photography, her capabilities in terms of exposure are quite basic and her expertise and interest in the technical sides of photography are superficial at best. So whenever I've processed a roll of her film, I find myself looking at maybe 6-12 frames that are properly exposed, with the rest being generally frighteningly thin in the shadows.

At at that point, I could go back to her, give her my sternest look, set her film on fire with a zippo and tell her she should damn wel do a better job because there's no polishing to these turds. Alternatively, I can gently explain that maybe she could set her camera to 200 instead of 400 for this film and see if she can find the time between changing diapers, getting the kid back from daycare and consulting with a neverending string of diabetes patients to have a look at one or two YouTube videos on exposure (but who am I kidding given her valid priorities). And I'll Wetransfer her images which I've diligently scanned, applied a totally dramatic, outrageous and insane contrast adjustment so that at least my niece will recognize her toddler without having to wonder if it's maybe the neighbor's kid, and guess what - she's happy. Because THAT's what reality often looks like. Whether you like it or not, that's what happens, and we're going to talk about that on a forum as well from time to time.

 

[albireo]

Even so, the 'negatives for scanning should be a little lower in contrast' argument still does not constitute the same as 'ah, just do whatever and the scanner will sort it.' For all intents and purposes, methodologies that aim for a higher vs. a lower gamma can still be as rigorous or lax, depending on the habits and preferences person(s) involved. There's nothing inherent to the aim for a lower gamma that implies a lax attitude.

I was never replying to any of the posts suggesting 'negatives for scanning should be a little lower in contrast'. I replied directly to dcy with my own views on the matter, which then you quoted.

You seem pretty defensive about this whole thing, but my post wasn't directly about you or your views on the topic at all.

 

[Yezishu]

I believe that scanning or enlarging film never captures 100% of flim's information. Different papers and scanners each have their own subtle biases, extracting varying details from the same negative.

As Koraks noted, there’s debate over how to best prepare negatives for scanning. In my observation, toady's digitization devices—from smartphone cameras and modified flatbeds to DSLRs, new mirrorless cameras, professional film scanners, and drum scanners—vary widely in dynamic range. Even ideally, we should adjust negatives to maximize the capabilities of the scanner, the operator may not fully understand their equipment’s performance or proper settings. Sometimes the issue is whether the details on the film will exceed the scanner’s default settings range, when the operator does not know how to make changes.

The intended output medium also matters. Many newcomers today display their work primarily on social media at mobile screens, which changes expectations for quality and detail. For instance, some darkroom push processing or curve adjustments might not yield a satisfactory printed photo from thin film, but could be adequate for viewing on a phone screen. The quality will definitely still be affected and not ”recommended“, but... whether this is a kind of "save" depends on what is considered "acceptable"......

 

[Alan Edward Klein]

Kodak claims Ektar and Tmax films are better for scanning. Anyone know why?

 

[Alan Edward Klein]

One thing about scanning I discovered. Setting black and white points (levels) allow great changes in the toe and shoulder aspects of the film. I don't chemically develop or print. So I don't know if this same ability is available using chemicals to print? Comments?

 

[dcy]

There are negatives so thin you will find it next to impossible to enlarge but can get a picture from by scanning. It should be noted that it works better if the negative is very thin from underdevelopment rather than underexposure. A very thin negative that has detail over the entire surface can sometimes scan and almost look normal but be pointless in an enlarger.

Very thin negatives due to underexposure - the scanner can't find what's not there.

That makes a lot of sense ---- My limited understanding of the situation is this:

1) You have a density curve. A properly exposed but under-developed image will have a curve with a too-shallow slope, but the curve will start at the right place so there's information everywhere. I can scan that, open it in GIMP and crank the contrast slider up to 11. --- Not ideal; I've still lost tonal gradation, but the photo is not a complete loss.

2) An under-exposed but properly-developed image will have a very long toe, followed by a curve with the correct slope. A long horizontal line / toe has zero information. You're screwed.

 

[Bill Burk]

This negative was thin and scanned nicely. It was a devil to print in the darkroom because there was no detail in my friend's sweatshirt. Scanner had no trouble.

 

[albireo]

This negative was thin and scanned nicely.

It scanned terribly, as expected. No offense intended at all, but that is a poor negative scan.

Scanner had a lot of trouble, as you can see from the thermal noise visible in the black shirt of the man, which was basically a transparent bit of negative with no info to scan.

 

[albireo]

Very thin negatives due to underexposure - the scanner can't find what's not there.

Yep - and given many people tend to overdevelop rather than underdevelop, 'thin as in underexposed', rather than 'thin as underdeveloped' tend to be, I'd expect, the vast majority of 'very thin' negatives. Very little to rescue, and a sub-optimal scan at best.

 

[qqphot]

It's not always apparent what sort of curve will be applied to a digitized negative when being inverted by the typical software people tend to use, such as Negative Lab Pro. I wonder i this also contributes to the perception that thin flat (black and white) negatives are "easier" than denser ones - they do invert in NLP to decent looking positive images with almost no manipulation, while negatives with shadow detail and a fuller density range tend to produce (by whatever algorithm NLP uses) a first attempt that looks drastically overexposed and with no detail in highlights, until you crank the sliders to their extremes. If you invert by hand by drawing a suitable curve this isn't so much the case. Maybe this is all my own ignorance though, as I'm sure will be pointed out to me shortly.

 

[DREW WILEY]

Anything that can be retrieved on a negative by means of a scan can also be optically printed. It just depends on how much trouble you are willing to go to. I've been tasked with printing some extremely difficult negatives over the years; and if someone was willing to pay that kind of premium, you'd be amazed at what can be done with an all-optical workflow and advanced masking techniques. Nowadays, drastic "salvage" image restoration is much more easily done via scanning instead. Yet extra hoops would still have to be jumped through if the client wants an actual darkroom print as the endpoint, rather than an inkjet print.

But in terms of my personal work, it would have to be some kind of very exceptional image to bother with all the extra fuss. Only a few of my earliest LF negs of special hard to reach places fall into that category, before I learned the correct level of development.

 

[Bill Burk]

It scanned terribly, as expected. No offense intended at all, but that is a poor negative scan.

Scanner had a lot of trouble, as you can see from the thermal noise visible in the black shirt of the man, which was basically a transparent bit of negative with no info to scan.

No offense taken. This was an example where the negative is too thin to print, but the scan implied there was a brighter image to be had. It stunk at grade 2 and I moved on after three terrible prints.

 

[dcy]

It's not always apparent what sort of curve will be applied to a digitized negative when being inverted by the typical software people tend to use, such as Negative Lab Pro. I wonder i this also contributes to the perception that thin flat (black and white) negatives are "easier" than denser ones - they do invert in NLP to decent looking positive images with almost no manipulation, while negatives with shadow detail and a fuller density range tend to produce (by whatever algorithm NLP uses) a first attempt that looks drastically overexposed and with no detail in highlights, until you crank the sliders to their extremes.

I think you're right. I use RawTherapee's Film Negative tool. I absolutely do not understand how it works. I just know I get better results than when I tried to adjust curves myself with GIMP. I have not scanned enough film to know if it does a better job with flat negatives, but if that were the case, I could easily see myself concluding that flat negatives is the way to go.

 

[dcy]

I think from what is now #1 it's clear that we're discussing a situation of negatives that didn't come out as intended, as implied by the mention of under- and overexposure and -development as well as 'lost' information. Hence, the term 'salvaging' seems appropriate to me.

I can confirm that this was my intent.

No, scanning and curve adjustments will not create a stellar (from a purely technical viewpoint) image from a significantly underexposed and underdeveloped negative. But it can be the ticket to a presentable or (for the purpose) acceptable image that the photographer prefers over...well, no image at all, or just a very muddy darkroom print that's barely recognizable.

Exactly. My holiday photo is my holiday photo. I can't take it again. If I screwed up the development or exposure, I still care about how much I can salvage.