overexposing and underdeveloping optimal for scanner?

vicmortelmans

Well-known
Local time
1:38 AM
Joined
May 2, 2005
Messages
316
overexposing and underdeveloping optimal for scanner?

I'm thinking (sporadically---time is scarse) about theoretical aspects of b&w development and film scanning and image processing.


high-signal noise

I'm looking for the optimal development practice for scanning. Something obvious, but still surprising about the principles of negative photography, is that the signal noise of a scanned picture is in the highlights, while intuitively one would associate noise to the low signal levels, shadows.

I started off developing my film much too dense, so the scanner has trouble getting enough exposure through the highlight area's. Once inverted, the noise is immense (not grain! scanner detector noise). I tried to increase the scanner's exposure, but then the shadows are 'blown out' (the high signal area's in the scanned negative are burnt).

Conclusion: does this mean that the density range of my (over-)developed negatives is too high for the scanner?


development technique

The solution is reducing the development time, which I will try next time. But I fear that this will cost me shadow detail... Or are there other tricks to 'soften' development, so the scanner will catch both highlights and shadows? Like reducing developer solution (now: rodinal 1+25) or changing agitation (now: continuous).


exposure technique

Can I tweak the ISO setting for better scanning performance as well? If I'm right about this, overexposing in combination with underdevelopment will give me a wider coverage of scene intensities (extending the cut-off in shadows and hightlights) and also a lower maximum density. Sounds ideal for scanning, as you then have a dataset that is more flexible towards image processing and will not have the ennoying noise in the highlights. The only cost factor here is the contrast at medium levels, which is relatively low.


Still, I don't get a good feeling of what a well-exposed-and-developed negative should look like. I really liked the overexposed negatives, and could see all highlight and shadow details with the naked eye, but after scanning it becomes a reall mess...

Going to experiment somewhat more...

Groeten,

Vic
 
There are too many variables that could be giving you a problem.

What you're describing sounds to me like a color-profiling issue.
 
Last edited:
If you're shooting traditional silver-based films, it's definitely a strong possibility that you're getting too much highlight density for your scanner. I've had problems with that myself. A bit more exposure and a bit less development should cut down highlight density without affecting shadow detail.

To get a sense of what a well-exposed and well-developed negative looks like: An old-time rule of thumb for judging highlight density is that in normal room light, if you lay your negative flat over a piece of ordinary-size printed text, you should just barely be able to read the text through the densest areas of the negative. If you can do this, the negative is more or less correctly developed. If you can see textural detail in these areas AND in the thinnest areas of the negative, it's also correctly exposed.

If you can't read the text through the densest areas, you need to cut back your developing time a bit, and if you can't see detail in the thinnest areas, you need to increase your exposure a bit. Keep proceeding in small steps until you get negatives that are well-balanced.

Another thing: I noticed that the 'auto exposure' setting on my film scanner didn't do a good job on silver-based films; evidently it was set up for chromogenic films, which have much less highlight density. I had to switch to the manual exposure setting and bump up the scanner exposure a bit to push more light through the densest areas of the negatives. If your scanner software has this option, you might give it a try if correcting your development time still doesn't give you good results.
 
vicmortelmans said:
high-signal noise

What you say on this topic is sensible and correct, for the most part (always some variables, of course). Overdevelopment == dense negatives == too dense for the scanner to punch through. The result is scanner noise. I find that in vuescan employing a variety of techniques can help, since I run into this a lot with IR film (still haven't quite gotten that down yet).

development technique

The solution is reducing the development time, which I will try next time. But I fear that this will cost me shadow detail

NO. (not in a mean way). Shadow detail is controlled by exposure. Reducing development time will not reduce the amount of shadow detail you obtained via exposure. I guess if you hadl ike a 30 second dev time you'd lose some shadow detail on the curve, but other than that what you're looking for is exposure, which means getting your EI dialed in.

... Or are there other tricks to 'soften' development, so the scanner will catch both highlights and shadows? Like reducing developer solution (now: rodinal 1+25) or changing agitation (now: continuous).

Overall contrast is controlled via development. So you can reduce your agitation (continuous>intermittent), for instance. Dilution with some developers gives you lower contrast but, more than anything else, it gives you more time with which to play with. for instance, if you're finding that developer X at 1+25 is already only 5 minutes, cutting down to, say, 4 minutes is creeping into uneven development areas. But if you diluate 1+50, you might be up at 8 or 9 minutes. Cutting by 20% then is still in a safe range.

You can also use divided developers, stuff like that. But I think time is your first thing to try. Then try changing agitation. 1 thing at a time.

exposure technique

Can I tweak the ISO setting for better scanning performance as well?

Yup. You have to dial in your exposure to give you proper shadow detail for your scanner. You are still tying exposure and development a little too closely, though they are absolutely related and, usually, reciprical (add exposure, decrease development to get same result). You get lower maximum density through less development. Period. You decrease contrast by overexposing (turning pitch black into detailed shadow, and detailed shadow into middle grey, for instance).

But, of course, overexposing also means your are pushing highlight detail into way blown out. So you have to then reduce development to bring that back. So they are related, as I said, but I think you haev them a bit too connected. It's more like cause and effect.

I would not blatantly overexpose, though. Find a couple of scenes which are representative of the type of stuff you do. Shoot it at, say, box speed and 2 stops under and over, in 1/3 or 1/2 increments - whatever you got. Then develop at the normal time. Some will end up massively overdeveloped (highlights too dense) but ignore that for now. Look at the areas that were shadows in your scene, after you've scanned them with nothing but auto levels. Ignore tha incredibly bright areas. Which EI gives you the shadow detail you want? That should be roughly your EI.

This is not a particularly controlled test, btw, but most RFF's aren't carrying around spot meters.

To then get your development time, set your camera at that EI and shoot a variety of scenes on separate rolls (or on a single roll but in stretches of low, middle, and high contrast so you can chop the roll up). Then develop at different times until you get the density in the highlights right.

Voila.

I really liked the overexposed negatives, and could see all highlight and shadow details with the naked eye, but after scanning it becomes a reall mess...

Negatives are inherently low contrast. This is something that people don't realize. If you look at a negative, and you think the contrast of the negative, unto itself, looks good, then it's too contrasty. You have to rely on the interpretive media or device - an enlarger and paper or a scanner - to add that contrast. So you must read the negative in relation to the end target, not unto itself.

allan
 
Wow, this is an interesting thread! It had never occurred to me that someone would want to process their film specifically so that the negatives are easy to scan, but it makes sense, I guess. Would this be detrimental to ease of printing? I would have imagined that a very good negative to print from would be a very good negative to scan from. Do negative scanners work so differently?
 
jlw said:
If you can't read the text through the densest areas, you need to cut back your developing time a bit, and if you can't see detail in the thinnest areas, you need to increase your exposure a bit. Keep proceeding in small steps until you get negatives that are well-balanced.

Hi,

I use vuescan and do a raw scan, with manual exposure. The postprocessing is out of hands of the scanning software.

Jlw, your statement above seems to indicate I'm on the right track: my densest area's certainly won't allow to read through, so I have to cut down development. On the other hand, I have the feeling that the shadow detail will suffer from this, so I'll overexpose.

So the next 125 roll I'll expose at 100. Or should I increase exposure more? And 400 at 250?

Can anyone tell what the influence of developer concentration may be? Can this help for better scannable results?

Learning on,

Vic
 
Wow, Allan, thanks for all that feedback. Gives some good answers and confirmations! And you already answered my dilution question!

Groeten,

Vic
 
The answer is.....kinda. You want a relatively low contrast negative for a scanner, because the light is fairly intense and focused.

There are, however, two types of enlargers. A diffusion enlarger has a softer light, so you usually want a more contrasty negative to make printing easier. But a negative that's aimed for scanning works well in the other type of enlarger, the condenser type.

If you are going to bother developing your own film, in my opinion, and not go the Diafine route (which is a perfectly fine way to go but doesn't give you as much control), then I think it is important to dial in both your exposure and development in regards to your scanner.

For instance, I expose TXT at 320 when using a spot meter because that gives me solid shadows with my scanner. But at 250 with a center-weighted meter since I don't have as tight of control over what I am metering. Right tools for the job. And the right methods with those tools.

allan



Crasis said:
I would have imagined that a very good negative to print from would be a very good negative to scan from. Do negative scanners work so differently?
 
kaiyen said:
NO. (not in a mean way). Shadow detail is controlled by exposure. Reducing development time will not reduce the amount of shadow detail you obtained via exposure. I guess if you hadl ike a 30 second dev time you'd lose some shadow detail on the curve, but other than that what you're looking for is exposure, which means getting your EI dialed in.

Allan, this is hard to understand if you don't see it happen! And of course I'm not going to ruin some film just to test it out... So I'll take it for granted.

Groeten,

Vic
 
I think gabrielma is pretty much on thhe right track, but I would heavily underscore the influence of both scanning hardware and software here, and the real necessity of knowing the parameters of both.

About the only generalization I can make regarding the scanning of conventional b/w film is that the basic rules regarding development for wet-darkroom printing (overexposure by a certain amount, normal development) aren't ideal. I've found that simple, by-the-book exposure seems to play better with the scanner, but then, I haven't played around much with different exposure/development arrangements (and I also use chromogenic film at least half the time I shoot b/w).

In any case: know your scanner, know your scanning software.


- Barrett
 
kaiyen said:
The answer is.....kinda. You want a relatively low contrast negative for a scanner, because the light is fairly intense and focused.

There are, however, two types of enlargers. A diffusion enlarger has a softer light, so you usually want a more contrasty negative to make printing easier. But a negative that's aimed for scanning works well in the other type of enlarger, the condenser type.
allan

Thanks, that's certainly interesting. I'm printing B&W with a colour head right now since the setup was cheap to get. Beseler 23CII with a colour head. I know that's not ideal and I want to eventually get a condenser head or a different enlarger with that head as a default, but that's money I don't have at the moment.

Hey, since you're here.. do you happen to know the drawbacks to using a colour head for B&W printing?
 
Color heads are great for B&W printing, actually.

They are diffusion enlargers so they aren't quite as tack sharp, but you can get away with contrastier negatives. Also, rather than using contrast filters, you can dial in magenta to produce the effect. And you can do, say, a 2.25 contrast filter rather than just either a 2 or 2.5.

Having said that, I do prefer a condenser enlarger myself. Sharpness is nice. But keep your negs clean.

allan
 
vicmortelmans said:
Allan, this is hard to understand if you don't see it happen! And of course I'm not going to ruin some film just to test it out... So I'll take it for granted.

Groeten,

Vic

Well, Vic, you wouldn't see it happen, because nothing would happen! :)

Expose correctly and the shadows will look the same. Or expose the same scene over and over at the same setting and develop differently and you'll still get the same shadow detail, even though the highlights will go up and down*.

allan

*Technically, shadow detail does go up slowly as you develop longer and longer. But for this discussion let's stick with shadow detail is controlled entirely by exposure.
 
patrickjames said:
I have been debating this myself lately, but I am definitely not going to give up better darkroom prints in order to get good neg scans.

These are not remotely mutually exclusive. If you calibrate for a scanner and use a diffusion enlarger, you need a 3 filter instead of the 2 you usually use, for instance. The problem is that if you calibrate for printing with a 2 on an enlarger you run the risk of getting a negative that would print fine but wouldn't scan as well. It's wiser, therefore, to calibrate exposure and development for a scanner. You can always print appropriately with the right filters.

I would also overexpose a bit too to take advantage of the fact that you will be adjusting contrast in Photoshop.

This is something I say over and over again, so I apologize beforehand. You shouldn't think of it as overexposing. You should think of it as just exposing correctly. But for a different target.

allan
 
Barrett,
I hope I don't offend at all with this post, I'm just responding.

amateriat said:
About the only generalization I can make regarding the scanning of conventional b/w film is that the basic rules regarding development for wet-darkroom printing (overexposure by a certain amount, normal development) aren't ideal.

damn straight. We need to take the concepts of wet-printing - expose and develop to get negatives that are easy to print. But we need to throw out the specific times/EIs that one used in the past.

I've found that simple, by-the-book exposure seems to play better with the scanner, but then, I haven't played around much with different exposure/development arrangements (and I also use chromogenic film at least half the time I shoot b/w).

What you're really doing, then, is just producing a thinner neg. through slight underexposure with normal development. Unless, of course, you are using a speed-enhancing developer in which case you got some magic going on :). But ideally you want proper exposure with the appropriate normal development time (normal to that exposure time, not to what you've done in the past, necessarily).

Chromogenic film is a whole different ballpark. Ridiculous latitude and no control over development time are both good and bad.

allan

ps - at this point, I should just write a whole report on this...


- Barrett[/QUOTE]
 
Last edited:
kaiyen said:
Color heads are great for B&W printing, actually.

They are diffusion enlargers so they aren't quite as tack sharp, but you can get away with contrastier negatives. Also, rather than using contrast filters, you can dial in magenta to produce the effect. And you can do, say, a 2.25 contrast filter rather than just either a 2 or 2.5.

Having said that, I do prefer a condenser enlarger myself. Sharpness is nice. But keep your negs clean.

allan

Ah, well that's interesting. Yes, I do dial in yellow and magenta.. I have a chart somewhere saying 10 magenta is grade 2, 0 magenta and 30 yellow is grade 1.5 or something. *shrug*

I didn't realize that diffusion was worse than colour. That's rather nice. I'm a fan of sharpness which is why I'd like a condensor head but if it's not that much better, I won't start saving.

Thanks lots for the informativeness :)
 
No, diffusion enlargers aren't worse than color enlargers. It's that color enlargers are _all_ diffusion enlargers, whereas you can find condenser enlargers but they are B&W only.

Condenser enlargers are sharper, but are less forgiving of contrasty negatives. And, as with many situations involving photography, getting a better enlarger lens will help a lot with sharper.

allan
 
Last edited:
I sense that this would create very flat negs - perhaps best for scanner with very narrow DMAX, but you'll have to curve it to get things into a realistic mode.
 
amateriat said:
I think gabrielma is pretty much on thhe right track, but I would heavily underscore the influence of both scanning hardware and software here, and the real necessity of knowing the parameters of both.
This is something that is always neglected, and it causes a lot of pain for many. But looks like this is not what was asked for, so I'll just reduce my initial post to mentioning the fact, and zap the techno jambo.
 
shutterflower said:
I sense that this would create very flat negs - perhaps best for scanner with very narrow DMAX, but you'll have to curve it to get things into a realistic mode.

Shutterflower,
What is the "this" to which you refer that would produce flat negatives?

allan
 
Back
Top Bottom