How to expose B&W film for scanners that aren't good at picking up shadows?

[moodlover]

Portrait shooter here, using Tri-X 400. Key light f/4, fill light f/2 (so fill is two stops lower aka 1:3 lighting ratio). Everything shot box speed (so camera set to f/4, 1/200s).

I use a Epson flatbed scanner and due to it's low dmax it can't penetrate the shadows very well. If my subject is wearing black, it has a tough time picking up dark details and usually bunches everything into black mush.

I am wondering if there is a way to pump more light into the shadows of the scene without changing the key:fill lighting ratio on the face. Would it be as simple as boosting the key to f/5.6 and fill to f/2.8, while leaving my camera set to f/4? How well does Tri-X 400 handle overexposure without getting too grainy? What about Fuji Acros 100?

 

[moodlover]

As I think about this more logically, I am realizing that I should be metering for the shadows and not the highlights since with film we try to save shadow detail. My camera's aperture setting should be set to the weakest light in the scene to meter that as a middle gray. Since the fill is f/2, my aperture should be f/2 (yes, I am shooting shallow depth of field on purpose).

This way, the darkest parts of the scene render middle gray (and I can darken that back down during scanning) and the highlights fall where they may, as long as Tri-X can handle the 2-stop overexposure of the highlights. Maybe I can just cut off 15-20% of development time?

This would effectively be the same as leaving my camera at f/4, raising the fill power to f/4, and key to f/8. Thus, i've added much more light to the scene without changing the lighting ratio!

Does anyone agree?

 

[Chriscrawfordphoto]

Scanners pick up shadows easier than anything. Remember that shadows are the least-dense part of a negative. Dense highlights are harder to scan, but with my Nikon LS-8000ED I have scanned some badly overexposed films and gotten great scans, though they required some Photoshop heroics.

Generally, my experience is that with a good scanner a film exposed and developed normally is easy to scan.

 

[moodlover]

Scanners pick up shadows easier than anything. Remember that shadows are the least-dense part of a negative. Dense highlights are harder to scan, but with my Nikon LS-8000ED I have scanned some badly overexposed films and gotten great scans, though they required some Photoshop heroics.

Generally, my experience is that with a good scanner a film exposed and developed normally is easy to scan.

Not sure how you figure that. Least dense = least light = least information. The light that penetrates the film is not powerful enough to recover detail in shadows. Hence why a Noritsu or Frontier can do density correction for under/overposed film and Epson can't. Flatbed scanners have the most trouble picking up shadow detail without noise if there is not enough light hitting the dark areas. Highlights are almost impossible to clip in Epson Scan or Silverfast, but shadows have been the hardest to recover in my experience scanning hundreds of rolls.

 

[Corran]

Chris is right. The shadow information just has to be there in the first place. You can't "recover" what wasn't exposed. But the highlights are all there, it's just hard to push through the density.

You are correct with regard to positive film though.

Expose the shadows correctly and develop for the highlights, and your b&w neg scans will be easy enough to do. But optimal scanning density and wet printing densities aren't necessarily the same, if that matters...

 

[dshfoto]

Testing Film Exposure for Scanning

Testing Film Exposure for Scanning

I had a similar question, because I use a Imacon scanner, I wanted to know what was the "best" speed to expose differing films to get an optimal histogram. I also wanted this info for Diafine developer.
That is, I wanted the film when scanned not to loose the shadows or highlights.
So, I tested Tri-X (and a lot of other film) at differing ISO's. I happened to use a flash, so all I had to change was the aperture.
I used a Kodak Gray Card and a set up like in the photo. A white board helped to keep track of the settings for the shot.
Just shoot a roll at varing ISO's. I found that with Diafine, the highlights might have come out a little gray, bu that could be adjusted out in Photoshop (I actually use "Gimp") or in the Scanner Settings.
With Tri-X and Diafine, the conventional wisdom is to use an ISO of 800, when developing in Diafine, but I found that an ISO of 640 actually gave me a little more shadow detail.
Just take the attached photo, and open it in Gimp or Photoshop and check the histogram (levels) to see what I found.

 

[dshfoto]

Testing Film Exposure for Scanning

Testing Film Exposure for Scanning

Here's the file with curve adjustment applied. -- The point being that you can always increase the white value in the highlights with a curve adjustment. What you need is an exposure, film & developer combination that will capture all the shadow detail, without blowing out the highlights. The highlights can always be increased in the software, so long as you have all of the information in the scene (no high light blowout or shadow block out.

 

[rscheffler]

As I think about this more logically, I am realizing that I should be metering for the shadows and not the highlights since with film we try to save shadow detail. My camera's aperture setting should be set to the weakest light in the scene to meter that as a middle gray. Since the fill is f/2, my aperture should be f/2 (yes, I am shooting shallow depth of field on purpose).

This way, the darkest parts of the scene render middle gray (and I can darken that back down during scanning) and the highlights fall where they may, as long as Tri-X can handle the 2-stop overexposure of the highlights. Maybe I can just cut off 15-20% of development time?

This would effectively be the same as leaving my camera at f/4, raising the fill power to f/4, and key to f/8. Thus, i've added much more light to the scene without changing the lighting ratio!

Does anyone agree?

Yes, this is in the direction you should be going, IMO. Because it's negative film, you need to expose for the minimum shadow detail you wish to retain and if the scene has a wide exposure range (high dynamic range), adjust development times shorter to prevent excessive highlight density. Remember that on negs the shadow areas are lowest density and therefore allow the most light through to the scanner sensor, whereas the highlights block the most light.

The film develops uniformly but at some point shadow values stop increasing in density while mid and highlight values do. By controlling development time, you control density in the mid to highlight areas because shadow development will have already ceased. It will result in a meatier, flatter tonal range, but that can be tweaked after scanning.

Whether or not it makes a difference, a method I often used was to scan B&W negatives as positives. This method results in a negative image, which one then inverts in Photoshop. It avoids any hidden adjustments beyond your control by the scanner software when it makes the conversion for you. This way it's more or less a 'straight' scan which you can then more easily compare to the actual negative to determine if any values have been clipped.

 

[charjohncarter]

Scanners pick up shadows easier than anything. Remember that shadows are the least-dense part of a negative. Dense highlights are harder to scan, but with my Nikon LS-8000ED I have scanned some badly overexposed films and gotten great scans, though they required some Photoshop heroics.

Generally, my experience is that with a good scanner a film exposed and developed normally is easy to scan.

That is the way it should be, but I find some film scan better than others. I like Tmax 100 for shadow detail, and to keep it in the Kodak family I like Trix less for shadow detail scanning. Some traditional films like Arista EDU ultra 400 also scan poorly for me with shadow detail. Wet printing may be a different story.

Here is a Tmax 100 35mm scanned with lots of shadow detail, and next an Arista EDU 400 120 film: look at the 120 Arista scan in the upper middle.

Tmax 100 HC-110h by John Carter, on Flickr

Arista EDU ultra 400 HC-110h by John Carter, on Flickr

 

[moodlover]

Chris is right. The shadow information just has to be there in the first place. You can't "recover" what wasn't exposed. But the highlights are all there, it's just hard to push through the density.

Ah okay yes in this context Chris is correct, my apologies. I guess what I'm saying is my flatbed scanner has a tough time picking up shadow detail when there is not enough information there to begin with, and I'm wondering if the best way to boost shadow detail would be to just pump up the overall fill light in a scene. Scanning is all that matters to me, I am not printing. Thanks!

 

[moodlover]

I had a similar question, because I use a Imacon scanner, I wanted to know what was the "best" speed to expose differing films to get an optimal histogram. I also wanted this info for Diafine developer.
That is, I wanted the film when scanned not to loose the shadows or highlights.
So, I tested Tri-X (and a lot of other film) at differing ISO's. I happened to use a flash, so all I had to change was the aperture.
I used a Kodak Gray Card and a set up like in the photo. A white board helped to keep track of the settings for the shot.
Just shoot a roll at varing ISO's. I found that with Diafine, the highlights might have come out a little gray, bu that could be adjusted out in Photoshop (I actually use "Gimp") or in the Scanner Settings.
With Tri-X and Diafine, the conventional wisdom is to use an ISO of 800, when developing in Diafine, but I found that an ISO of 640 actually gave me a little more shadow detail.
Just take the attached photo, and open it in Gimp or Photoshop and check the histogram (levels) to see what I found.

Interesting, you are underexposing your Tri-X 400 by 2/3 of a stop? I am not sure if I have the courage to do that! Im more interested in overexposing since I never had a problem with highlights. Its always the shadows that I am having trouble with and want more detail out of. But your test is helpful, I might have to do the same!

 

[moodlover]

Yes, this is in the direction you should be going, IMO. Because it's negative film, you need to expose for the minimum shadow detail you wish to retain and if the scene has a wide exposure range (high dynamic range), adjust development times shorter to prevent excessive highlight density. Remember that on negs the shadow areas are lowest density and therefore allow the most light through to the scanner sensor, whereas the highlights block the most light.

The film develops uniformly but at some point shadow values stop increasing in density while mid and highlight values do. By controlling development time, you control density in the mid to highlight areas because shadow development will have already ceased. It will result in a meatier, flatter tonal range, but that can be tweaked after scanning.

Whether or not it makes a difference, a method I often used was to scan B&W negatives as positives. This method results in a negative image, which one then inverts in Photoshop. It avoids any hidden adjustments beyond your control by the scanner software when it makes the conversion for you. This way it's more or less a 'straight' scan which you can then more easily compare to the actual negative to determine if any values have been clipped.

Thanks for the confirmation. I have never lost highlight detail even in films exposed 2-3 stops over (which is usually how I shoot) so that is not my issue really. But lately for my studio shots I've been shooting completely at box speed and not satisfied with the shadows mushing together with no detail.

I don't mind a flatter scan as long as all the detail is there, I love when it's flat actually. Feels like color grading a movie still from a high end camera that starts really flat then you have all the latitude you need. Do you happen to know at what time Tri-X 400 ceases developing shadows? Ive never heard of such a thing, this is new to me!

I will try your scan-as-positive method and see if it works better, thanks!

 

[moodlover]

That is the way it should be, but I find some film scan better than others. I like Tmax 100 for shadow detail, and to keep it in the Kodak family I like Trix less for shadow detail scanning. Some traditional films like Arista EDU ultra 400 also scan poorly for me with shadow detail. Wet printing may be a different story.

Here is a Tmax 100 35mm scanned with lots of shadow detail, and next an Arista EDU 400 120 film: look at the 120 Arista scan in the upper middle.
r

These are amazing, I love how much shadow detail you get in these. What was this scanned on? Looks like something far superior to an Epson flatbed. Also, did you overexpose at all?

 

[x-ray]

Since TX was reformulated in 2006 ive had to rate it at ISO 250 running it in HC110 B.

Actually your light ratio is 4.1 not 3:1.

Show us some of your negs on a light box so we can determine if they're under exposed. Exposure controls shadows and development controls highlights.

If you're using something like an Epson v750 for example and selecting a film profile, the software will apply level and curves adjustments to your scan automatically. These are based on a perfect negative which doesn't sound like what you have. In reality it might cause clipping in the shadows and or highlights depending on how close your negative is to their idea of what perfect is. Most of us don't make perfect negatives no matter how much experience we have.

On my Imacon and V750 I never select a film profile. I set the scanner up to scan a linear scan then I make the decision on levels and curves that I want to apply. Every neg almost requires individual adjustment.

Show us a few negs.

 

[moodlover]

Actually your light ratio is 4.1 not 3:1.

How so? If the ratio is written fill:key then:

Key f/4, fill f/4 is 1:1
Key f/4, fill f/2.8 is 1:2
Key f/4, fill f/2 is 1:3
Key f/4, fill f/1.4 is 1:4

Show us some of your negs on a light box so we can determine if they're under exposed. Exposure controls shadows and development controls highlights.

My black areas are clear on the film, they are definitely underexposed. Hence me asking how I can expose my shadows better so they are not clear, but have the minimum amount of detail necessary.

If you're using something like an Epson v750 for example and selecting a film profile, the software will apply level and curves adjustments to your scan automatically. These are based on a perfect negative which doesn't sound like what you have. In reality it might cause clipping in the shadows and or highlights depending on how close your negative is to their idea of what perfect is. Most of us don't make perfect negatives no matter how much experience we have.

Very true. Im not using any profiles, I am scanning completely manually and attempting to boost the shadows but the information is not there to begin with so it just boosts blocked up blacks into gray.

 

[x-ray]

F2.8 allows half the light of f2 to enter and f4 allows half the light of f2.8. 1/4 the light or 1:4. The difference in light between 1.4 and 4 is 1:8 and if you use f5.6 it's 1:16. F8 would be 1:32 and so on.

No scanner will add information. At best it can only give you back what's there.

Rate your film at ISO 250 and give that a try.

 

[charjohncarter]

These are amazing, I love how much shadow detail you get in these. What was this scanned on? Looks like something far superior to an Epson flatbed. Also, did you overexpose at all?

I'll PM you tomorrow.

 

[Chris101]

Scanners pick up shadows easier than anything. Remember that shadows are the least-dense part of a negative. Dense highlights are harder to scan, but with my Nikon LS-8000ED I have scanned some badly overexposed films and gotten great scans, though they required some Photoshop heroics.

Generally, my experience is that with a good scanner a film exposed and developed normally is easy to scan.

Right? Under exposed film is very difficult to bring anything out, but if there is anything in there, 'chopping' can bring it to fore, but the picture changes considerably.

The best way to approach the problem is to expose the film to fully differentiate the shadows. Then, when developing, keep the highlights within the maximum density that your scanner can see, before it all becomes black.

Oh wait. Expose for the shadows, and develop for the highlights applies to scanned images as well??

Cool. (print and scan from the same negative)

 

[x-ray]

Expose for the shadows and develop for the highlights applies to all negative films whether scanned or wet printed.

Like I've recommended in threads before you should buy the later books by Ansel Adams on exposure and darkroom (processing). Before you go much farther you need to study the fundamentals. You need to understand the relationship of aperture, shutter speed and ISO. You need to know how changes in each effect your results and gain a better understanding of light. This is why you thought you were getting a 1:3 light ratio when it was actually 1:4. You also need to learn how to determine the ISO of your film and correct development. You've got to know certain fundamentals if you want to be successful with your project.

 

[charjohncarter]

Expose for the shadows and develop for the highlights applies to all negative films whether scanned or wet printed.

Or expose for the shadows, develop for the middle tones, and agitate for the highlights.