Developing negatives for scanning

I thought I might get this forum off to a start with a discussion about developing negs for scanning. I've been scanning B&W negs for several years now and customizing the process for the needs of film scanners is something I think about quite a bit. I use a Minolta Scan Dual II for 35mm and an Epson 4180 for 120.

There are tons of discussions about this on PN and other forums, but overall the consensus seems to be that aiming for fine grain (rather than acutance) and taking care not to overdevelop are the best ways to ensure nicely scan-able negatives. I also think that the scanner used can make a big difference.

These days, I try to use relatively fine-grained film-developer combinations. I can get away with a lot for slow films (Efke 25 in Rodinal scans nicely) while faster films like HP5 Plus require a bit more care (the most scan-able HP5 negs I have were developed in Perceptol).

Any thoughts?

Also, for the most part, a thinner negative is preferred to a thicker negative as the majority of the scanners are over-rated by the manufacturers on the max dMax they can scan.

I find Ilford's XP2 super is great for scanning. I shoot this film alot for MF when I plan to make digital negs. Super fine grain, very little reciprocity effect, easy to scan since it's a dye-based C41 film.

I shoot HP5 and develop in tmax. I scan 35mm film in a polaroid sprintscan 400.
These images are the all negative an then a crop from a 18 X 12 inches print. I think the grain is very acceptable.

If you read the DR5 website, they suggest that developing film to be a positive is the best way to go.

Thanks for your input, everyone. Kerik, I agree that Ilford XP2 scans wonderfully... if I wasn't so addicted to playing with chemicals, I'd use it all the time. The "grain" is nice and smooth and renders well on scanning.

Aside from Ilford XP2, the best-scanning negs I have were made with the Efke 25 / Rodinal and Delta 100 / HC-110 combinations. (Efke 25 is virtually grainless, even in Rodinal). The worst-scanning negs I have are basically anything (except Efke 25) in Rodinal and HP5 Plus / Diafine.

The scans I have seen from DR5-processed positives are very nice but I think it isn't feasible (for $ reasons) for most amateurs to switch their B&W work to DR5. My own home-processed B&W positives scan quite nicely, though I can get the same quality from a negative process as well.

You can get chemical kits for the B/W reversal process, but I haven't tried any yet.

Kerik said:

I find Ilford's XP2 super is great for scanning. I shoot this film alot for MF when I plan to make digital negs. Super fine grain, very little reciprocity effect, easy to scan since it's a dye-based C41 film.

Click to expand...

This is very true in my experience. Less apparent grain, long tonal range and a lower Dmax make the chromogenic c41 films excelent for scanning.

Kerik said:

I find Ilford's XP2 super is great for scanning. I shoot this film alot for MF when I plan to make digital negs. Super fine grain, very little reciprocity effect, easy to scan since it's a dye-based C41 film.

Click to expand...

How does one control contrast with a C41 film, say with conditions that normally require very severe contraction development? I assume you do not develop in C41 yourself? I shoot a lot of color negative film with the intention of scanning but have not found a way to control high contrast scenes. Is it possible/feasible to requested reduced development from labs that process C41 film.

Sandy King

Sandy,

I don't process it myself, just take it to the local lab for standard C41. I haven't had trouble with excessive contrast, so I've not pursued how to reduce it. I've not had highlights that are so dense they can't be scanned. OTOH, I generally don't shoot under extreme contrast conditions.

Sandy,

I've never had big problems scanning the highlights with colour neg film, and I often use all the density range the film offers. The only channel likely to disappear off the scale is the blue channel - because the mask density adds to the image density - but that is an extreme case. Though the shadow range that exists on the negative is compressed for the jpg to keep midtone and highlight separation, the view of Midtown Manhattan in my gallery here should be a reasonable example of highlight separation with film that has had plenty of exposure.

What scanner and software do you use?

Best,
Helen

Helen,

I have not had a problem in the past with highlights in scanning color negatives made in scenes of normal and moderately high contrast. But how do you handle scenes where the contrast range indicates a SBR of +14 or higher? I just returned from a trip to Turkey where I photographed in a 6th century Byzantine monastery with both color and B&W film, where SBR values varied from +11 to +16. I tamed the contrast with the B&W film by developing in a very dilute solution of Pyrocat-HD with minimal agitation. But what should I do with the color film? Just have it developed by my local lab with normal C41 processing?

For scanning medium format color negatives I use a Leafscan 45 wtih the Leaf software, and scan in a dedicated B&W mode at 2450 dpi. This is an old, but very high quality scanner. It has a fairly high dynamic range, stated to be 3.7, tested at 3.5.

Best,

Sandy

Helen Bach said:

Sandy,

I've never had big problems scanning the highlights with colour neg film, and I often use all the density range the film offers. The only channel likely to disappear off the scale is the blue channel - because the mask density adds to the image density - but that is an extreme case. Though the shadow range that exists on the negative is compressed for the jpg to keep midtone and highlight separation, the view of Midtown Manhattan in my gallery here should be a reasonable example of highlight separation with film that has had plenty of exposure.

What scanner and software do you use?

Best,
Helen

Click to expand...

Sandy,

I have no experience of situations where I have tried to record detail over a 16-stop range on one piece of film, so can't be much help. Twelve or thirteen stops is about my limit for a single exposure on colour film - the limit being the film shoulder, not the scanner. I've tried it with two pieces of film, but that's not an ideal method.

Now that you have mentioned it, I feel motivated to do some tests with pulled C-41. It's a long time since the last tests I did. Those tests did not produce useful results, but scanning has changed things since then.

Best,
Helen

Helen Bach said:

Sandy,

I have no experience of situations where I have tried to record detail over a 16-stop range on one piece of film, so can't be much help. Twelve or thirteen stops is about my limit for a single exposure on colour film - the limit being the film shoulder, not the scanner. Best,
Helen

Click to expand...

Helen,

When you say that the limit is the film shoulder, are you saying that color negative films shoulder a lot with over-exposure? If so, do you know off-hand if the maximum Dmax is more or less than log 3.5? If less, I may be able to scan these negatives OK, irrespective of SBR conditions, since the dynamic range of my scanner is about 3.5, or slightly higher.

As always, thanks so much for your advice and keen insights.

Regards,

Sandy

Sandy,

I haven't come across a colour neg film with a density higher than 3.5 (or, more accurately, I can't find any of my test results that show a higher density). The densest that I've measured recently is 100UC, which has a comparatively high recommended density for an 18% grey card exposure (1.0 for Status M red) - ie it is denser than most other films when exposed as Kodak recommend. The maximum density I've measured with 100UC is 3.3 (Status M blue); 2.9 (Status M green); 2.3 (Status M red). That's at 13 stops over the exposure that gives a density of 0.15 above fb+f. The red curve falls away from the blue and green curves at the top end.

The colour neg films that I've looked at (Portra series and Fuji NPS, NPH, NPZ) have rounded shoulders (!). They fall off the straight line, then slowly flatten over about four stops - and the R, G and B curves lose their parallelism (the right word?). The manufacturers' curves generally don't continue beyond the straight line section.

I hope that makes sense.

Best,
Helen

sanking said:

How does one control contrast with a C41 film, say with conditions that normally require very severe contraction development? I assume you do not develop in C41 yourself? I shoot a lot of color negative film with the intention of scanning but have not found a way to control high contrast scenes. Is it possible/feasible to requested reduced development from labs that process C41 film.

Sandy King

Click to expand...

Sandy

You can get some contrast control with C41 BW film through exposure. At least that's what I got from the older Ilford XP2, Kodak TCN or even Konica Monochrome VX 400. Overexposure (use EI 100 instead of 400) should give you more contrast, and underexposure, less contrast. These films had tremendous latitude (except maybe for Konica's monochrome, long discontinued btw) which allowed generous over- and under-exposures. Nominally you could set your camera/meter to expose between EI 100 to about 1600 and still get useable negatives. And you need not alter the processing. A roll exposed with variable EIs can go straight into the processor and be processed normally.

For traditional silver negatives, thinner is indeed better. The exposed and developed film leader should show some transparency. You should be able to see through it easily when held against a bright area. My film developed in Rodinal or paRodinal scan very well with minimum of the noise produced by high densities.

Pushed negatives, having denser developed areas, usually don't scan well. Same could be said of overdeveloped or contrasty negatives.

Jay

Jordan said:

There are tons of discussions about this on PN and other forums, but overall the consensus seems to be that aiming for fine grain (rather than acutance) and taking care not to overdevelop are the best ways to ensure nicely scan-able negatives. I also think that the scanner used can make a big difference.

Click to expand...

As you say, it depends on the scanner. It also depends on the film, the exposure, the processing, the image itself, etc.

I think we can split the discussion neatly between emulsions which form images with metalic silver and those that form images with dye clouds. The discussion below just pertains to silver films.

Silver images are subject to the Callier Effect which is light scattering due in large part to the light not being able to penetrate the silver and instead reflecting off the silver structure. The basic result, in enlargers and scanners both, is that light scatter effects image contrast, with the dense areas of the film having more light scatter and therefore lower local contrast than the thin areas of the film. This type of local contrast non-linearity is problematic; many people find it objectionable.

Since there is a clear relationship between film density and the amount of Callier Effect we can see, it stands to reason that a less dense negative will give one better scanning results. In my own tests using a drum scanner and 5x4 Tri-X I have found this to be true. I develop my negatives to what a Zone System practicioner would call about N-1 to N-1.5. Below this and the tonal compression starts to become problematic for me.

The same should be true for CCD scanners, but my testing here has been limited. I suspect that the improvements I see with a drum scanner will be somewhat less with a CCD scanner. I'm postulating this due to the different ways the two types of scanners illuminate the film (drums use a spot, CCDs use at least a line).

Part of what I'm getting by decreasing density is a corresponding decrease in graininess. More than I could get with a solvent developer I suspect. When I tried different dilutions of XTOL to see what gave me good scanning results, I found that a 1:3 dilution worked best. That is, I biased the developer toward accutance over small grain. The results are excellent.

That's what I've found that works for me. Of course YMMV. And why guess when you can so easily know? Do some testing and see what you think.