Best BW film for scanning

[Steve Pelton]

Greetings all,
I am looking to get the best possible scans of B&W with an Imacon 343.
I have tried Delta 100, HP5, TMX 100 & 400 and Acros. I have souped them in TriX dev and Xtol. So far Acros/Xtol has given the best contrast for scanning but I think there is still room for improvement. Before I spend the next month trying different film/dev combos, what are you guys using? What density of negative should I be targeting, and is this different than for a wet darkroom?

thanks,
Steve Pelton

[Brian Ellis]

I don't know that any one film is inherently better than another for scanning. But regardless of that, in general I think you'll be better off developing to a lesser contrast for scanning than you might have for darkroom work. I no longer do plus development at all and have reduced my N time by 10%. The 10% is more or less an arbitrary compromise between developing everything at N - 1 which seemed a little extreme but still making sure that the negative is on the flat side. Maybe others have a more scientifc approach but this has worked pretty well for me.

[Ron Marshall]

Exactly what Brian has said. Negatives slightly thin and slightly flat, as compared to the optimum for silver printing have worked the best for me.

[Ed Richards]

Some of us have found Tmax in Xtol 1:3 to be esp. good for scanning. 100 if do not need the speed, 400 if you do.

[Helen Bach]

I beginning to think that Fuji Pro 160S is the best B&W film for me.

Best,
Helen

[Bruce Watson]

Exactly what Brian has said. Negatives slightly thin and slightly flat, as compared to the optimum for silver printing have worked the best for me.

I'll third that. Unlike Brian though, I now develop everything to N-1. My film would be a pain to print in a darkroom. For the record, I'm using 5x4 Tri-X souped in XTOL 1:3. This fall/winter I'm going to be investigating 400Tmax, first in XTOL 1:3, then maybe in Pyrocat-HD.

I'm thinking of making the switch for the improved reciprocity performance of 400Tmax. If it scans a little better (which it might, or might not) I won't complain.

[Ron Marshall]

I beginning to think that Fuji Pro 160S is the best B&W film for me.


It's a wonderful film that has weaned me off of chromes. Scans very well.

[Leonard Evens]

If you develop b/w film normally, the density range is well less than anything a modern scanner can handle. So, theoretically you should be able to increase the density range by overdeveloping and do better. That way you won't end up with values spread out too far in the range 0..255. But, in practice, it doesn't seem to work out that way. Perhaps, overdeveloping increases graininess too much, and perhaps scanning software is not optimized for developing b/w film that way. In practice I use normal development, but when I've underdeveloped by mistake, I haven't had any problems scanning.

Some poeple mentioned densities. I use Vuescan, which allows you to read what are purported to be raw densities above base rather than RGB values in the range 0...255. As best I can tell, it gives the kinds of values I would expect, so, even if it isn't exactly right, I think it is close. Which other scanning packages allow you to display actual density values?

[Helen Bach]

If you develop b/w film normally, the density range is well less than anything a modern scanner can handle. So, theoretically you should be able to increase the density range by overdeveloping and do better. That way you won't end up with values spread out too far in the range 0..255. But, in practice, it doesn't seem to work out that way. Perhaps, overdeveloping increases graininess too much...

...which is exactly why I put a lot of effort a while ago into finding a good reversal process for B&W film tailored for scanning rather than projecting. That is one way of increasing the density range without the increase in graininess that comes with simply increasing the development of a negative*. Because the image is not intended for viewing the image colour is not important, so the choice of second development method is wide.

*The increase in the graininess of the original negative image is offset by the reduced amount of contrast, and hence graininess, amplification that is required to get to the print, of course.

Best,
Helen

[Bruce Watson]

What density of negative should I be targeting, and is this different than for a wet darkroom?

To optimize a B&W negative for scanning it should be somewhat less dense that a negative optimized for the darkroom.

B&W negatives build up density by building up metalic silver. Scanning and darkroom printing both suffer from the Callier effect. That is, the light projected through the film is scattered by the metalic silver. What you get is a non-linear loss of local contrast. That is, the contrast in the highlight areas (dense on the negative) is less than the contrast in the shadows. Another way to say it is the highlights tend to block up.

If you go more dense, you get more density and thus more Callier Effect. You also start to get interesting but not very pretty artificacts in the negative. In particular, graininess increases with density. The grain structure gets mushy as well. Ugly.

The reason going to lower density works when scanning is the nature of the beast. If you set your black and white points properly most modern scanners/software will fit your density range (whether 0.3 or 3.0) into the scanner's digital range (0-255 if 8 bit, 0-4095 if 12 bit, etc.). It's an exact fit by definition.

That said, if you are ever going to use the negatives in the darkroom, then optimize for darkroom work. If it will print easily in the darkroom, it will scan just fine.

But you don't have to believe me. Or anyone else. Run some tests and see it for yourself. Why guess when you can know?

[Kirk Gittings]

Negatives slightly thin and slightly flat, as compared to the optimum for silver printing have worked the best for me.

In the old days (like last week) a "thin" negative meant one which was underexposed, though it appears you are talking about under developement. I do not recommend underexposing negatives for better scanning though slight under developement is not a bad idea.

And contrary to some opinions, I still do +1 and +2 developements, because the increase in grain in the negative from the developement is less problematic than applying a steep curve to the file which produces increased granularity and noise in the transition areas between tones, like in cloud filled skies.

[JoelBelmont]

I try to shoot TMAX 100 in flat light and do N+1 in 1:2 Xtol. This has been tailored for darkroom printing to require little work in getting the contrast I want.

Will it be hard to digitally scan and print a neg with less info, or does it matter since it is similar to how the print should look anyway? It seems the consensus is to develop with less contrast. But I still prefer the option of darkroom printing. Will the digital prints look bad if I continue to develop with more than average contrast? I don't mind a slightly darker, printed-down look.

Thanks,
~Joel Belmont

[Steve Pelton]

Hmmmm, seems a lot of you go thin. I started this way, but found the files grainy and very very flat. I had to build them up with adjustment curves in PS to get acceptable density, and in the end the results look OK.

Through experimentation I have also found over developing to increase contrast works very well but I have not taken a very scientific approach to it so far.
Sounds like more experimentation is in order.

thanks for the input,
Steve

[buze]

I am also of the school of "slightly thin" development for scanning. Best way to work on constrast in PS /without/ touching a curve is to change the GrayScale "color" profile; between 10% and 30% will surely give you a nice constrast for almost any scan, even on thin negatives.
And that's "free" contrast, it doesn't touch a pixel of the file, just the way it is rendered eventualy. You can then do you normal processing without having to raise the grain level etc. In particular the "threshold" in sharpening can be kept low without picking up the grain and getting overboard.

Experiment with that, it's magic.

[paulr]

To optimize a B&W negative for scanning it should be somewhat less dense that a negative optimized for the darkroom.

I haven't experimented with optimizing negs for scanning, but my darkroom negs are all on the low density, low contrast side ... this just happens to be what works best with my paper and other chemistry. And these negs indeed scan really well. No complaints at all. I do find that some of my older negs, which are denser and more contrasty, are harder to get a great scan from.

It seems to me that picking a film for scanning is easy compared with picking one for traditional processes. Since you have absolute control over the tonal scale, you don't have to worry so much about it. You can draw whatever curve you like. This allows you to pick film based on more objective and quantifiable qualities ... speed, resolution, grain, freedom from streaking, etc. etc.

I've had great luck with TMX. detail for miles, no visible grain, even development, and a long scale that gives me more detail in the extremes than i even use. Speed is acceptable, but not stellar (with my developer I get around asa 50). This might be the biggest shortcoming ... can be hard to use if i need much depth of field and if things are blowing in the wind. But at night, because of its great reciprocity characteristics, it feels like a very fast film.

[PViapiano]

Paul...

What do you dev your TMX in?

[paulr]

i've been using this for the last ten years:
http://www.jackspcs.com/gpq.htm

although now that i'm mostly scanning and printing digitally, i don't think i'll be as picky about it ... anything with decent speed/grain/resolution characteristics should let me get my prints the way i like.

[JoelBelmont]

I am also of the school of "slightly thin" development for scanning. Best way to work on constrast in PS /without/ touching a curve is to change the GrayScale "color" profile; between 10% and 30% will surely give you a nice constrast for almost any scan, even on thin negatives.

Can you explain how to change the greyscale 'color' profile? I am assuming this is a similar concept to color filtration.

Thanks,
~Joel Belmont

[Kirk Gittings]

I use Vuescan, which allows you to read what are purported to be raw densities above base rather than RGB values in the range 0...255. As best I can tell, it gives the kinds of values I would expect, so, even if it isn't exactly right, I think it is close. Which other scanning packages allow you to display actual density values?

Well, Silverfast for one.