scanning b/w issue

[Herb Cunningham]

I did not find my answer in any threads, so the question:

I am working diligently to get my skill levels up several notches, and to that end have done some film/developer testing, with the result that my Efke 4x5 Pl 100 developed in PMK 1 2 100 came out at 80 film speed, and I got what I think are really good negatives.

I shot several tests at slightly different exposures of the same scene, a woodland trail, with lots of pine trees and some sky, looking for the best depth of field and lens combination.

Most of the shots came out looking good to my feeble old eyes, very punchy and good shadow detail. I had them contact printed at a local pro lab, not have the time to do them myself, thinking I would later scan and print on my 2200 epson.

When I try to scan the negative on my 3200 epson, I cannot get the %$#@#$ thing to come out anywhere near a decent density. It is way too dense, i.e. the positives are hugely overexposed.

I did a Stouffer step wedge to arrive at the film/developer/time setup, and, while not any expert, I have been a serious photographer

for a very long time. I think I can tell a decent negative when I see one, but this has me flummoxed.

Any help would be GREATLY appreciated.

[Leonard Evens]

I'm not sure what to say. I've had no trouble scanning black/white negatives. I've scanned both my old 6 x 7 and 6 x 9 negatives, which were developed with conventional printing in mind, and my recent 4 x 5 negatives, which I planned to scan.

I use an Epson 3200 together with Vuescan under Linux. But Vuescan works the same way under Windows. It also allows you to measure the actual densities as opposed to the RGB values by right clicking and using the left mouse button. I haven't done an independent check with a densitometer, but the values Vuescan reports to me are pretty much what I would expect. They maximum densities seldom are over 2.0, and for properly exposed and developed black/white negatives they are usually just where I would expect them from Zone System considerations.

There are only two possibilities that occur to me. One is that you aren't setting the white and black points properly. If you can't set the white point high enough to bring down the RGB values for highlights to a reasonable range, That would suggest that your negatives are way overexposed. But that should be immediately apparent by looking through them. The usual criterion is that the densest parts of the negative shouldn't be so dark that you can't see anything at all through them.

[Donald Qualls]

I'm afraid you're up against a limitation of your scanner -- many/most current generation scanners have trouble handling a B&W negative of any sort (they're optimized for the dye-cloud color films, which pass infrared even in the densest areas), and may also have limitations on the maximum density they will accommodate.

I've gotten very good results scanning B&W with my old Arcus 1200, but it predates all the fancy ICE dust-removal systems that depend on IR transmission, and handles much higher density than most common scanners of the last couple years.

First thing I'd suggest is to assume manual control, insofar as your software will permit, of the scanning process -- set the levels yourself after previewing, turn off all sharpening and dust-removal features, etc. I don't see it making a lot of difference to scan as a positive transparency and invert manually, vs. letting the software handle the inversion, but it's a lot easier to set the levels when looking at a positive image on the screen, IMO. Also, if your scanner will permit, scan in 16 bits per channel, even if you have to demote the image to 8 bits to do any other editing; even very old versions of Photoshop will allow setting the levels (again, after scanning) before demoting, and that will do a tremendous amount to let you scan denser negatives.

At the bottom line, you may have to reduce development somewhat (and further increase exposure a bit) to control highlight density for negatives that will be scanned -- printing these will require a higher contrast filter or higher paper grade, but you should still be able to make good prints from negatives thin enough to scan well. Almost all consumer scanners work better with a negative on the low-contrast side of "normal" -- not quite N-1, more like N-0.5.

[Roger Hein]

Herb,

I agree with the point Donald made regarding neg densities - a slightly 'thin' neg seems to better match the capabilities of most scanners. Also I question whether the pyro 'stain' may be causing some problem also? One thing to try is scanning twice and adjusting the histogram in the scanning stage - one to favour shadows, one to favour high values - then 'joining' the 2 files to help increase dynamic range. Best scans I've been able to make are from plain old txp negs processed in hc110 or d23.

[Herb Cunningham]

Thank you guys both-I dug out a thin negative, and voila! The Silverfast 6ai that I am using may be the culprit, also-after Donald's post, I went to the epson pro scan mode and got an improvement, not enough for what I really needed, but noticeable. I will do some more tests next week and post the results on this thread.

I love this forum

[Erik Sherman]

I also agree that the problem may be with the scanner. Many simply don't have the dmax to handle serious negative density. Something I've done was to shoot T-Max and then use DR5 (www.dr5.com). The owner processes many types of B&W negative films to get transparencies. The T-Max ends up with something that scans pretty easily, and the results are nice. Or you can go with some other film for different feelings to the transparencies. The density range isn't what you'd get in a B&W print, but I think you'd need a much better scanner to get much of that. You could do double shots - one transparency and one negative so you can get the best of both worlds.

One other possibility that might help is doing multiple scans. You can always scan for the highlights and for the shadows separately and then piece them together digitally.

[Leonard Evens]

The Epson 3200 doesn't use infrared.

Also, I'm not the only one who manages with b/w on this and similar scanners.

IT IS NOT THE SCANNER!

[Leonard Evens]

Modern scanners like the Epson 3200 can easily handle the densities in a properly exposed black/white negative. As I noted, such densities are unlikely to be much above 2.0. It is transparencies with dmax's above 3.5 that such scanners have trouble with. A black/white negative with such a high density would not be printable by any means whatsoever.

I suggest you try Vuescan and take your time learning how to do it.

[Brian Ellis]

It is not the equipment (the 3200 works fine with black and white film), it is the person operating the equipment. You need to spend some time learning your software and learning how to scan. Theres a lot more to scanning than just popping the negative in and pushing the start button. I believe the Luminous Landscape web site has a tutorial on scanning and there are many other sources. Wayne Fulton's self-published book called "A Few Scanning Tips" was very helpful to me (despite the title, the book runs several hundred pages). It's dated to some extent because the equipment has changed so much since the book was written but the basic information is still very good.

[julian_4860]

Just as an experiment: get the demo of vuescan. Turn on all the 'advanced' options. Set the 'colour balance' to off. Scan as a 16bit 'image'. Don't mess with anything in the scanner and turn everything that sounds automatic to off (contact me off list if you want my settings). Set the exposure clipping to 0. You'll get a 'negative' in PS. Invert. Then adjust the levels so that only one pixel goes to black and NO pixel goes to white (but is only one step way from it). Thats about the best you can get from the neg - then add curves etc for contrast

[Bruce Watson]

Leonard and Brian are right. Scanning, like everything else in photography, has learning curves that must be mastered. Julian is also right - auto anything in scanning is going to be less than optimal for B&W - scanners and scanner software (all that I'm aware of) seem to be optimized for chromes.

The only way to get good at scanning is to scan. Scan the same negative over and over, changing one setting at a time until you learn what the various controls mean to image quality. You didn't learn how to use a view camera in a day, and you aren't going to learn how to scan in a day either.

[Herb Cunningham]

Thanks guys: as I said, I did not find any threads, which promted the question.

I have been scanning for about three years, this was the first time I had run into this problem. I have sold a few b/w prints, so while not real smart, not a complete idiot, just partially dumb.

What the problem was that I was using Silverfast's scan pilot, which has limited ranges. Once I quit using the pilot and did my own adjustments, everything came into place.

I really appreciate the inputs- and IT WAS NOT THE SCANNER

[Keith S. Walklet]

I personally prefer the straightforward interface of the Epson scanner software over Silverfast (I run a Mac). But if you are interested in trying to learn how to use Silverfast, there is a tutorial at lynda.com.

If you are not familiar with lynda.com, it is a terrific resource, an on-line library of video tutorials that one can access 24/7 for a month for only $25. The tutorials are excellent and the library covers most of the more popular graphics software out there. Before you commit to the monthly subscription, it is possible to browse sections of the library for free to get a feel for how they are put together. I've used the library for numerous applications. Give it a look.

As for setting the white and black point, the Epson interface is kind of cludgy. They have a histogram (at least on the 4870 driver), but it is not very accurate. I typically do a test scan, open the file in Photoshop and check my histogram there to make sure I got all the data. More often than not, the Epson histogram fails to show the trailing edge of data on both highlight and shadow ends of the tonal range, so it is necessary to leave room on both ends when adjusting the prescan histogram to ensure I get all the information.

[julian_4860]

Keith, one of the things you may like to try to get round that is to scan the neg as a positive. That worked for me on an epson 1680, although for BW Vuescan always got me better files.