Scanning a negative as a document

[Barry Kirsten]

There's been some recent discussion on RFF about using an ordinary scanner to scan 4x5 negs. Some have asserted that a negative can be scanned in a conventional scanner then inverted by software to achieve a positive. I was skeptical of this as I'd never heard of it being done, especially here on LFPF where a lot of scanning is done. To test it I scanned one of my 8x10 negs on my V600, as a 16 bit greyscale document ad 1200 dpi. The result was barely passable, requiring a lot of sharpening and some adjustment to curves. I then tried a 5x7 double sided x-ray neg and the result was awful. I assume the double density was too much for the scanner to handle. My conclusion is that it's a hit-and-miss technique and not up to the quality of 'ordinary' LF scanners like the V700. It got me thinking about why this is so. One obvious difference between the two processes is that one is reflective and the other transmissive. I also know that LF scanners seem to have a two lens system, obviously for better resolution. But I don't do much digital processing and am wondering if there's more to it than that. Any thoughts?

[Corran]

I saw that and I have no idea what those folks were smoking. Yes, film is transmissive, and the Epson and all negative scanners use a light source behind the negative (in relation to the lens/sensor) to get an image.

Scanning in reflective mode uses a light from the side of the lens/sensor, bouncing off the opaque paper. With film the results are as you found, bad - probably most of the light just bounced back after passing through the film. The multi-lens system does not relate to these two different scanning setups.

Here's a silly question, what would happen if you put a mirror behind the film? Probably just as bad. That's rhetorical, there's no reason to do this.

Notice non-film scanners only have a thin top lid, while film scanners have larger tops to hold the backlight. Put your film in front of a white piece of paper and you'll see exactly what a document scanner would see.

[Tin Can]

Abstract art

https://www.tate.org.uk/art/artworks/rothko-red-on-maroon-t01165

weed is becoming legal and LSG is becoming medicine...

[Jim Jones]

Many years ago I improvised a light source to mount above a negative on a cheap document scanner, but could never could get decent scans. Perhaps it would have worked better if I had disabled the scanner's internal lamp.

[C. D. Keth]

You could probably get something interesting, if very different to a transmitted light scan, by scanning with something matte black behind the negative to treat it like an ambrotype or tintype. Might have to tent in the scanner with black rather than press something to the back of the negative to keep the black whatever from being in focus.

[Tin Can]

Just tried scanning a 5X7 neg as document on V700, with OE white cover, got an image with glare, not sharp

With black folder cover I got almost nothing

Light matt grey may work better, I don't have any

[Barry Kirsten]

Thanks for the interesting thoughts. Actually Christopher's comment struck a chord with me. The second image, the 5x7 pinhole x-ray of the tree came out looking somewhat like an ambro- or tintype, which I thought at the time looked like a solarised print. In actual fact none of my image manipulating programs (Gimp, Pixelmator, Darktable) were able to invert this negative. Why? I don't know - apart from the obvious that the scan was not a proper negative scan. In both of these little tests the histogram was incredibly narrow, indicating that a lot of tones were not being seen by the scanner. My conclusion is that this is a dud process and not to be used if anything like decent results are required. I'll add this to the RFF thread.

[C. D. Keth]

Bummer, I thought that would work better. Ferrotypes are always pretty high contrast but they're workable. I wonder if the silver in film emulsion is too reflective or not reflective enough?

[Tin Can]

https://live.staticflickr.com/65535/49534374553_d361b0a692_c.jpg (https://flic.kr/p/2itbBNz)DOC SCAN before PS Invert (https://flic.kr/p/2itbBNz) by TIN CAN COLLEGE (https://www.flickr.com/photos/tincancollege/), on Flickr
https://live.staticflickr.com/65535/49534880031_0a9dffd0af_c.jpg (https://flic.kr/p/2ited4H)DOC SCAN AFTER PS Invert (https://flic.kr/p/2ited4H) by TIN CAN COLLEGE (https://www.flickr.com/photos/tincancollege/), on Flickr

[Barry Kirsten]

Bummer, I thought that would work better. Ferrotypes are always pretty high contrast but they're workable. I wonder if the silver in film emulsion is too reflective or not reflective enough?

Yes I've been thinking along these lines. In the case of my 5x7 x-ray neg there are two emulsions, one of which is pressed hard against the white backing in the scanner. Also the thin space of the polyester base between the emulsions may allow light piping to cause crazy flares and reflections. Who knows? Anyhow scanning negs in this way isn't reliable.

In Randy's example above I'm sure a lot of digital manipulation would be needed and would still leave a lot to be desired compared to the potential of the negative.