Author Topic: A new scanning technique I discovered  (Read 7614 times)

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
A new scanning technique I discovered
« on: April 08, 2010, 11:30:19 PM »
I just discovered this technique while reading a French magazine. And to my great amazement, it really works!

They published an article on negative scanners which made me smile a lot for a good reason. They compared a 120 negative scanned with a Hasselblad Flextight X1 and an Epson V700. And by comparing images, they came to the conclusion that the results between both are so close that no real difference can be seen between both images.

But back to the technique...
Most flatbed scanners suffer in resolution because of the glass which changes the light path slightly. When scanning a B&W negative, you can get much greater resolution by scanning as 48 bit RGB than grayscale. When scanning grayscale, the driver mixes all colors equally which results in a slightly fuzzy image. Here is a simpler step by step of what they suggest.
1 - Scan as RGB 48 bit (16 bit per color) slide material with medium sharpening, no grain reduction and no DigitalICE.
2 - Invert the image to a positive (ctrl-i in photoshop)
3 - Open the channels layer box and inspect each colored layer individually. You will notice that some will be a lot sharper than others!
4 - Open the Adjustments/Channel mixer, check the monochrome box and set to 100% the layers you want to keep. Put the undesired layers to 0%
5 - Convert the image to 16 bit grayscale.
6 - Put the color profile to Gamma 2.2

That simple. Here are the screen captures... tell me what you think :)

[Sorry, image deleted during forum software upgrade. Please re-upload if so inclined.]
Francois

Film is the vinyl record of photography.

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
Re: A new scanning technique I discovered
« Reply #1 on: April 08, 2010, 11:39:26 PM »
And here's the final image.
Shot on Kodak Tmax 3200

[Sorry, image deleted during forum software upgrade. Please re-upload if so inclined.]
Francois

Film is the vinyl record of photography.

wavebeat

  • 120
  • **
  • Posts: 52
Re: A new scanning technique I discovered
« Reply #2 on: April 09, 2010, 01:44:03 AM »
This sort of thing makes my head burn. I'm sure it would burn my poor old emac too. ;D

Would it work in PSE2?

Good to know about the epson tho'

I like the look of the red channel best.

original_ann

  • Self-Coat
  • *****
  • Posts: 1,276
Re: A new scanning technique I discovered
« Reply #3 on: April 09, 2010, 02:22:26 AM »
I've read this before... I need to remember it and really try it out.  Thanks Francois!

LT

  • Global Moderator
  • Self-Coat
  • *****
  • Posts: 5,030
Re: A new scanning technique I discovered
« Reply #4 on: April 09, 2010, 03:37:57 AM »
there is a claim that it makes even more sense of you're using a tan and stain developer as you're supposed to see some of the grain masking benefits of the stain and better highlight separation.  I'm not sure about it myself, but that's what some say.
L.

vicky slater

  • Self-Coat
  • *****
  • Posts: 820
    • vicky slater
Re: A new scanning technique I discovered
« Reply #5 on: April 09, 2010, 10:57:35 AM »
francois, I'm picking up a diana roll today (probly not the best example :) so i'll try it later.
i always scan my b/w as col neg anyway but never tried doing it as a positive.

if i get any results worth showing i'll post it up.

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
Re: A new scanning technique I discovered
« Reply #6 on: April 09, 2010, 01:12:55 PM »
Actually, the Diana negatives probably won't change anything to the process.
On the neg I scanned, you could clearly see the grain on one of the layers. The screenshots had to be shrunk for the forum as I don't think anybody would have liked to have some 1900px wide images on their screen. This greatly affects how they look so consider them mostly for illustration purposes. For such results, I scanned at 3200DPI on an Epson 4490.

For now, I can surely say without any doubt that this was the sharpest results I ever got from this scanner without resorting to any kind of sharpening!

It probably would work too with stained negs though I think it would make things a bit harder considering the very greenish tones of the original. I suspect that not very much information would be visible on the blue and red layers... though never having tried such developers, I can't say for sure...

Also, I don't see why it wouldn't work in Elements 2. If it gives access to the RGB layers and the channel mixer, that's really all you need for this to work. If the channel mixer is missing, maybe you can just thrash a layer directly (I haven't tried) and convert everything to gray.

Francois

Film is the vinyl record of photography.

vicky slater

  • Self-Coat
  • *****
  • Posts: 820
    • vicky slater
Re: A new scanning technique I discovered
« Reply #7 on: April 09, 2010, 02:47:03 PM »
okay, i scanned a hassy pic instead.
but i realised i don't really understand the directions, i've never used layers but i did scan as slide and invert and mess around with channels etc...
so....first one is after trying to follow your directions, second how i normally do it and third in the middle of trying to do what you said and i really like it :)
Anyway, it was fun trying even though the results are meaningless because I didn't follow directions.




[Sorry, image deleted during forum software upgrade. Please re-upload if so inclined.]

astrobeck

  • Guest
Re: A new scanning technique I discovered
« Reply #8 on: April 09, 2010, 05:16:46 PM »
I scan in RGB as well, but never tried the slide trick.
Thanks for the screen captures, as I read instructions poorly and always need a "show and tell" to get the whole picture in my thick head of what's going on.   :)

I'd really like a scanner that sucks all the dust and teeny debris off my neg as it scans  too!

Thanks Francois!

Phil Bebbington

  • Self-Coat
  • *****
  • Posts: 2,568
    • Phil Bebbington
Re: A new scanning technique I discovered
« Reply #9 on: April 09, 2010, 05:55:34 PM »
Certainly gonna give this a try when  I next scan. Much of my scanning has been hit and miss, so trying something new might me good for me.

LT

  • Global Moderator
  • Self-Coat
  • *****
  • Posts: 5,030
Re: A new scanning technique I discovered
« Reply #10 on: April 09, 2010, 07:10:15 PM »
Pyrgallol stain is yellowy green, but catechol is browny red. 

Barry Thornton said this:

Quote
It turns out that a negative from a moderately fine grain [and staining] developer (like Exactol or PMK) requires less post-scan sharpening. This means less grain exaggeration, and the stain between the grains registers on the scanner to help ?smooth? the resulting file. Tanned/stained negatives scan beautifully.

Peter Hogan says (re his catechol based developer):

Quote
If scanning, scan in RGB and convert to monochrome in Photoshop to give a greater, smoother range of tones.

mind you - they were both trying to sell their products ... I've never really been into scanning negs to any great detail level etc so have never made a comparison.

L.

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
Re: A new scanning technique I discovered
« Reply #11 on: April 09, 2010, 07:32:04 PM »
okay, i scanned a hassy pic instead.
but i realised i don't really understand the directions, i've never used layers but i did scan as slide and invert and mess around with channels etc...
so....first one is after trying to follow your directions
I use Photoshop CS3 so I can't tell the exact order for your computer. The layers themselves remain the same, it's the channels I play with. But since they are RGB layers, I just named them so...

So, here is the short sequence:
Scan as slide, invert to positive, zoom in to grain size, open channels (Window-Channels on the top menu bar), Select Red, observe shadow detail and grain detail, select Green, observe, select Blue, observe, choose the one you like best, open channel mixer (Adjustments-Channel mixer), Put selected channels to 100% and undesired channels to 0% with the Monochrome checkbox (at the bottom) on, convert the image to grayscale.

I know it sounds like a lot of work but the detail is in the grain, something which we don't see well in web sized images (that's why I put blow ups of the picture I scanned). The image was zoomed about 130% for evaluating a single frame of 35mm. Also, if you scan at too low a resolution, the results won't be as drastic.

On medium format film, the most a Hasselblad Flextight can do is a 3200DPI. Same scanner does 6300DPI for 35mm. Sort of makes you think...
Francois

Film is the vinyl record of photography.

LT

  • Global Moderator
  • Self-Coat
  • *****
  • Posts: 5,030
Re: A new scanning technique I discovered
« Reply #12 on: April 09, 2010, 08:55:54 PM »
it dont know if this is breaking news or anything ... but, i find i get much sharper scans if i ignore my epson instructions and scan with the neg facing up then flip it post scan.  the other thing i found to help was to use a mamiya 6 to take your photographs - so effing sharp that it cuts your eyes ;) i miss it already.
L.

Phil Bebbington

  • Self-Coat
  • *****
  • Posts: 2,568
    • Phil Bebbington
Re: A new scanning technique I discovered
« Reply #13 on: April 09, 2010, 10:00:13 PM »
Any tips are always good, Leon. I use an Epson so will try flipping it next time - do you scan with everything turned off? You know, sharpening, ICE etc? I always have.

LT

  • Global Moderator
  • Self-Coat
  • *****
  • Posts: 5,030
Re: A new scanning technique I discovered
« Reply #14 on: April 09, 2010, 10:43:26 PM »
I do phil
L.


LT

  • Global Moderator
  • Self-Coat
  • *****
  • Posts: 5,030
Re: A new scanning technique I discovered
« Reply #16 on: April 09, 2010, 10:52:44 PM »
that's fine phil.
L.

vicky slater

  • Self-Coat
  • *****
  • Posts: 820
    • vicky slater
Re: A new scanning technique I discovered
« Reply #17 on: April 09, 2010, 11:00:53 PM »
heh heh :)

sapata

  • Self-Coat
  • *****
  • Posts: 2,073
  • "I want to be plastic" Andy Warhol
    • picturenoise
Re: A new scanning technique I discovered
« Reply #18 on: April 09, 2010, 11:25:26 PM »
Thanks for sharing Francois...

Since we're on the subject, has anyone got a good way of scanning 35mm negs without showing too much  of the film grain on the results?... I have an Epson V700 and it deals brilliantly with medium and large format but the 35mm are a real nightmare for me and I often get better results scanning the prints! 

LT

  • Global Moderator
  • Self-Coat
  • *****
  • Posts: 5,030
Re: A new scanning technique I discovered
« Reply #19 on: April 09, 2010, 11:29:58 PM »
has anyone got a good way of scanning 35mm negs without showing too much  of the film grain on the results?...  I often get better results scanning the prints! 

I think you've answered your own question there Sapata ;)
L.

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
Re: A new scanning technique I discovered
« Reply #20 on: April 09, 2010, 11:34:08 PM »
I use only a bit of sharpening (set to medium).

No grain reduction since my goal with a high res scan is to actually see the grain...

And absolutely no DigitalICE... probably the most important setting.
Not that DigitalICE doesn't do a great job, which it does... but only for color slide and color negative.
Once you understand a bit how the thing works, it becomes obvious.

DigitalICE works on the principle that color dyes in the film are totally invisible to infrared radiation. So after the first scan, you hear a click; this is when the scanner puts the infrared filter over the backlight. The scanner does a second pass looking for things which are opaque to the infrared. Scratches, by changes in the base shape tend to show very clearly... same thing with dust. The software then blends the surrounding area to make the artifact disappear.

Now with B&W film, the image is made from solid and opaque silver particles. When the scanner sees the grain, it thinks it's actually dust and attempts to remove it... not a pretty sight since it merges surrounding pixels.

Anyways... that's the long explanation to something simple.

Flipping the negative can work since you're moving the focal plane of the negative to what is probably a sweet spot on your machine.

As for excessive grain and removing it from 35mm negs, there are a few things to consider. First thing is that the scanner CCD's tend to exaggerate grain. This is probably due to the fact that a sensor is very precisely linear while the negative is a perfectly random source (think of photographing a TV screen with a digicam). This is why the best scans are done with a drum scanner which uses a laser to reach incredible resolutions. To minimize it, simply remove sharpening and add grain reduction. That should fuzzy-up the image to some extent. I noticed it works well on toy camera negs to increase the low resolution look of the images...
Francois

Film is the vinyl record of photography.

vicky slater

  • Self-Coat
  • *****
  • Posts: 820
    • vicky slater
Re: A new scanning technique I discovered
« Reply #21 on: April 09, 2010, 11:47:20 PM »
Can you ever have too much grain on a 35mm b/w?

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
Re: A new scanning technique I discovered
« Reply #22 on: April 10, 2010, 03:31:19 PM »
Well... not in my book.
I just always loved chunky grain... It's the type of thing which gives so much moodiness to the image.
Francois

Film is the vinyl record of photography.

sapata

  • Self-Coat
  • *****
  • Posts: 2,073
  • "I want to be plastic" Andy Warhol
    • picturenoise
Re: A new scanning technique I discovered
« Reply #23 on: April 10, 2010, 09:20:07 PM »
I also like grains on my pictures, it's just that sometimes I don't want to for a particular shot, especially when I shooting with a fine grain film...

sapata

  • Self-Coat
  • *****
  • Posts: 2,073
  • "I want to be plastic" Andy Warhol
    • picturenoise
Re: A new scanning technique I discovered
« Reply #24 on: April 12, 2010, 10:50:26 PM »
Inspired by Francois' discovered I decided to dig  some old negatives and experiment a bit with the scanning process and for this time I only used 35mm negs.

Since I haven't been able to do proper prints in a dark room, the whole scanning process puts me off when using 35mm cameras due to the grain effect.  Not that I don't like grain,  it's just hard to get a nice effect as they usually looks fake, especially on colour negs.

For these ones I used Epson V700 and  scanned the negs in positive 48 bit RGB ,  inverted and then I played with each of the slides.  Some small extra adjustments and here are the results,  all taken with a Voigtlander VitoII apart from the snow shot which was taken with a Canon EOS5.

I'm happier I have to say... and these are low res as the original ones are quite good !

[Sorry, image deleted during forum software upgrade. Please re-upload if so inclined.]
« Last Edit: April 12, 2010, 10:53:13 PM by sapata »

Urban Hafner

  • Self-Coat
  • *****
  • Posts: 1,545
    • Urban Hafner
Re: A new scanning technique I discovered
« Reply #25 on: April 13, 2010, 07:20:10 AM »
Well done Mauricio. Oh, and another Vito II user :) Great little camera, isn't it? Especially with slide film.

Urban

PS: And now image my frustration of trying to scan a Kodak Portra 800 shot in a Diana Mini (half-frame) :/

Francois

  • Self-Coat
  • *****
  • Posts: 15,552
Re: A new scanning technique I discovered
« Reply #26 on: April 13, 2010, 03:09:43 PM »
Really nice Mauricio...

Glad to see this trick is proving to be useful.
I for one won't go back to the old straight scan for B&W any time soon :)
Francois

Film is the vinyl record of photography.

sapata

  • Self-Coat
  • *****
  • Posts: 2,073
  • "I want to be plastic" Andy Warhol
    • picturenoise
Re: A new scanning technique I discovered
« Reply #27 on: April 13, 2010, 05:40:45 PM »
Well done Mauricio. Oh, and another Vito II user :) Great little camera, isn't it? Especially with slide film.

Urban

PS: And now image my frustration of trying to scan a Kodak Portra 800 shot in a Diana Mini (half-frame) :/

Yes... what a great little camera ! The design it's what got me at first... so pretty, haven't tried the slide film yet.
Cheers !

roryot

  • 120
  • **
  • Posts: 107
    • Cork Analogue Photographers
Re: A new scanning technique I discovered
« Reply #28 on: April 17, 2010, 08:38:28 AM »
If you don't a film scanner, but do have a digital slr and flash (shh I know, sacrilege around here), then this post on Nitsa's site might work for you. Haven't tried it yet, but it looks pretty good

http://nonphotography.com/blog/experimental/if-you-don%C2%B4t-have-a-film-scanner-then-diy/