Sunday, 3 August 2014

George Peter Gatsis: Scanning Guide

Maybe you could make something out of this and post it on AMOC? This is the education I emailed to John Funk and Sean, for getting the detail in the darks and lights to co-exist in one digital page... This is what I did for the 1000 plus pages of Cerebus v1 and v2.

Cerebus #79 (October 1985)
Art by Dave Sim & Gerhard

---------- Forwarded message ----------
From: George Peter Gatsis
Date: Fri, Jul 11, 2014 at 2:04 PM
Subject: the steps...
To: Sean Michael Robinson

I'll use the following screen snapshots to illustrate my process in detail.

[Click images to enlarge]

This is the area I will be using to illustrate the process

This is to show a side by side comparison of the raw 2400 dpi and standard overall adjusted 2400 dpi.
Notice the detail is missing in the adjusted.

Here I have adjusted the raw scan with the LEVELs window floating, so you can see the manual adjustment that I made... with a comparison of the standard scan below.

Notice in my adjustment that the details in the white have swelled up... THIS IS IMPORTANT, the greater the detail we can get from the white areas that are surrounded by a sea of black will go a long way of maintaining a good print when the ink swells into the white area.

IF THIS PAGE was more white covered, than black, as it is now... the adjustment points would be shifted slightly to the right, somewhere between where it is now and the middle.

A larger area for a side to side comparison.
Notice the detail that I have maintained and the detail that the standard adjustment has lost.

A close up of standard adjustment

A close up of my adjustment.
Notice the more detail.

A close up of my adjustment.
Notice the less detail in the white areas around the lamp glow.

A close up of standard adjustment. 
Notice the less detail in the dark areas in the bottom right.

Overlaid standard adjustment and mine.
I only exposed or kept the detail of standard adjustment around the lamps glow area and the rest of the areas around are my adjustment. So we get the detail in the dark areas and in the light areas.

Notice the lack of detail all over this standard adjustment. BUT in the light area, at the base of Cerebus neck you'll notice 3 horizontal lines.

My adjustment has all the detail in the dark areas, BUT lost the detail in the light area, which is at the base of Cerebus neck - the 3 horizontal lines.

So, I've kept all the detail of my adjustment, which maintained detail in the dark areas and brought in the detail of the standard adjustment to keep the line detail at the base of the neck.

So in conclusion, it is a VERY manual process, marrying the best of both worlds DARKS and LIGHTS. Rarely one level adjustment works.... Most of the time you would have to:
  1. Take the raw scan
  2. Keep the source file on the base layer
  3. Duplicate the base layer twice
  4. Make your dark adjustments on layer 1... which you DODGE the dark areas to make the detail more pronounced for printing... so when the ink swells it won't fill in.
  5. Make your light adjustment on layer 2... which you BURN the light lines to make them thicker, so they won't break or disappear in printing.
  6. MASK layer 2
  7. And then expose only the light area details in layer 2 over layer 1
  8. Once you have the best of both worlds for layer 1 and 2...
  9. Group layer 1 and 2 into a folder and then flip it on and off to see what may have been missed or lost.
  10. Then if everything is good...
  11. Save PSD file
  12. Then flatten and SAVE a JPG file
  13. DONE.
Now do this 1000 plus times for Church & State 1 and 2... DONE!


AND... I recommend:

  1. RGB scanning
  2. Sticking with 1200 dpi... or 2400 if you can deal with it.
  3. Scan in everything RAW unadjusted
  4. Make a back up of the raw scans that are scanned every day to the CLOUD STORAGE and a local hard drive
  5. Go with the 13 steps above.

Chive on,

George Peter Gatsis
CEO / Executive Producer
The Black Diamond Effect Inc.
2200 Gerrard Street East,
Toronto, Ontario, M4E 2C7
T 416-440-4031
F 206-309-0055


David Birdsong said...

Thanks for this. I have a lot of my own work to scan and between this and SMR's information I think I can get the best versions out of the scans now. Priceless.

Sean Michael Robinson said...

Thanks for sharing, George.

Just wanted to mention that my approach is very different, although the goals are mostly the same. I'm using a combination of sharpening approaches (mostly Smart Sharpen and high pass/overlay method) towards the same ends.