Tuesday, 23 May 2017

Paper to Pixel to Paper Again: Part 17

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 9 | 10 | 11 | 12 | 13 | 14 | 15
16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29

A guide to creating the best looking line art in print in the new digital print world

Part 17
Working with "Bad" Negatives, B


This is the seventeenth installment of Paper to Pixel to Paper Again, a series that explains (in an overly thorough manner) the how-to's of preparing line art (and later in the series, color art!) for print.

And as always, if you have any questions, please let me know in the comments!


It would be hard to talk about bad photography in the original Cerebus series without discussing Gerhard's debut on the book, which happened to coincide with the with the worst photography since the issue we discussed last week. Did Preney Print and Litho have a turnover of their technical staff? Were they experimenting with different film, or new techniques to deal with the increasing levels of fine detail they were being asked to print on pulpy newsprint? We'll never know for sure, but if you have an original printing of issues 65 and 66 handy, the results speak for themselves.

The problem seemed to be the same as the issue we discussed last week— the exposure contrast point being set in a place that caused all kinds of unintentional "detail" to be printed on the negative along with the intentional. Sometimes this manifested itself as clogged Cerebus dot-tone (faint unerased pencil coming through the image). And other pages?

In issue 66, Gerhard's second issue on the book, there's a six page scene in which Posey and Cerebus discuss/rehearse the speech he's planning to give later on. The strange fog in the basement is suddenly cut by the "strange glowing white thingy."

Here's a peek at what that sequence looked like in print—

I actually asked Gerhard about this sequence in my interview with him way back in December 2010. Here's what he had to say:

Robinson: When we hit 305, is this some type of splatter on top of a Letratone?
Gerhard: Nope, again it’s the stipple tone. I would use two layers of it. The lighter gray is one layer, and I would put another layer of the stipple tone on top of it. If you look at the original pages it looks really good. If you look at this page reduced and printed on newsprint, it’s like “Ugh, that looks muddy. Don’t do that again!” That was the other thing learning to draw for reproduction. Most of the stuff I had done up till then was for framing, not reduced and reproduced. I would do the pages and I wouldn’t actually see how it turned out until the printed book came in. And I would look at it and go “Oh, that didn’t work; that did. Do that again; don’t do that again!” These issues were pretty much done without the knowledge of what it was going to look like in the final book.
When I started to work on this section of the Church & State I restoration, I wanted to know what the original art actually looked like. What would it have reproduced like under the best of circumstances, and could I bring the negatives to that state?

Fortunately, I was helped by a visit by the very generous James Guarnotta, who came down to San Diego from his native Los Angeles and brought with him his large Cerebus collection to be scanned for the restoration project.

And he happened to own one page from this sequence!

As you can see from the reproduction here, the original looks extremely different from the reproduction. Gerhard used what's been called a "mezzotint" tone (emulating the traditional mezzotint pattern produced by roughening the printing plate with a rocker, and then smoothing/shaving areas to lighten it). Gerhard used multiple layers of this light fleck tone to create the streams of rolling fog, visually adding to the density of the effect with each layer. He used up to three layers per page to "paint" with the fleck tone, in a really interesting way.

But as you can see from the image above, it reproduced horribly, the two-tiered areas of tone reproducing almost completely black and obscuring almost all of the hatching below with the exception of the thicker contour lines.

Here's a peek at the actual negative.

So, in the grayscale scan, we can already see lots more information than in the newsprint reproduction. I suspect there's more hiding in there. Let's make a fairly extreme levels adjustment, to bring the lightest tone more in line with how it looks on the original art scan.

I bring up the Levels command (Ctrl-L) and move the Mid-point Arrow/Gamma Control to the left, fairly extremely. And lo and behold, the detail begins to appear, and starts to make clear what happened.

Look at the new detail that's opened up on the darkest patches. This is the same problem as the previously discussed issue, with a poorly calibrated stat camera. The overlapping tone has created unintentional dark patches, muddying up the entire effect even before the ink has hit the page.

Although our extreme Gamma adjustment has shown us the detail hiding beneath the fog, we still haven't made it stick. Now instead of running our negative Action, we're going to do the following manually—

a. resize to the desired size (you can even play back just this line of your previous Action)
b. make a copy of the layer, as normal, and name it "Sharpened" (or play back this line of your previous Action)
c. make a Threshold adjustment layer ( (or play back this line of your previous Action)

Now we'll sharpen and adjust manually, in an attempt to keep all the good information while eliminating the bad.

Bring up the Unsharpen Mask filter, 500 percent and 1.3 px radius, and then play with the Threshold command until it's grabbing all of the obvious detail in the light fleck tone but is largely not affecting the noise.

Now I'm going to go in really close, and bring up the Levels command again. Another reduction in the Gamma (Mids) control, and a reduction in the Lights control as well, trying to knock out some more of the noise.

Now repeat those two steps again, until you've knocked out the majority of the noise.

And here's the full result, in direct comparison. The first image below is raw scan, bitmap-converted. The second is the adjusted and enhanced scan, bitmap-converted.

There are still a few weird artifacts in some of the overlapping fog areas, and there's plenty of new noise to clean up in the black areas and lettering, but it's remarkable how similar the image now is to the page of original art we have. 

Next week: The last post on negatives—I swear!

To download the negative scan we've been working from this week, click here.

Sean Michael Robinson is a writer, artist, and musician. See more at LivingtheLine.com.


Damian T. Lloyd, Esq. said...

I don't really have anything to say other than the usual, but I didn't want you to think you're writing into the void: The restored art looks amazing, the process is really interesting, thanks for doing it and letting us know how. Cool!

-- Damian

Sean R said...

Thanks Damian! I really appreciate the kind words and support.