Wednesday, 19 July 2017

Paper to Pixel to Paper Again: Part 22

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

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

Part 22
Cleaning Up Line-up From Print Sources


This is the twenty-second 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!


In the last installment I suggested that there's a way to uniformly reduce the thickness of every single aspect of a line art scan after it's been cleaned, regardless of origin. This is the technique I used as a final step for the majority of pages in the new Cerebus Volume One to remove dot gain present from previous printings, and that I've used in the most recent issues of Cerebus in Hell? to ensure that the already-teeny-tiny-lined wood engravings survived being reduced in size so much. 

So, you wanna uniformly reduced the size of your artwork to reduce (or anticipate) dot gain? How do you do it?

Let's go back to the first page of Cerebus ever, restored from a print copy in a previous installment.

This is how the page appears after the cleanup.

 Although it looks excellent, and noticeably better than previous printings of the phonebooks, the entirety of the image has expanded from the negative because of the dot gain in the previous printing. If we allow this to go unchanged, it's very likely that details appearing here will fill in even more upon printing again.

Which leaves us with the question—how much expansion was there in the original printing? Well, we can make a pretty good estimate by looking at the Cerebus figure.

Although Dave (later, Gerhard) occasionally used a darker screentone when Cerebus was in the dark/under an object/ inside an enclosed space/and a lighter tone when he was drunk or sick, for the most part they used a straight 30 percent tone for his fur "color". So, keeping in mind any other suggestions on the page that the tone might have been different, I can use this tone to calibrate the amount of gain reduction I'm going to do.

Okay, let's get to the actual technique.

First, we're going to make a flattened version of our page, flattened and Threshold-converted! So turn on your Threshold Adjustment layer, and make every layer that you want to be visible in the final image visible.

Then hold Alt-Ctrl-Shift and hit E. This makes a copy of all of your work as a new flattened layer on top everything else. Lastly, drag this newly-created layer beneath the Threshold-Adjustment layer, for easy comparison afterwards.

(Normally we wouldn't Threshold-convert line art until we're totally done with it, but this is the most efficient way to accomplish what we want. Making it a new layer is an additional safeguard in case you want to return to your pre-adjusted work in the future!)

Now it's time for the magic trick. Go to Filter-> Blur-> Gaussian Blur, and use a blur with a radius of just under one pixel. (.9 px should work perfectly).

If you have the Threshold Layer turned on, you probably won't notice much of a difference except for the finest details shrinking or disappearing a bit. Now bring up the Levels control (Ctrl-L).

Now grab the Mid-control (the gray middle arrow) and bring it to the left. The further to the left you move the arrow, the more you'll shrink your content. Don't worry if smaller elements continue to disappear, just bring the tone (or whatever area you're using to calibrate the shrinkage) to go to the size you want. (At the end of this series, I'll upload a "tool kit" of Photoshop scripts that includes a tool to analyze tone density, so if you've already downloaded that, future person, use that :)  )

Lastly, we have to restore the details that we lost. If you've read and understood this series thus far, you probably know by now what the answer is—Unsharp Mask.

Bring the Threshold all the way down—we want this to affect the entirety of the image. The radius, as usual, should be just a little higher than 1 px, and the amount just under 200 percent. If you make the amount too high, you'll make the entire image unnaturally spiky—so watch this.

And if you've done all of this, assuming you're working in the required resolution space (2400 ppi), you should have a perfectly-reduced version of your line art!

(Tangential parenthetical—notice how the dots are the most round and smooth in appearance in the middle, Gaussian-blur applied example? Well, as you might guess, some very clever use of Gaussian Blur and sharpening can do wonders for damaged or digitally mangled tone. But there's no real formula, only adjustments that can be made on a case-by-case basis, and I've already dragged this series on far longer than its usefulness to most readers...)

As I mentioned above, the monthly schedule of Cerebus in Hell? has enabled me to take it to another level, which is a good thing, seeing how much the Dore images are shrunk on any given page, and seeing how dense with dark tones they are, and thus more prone to visual dot gain.

The problem is exacerbated on the Bible plates, where I'm working from scans of materials one generation removed from the original printings. So they need special treatment to ensure they don't close up upon printing.

And voila! The technique works great.

A "Solomon" panel from Batvark, before, then after:

One last thought. It took me less than an hour to type this installment up and put together all of these examples, but all-told hundreds of hours of work on the overall Cerebus restoration project before I figured this technique out and perfected its application. Of all the things I've figured out in the course of this project, this and a sophisticated use of sharpening are the two things that I think are most significant to future line art projects, and the least understood by other people. So, if you have any need for these techniques, read this installment, play with these techniques, until you understand them inside and out. And as always, if you have a better way to accomplish the same ends, let me know in the comments!

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


Jimmy Gownley said...

Brilliant as always!

Jeff Seiler said...

One of these days, you really have to explain this stuff to me in a way that I can understand it, me being digitally illiterate.

Nevertheless, like a bug to a zapper, I am fascinated!

A Moment Of Cerebus said...

"digitally illiterate"

I don't understand why any one would pride themselves on being digitally illiterate. I'm not saying we all need to be in Sean's league, but Dave and Jeff could do with a few basic lessons. PCs/Macs are here. They're not going away. So do yourself a favour and book an evening class or two. I guarantee you'll be more productive and effective in your work (whatever that may be)... less hassle for everyone else... and less irritating to listen to.

Sean R said...

Aww, thanks, Jimmy!

Hey Jeff--I think Tim's point is an excellent one. Although they can take over your life if you let them, computers are tools like any other, just incredibly multi-faceted tools. I could imagine you having a great time in a community college course focused on basic Photoshop or word processing (and would probably have some significant practical benefits too :) )

Part of the problem of writing this is there's no one audience. If I labored to create a bunch of real-work analogies for all of the things I'm doing here, I'd end up with people who didn't need all of the explanations rolling their eyes the entire time.... it's a difficult balance.

Jeff Seiler said...

Well, Sean (and, especially, Tim, who seems to seriously dislike me), I'm not complaining. I own who I am, as does Dave (he owns who he is *and who I am ;)). If I or he choose to be digitally illiterate, then that is my and his choice. Why should you choose to be irritated by that, Tim?

But, thanks for the advice, anyway.

A Moment Of Cerebus said...


I don't "seriously dislike" you. I don't even know you.

Help me understand this. Why would you choose to be "digitally illiterate"? Let's use the example of your proof-reading Cerebus for Dave. You could be way more effective/productive if you got to grips with MS Word (or similar word processing program). I have no interest how computers work in the nut-and-bolts sense, but if I have a job to do I want to be the most effective I can be at it, and that means mastering all the tools at my disposal. Taking pride in being "digitally illiterate" just seems like a pathetic affectation... IMHO.

As far as I know, Dave has mastered the use of a photocopier, fax machine and video camera. Why? Because he has a job to do and he wants to get it done efficiently. Why are computers any different? If you were born 100 years ago, would you refuse to use the telephone whenever it got invented because you want to be seen as "telephonically illiterate"?

Your posturing is irritating because I see you and Dave getting befuddled by simple IT concepts and it is harming the Cerebus restoration project.


Jeff Seiler said...

Hi, Tim! Glad to hear you say that you don't dislike me, actively or otherwise.

I don't take "pride" in being digitally illiterate--it just so happens that there was nary a computer in a classroom I attended until I was in graduate school. Even then, it was in the basement, and we plugged punch cards into it.

Nevertheless, I am typing this out on my handheld computer, which is multiple times more powerful and user-friendly than was the mainframe computer that filled that entire basement at OSU.

I know MS Word and have utilized it countless times. I prefer to do my proofreading by implementing a red pen upon the printed page, first. After that comes the computer.

Who are you to say that my way is worse than any other?

A Moment Of Cerebus said...

Hi Jeff,

I wouldn't dream of telling you how to proofread. You're the expert... but that wasn't my original point.