Wednesday, 3 May 2017

Paper to Pixel to Paper Again: Part 15

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 15
Production Negative Sharpening

Greetings!

This is the fifteenth 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!

***

We left off last week discussing production negatives and the wealth of information that can be hiding within them, ready to come out. Now we're going to write our upscaling and sharpening script for them.

Unlike original artwork, which I recommend be scanned with consistent settings and batch adjusted/scripted, I've found that the best results for working with negatives involve adjusting them each individually and then writing a script that "spits out" the finished page on the other end. The reason for this is the vast amount of variation that can be present from issue to issue or even page to page in the photography of the books. Though issues are mostly pretty consistent—i.e. if something is funky or adjusted in a particular way on Cerebus Issue 230 pg 1, then Cerebus Issue 230 pg 20 is probably pretty similar—there are pretty major exceptions, especially in portions of the book where the photography was particularly, let's say, adventurous. If you're batch processing an entire issue, you're missing out on seeing these oddities and thus missing your best window of correction.

Okay, here's our adjusted/Gamma-corrected example from last week. Now we're going to write a script that will scale it to our destination size and resolution and then make a sharpened version of it to retain all the detail. (Contrary to popular belief, production negatives can be scaled to virtually any size from your original scans, if you captured sufficient resolution in your initial scan! One of the many advantages to scanning to grayscale rather than the HORROR that is "copydot scans". Ugh. Don't get me started. Want to know why every single edition of Osamu Tezuka comics has moire in the tone? "Copydot"! If someone starts talking about "copydot", or talks about scanning your negatives direct to 1-bit bitmaps, RUN!)


Anyway.

Our raw scan was scanned at-size and at 1200 ppi grayscale 24-bit. After carefully measuring the image area, I've determined that the overall size needs to be enlarged by 106 percent, and then the resolution should be changed to our output resolution of 2400 ppi. NO SHARPENING UNTIL YOU'VE UPSCALED, ON ANY IMAGES WITH DOT TONE. DON'T SAY I DIDN'T WARN YOU.

I click on my Actions pane and click the Create New Action button on the bottom of the pane.

Then I'll do the following —

♠ "Flatten" image (in case I have to rotate any of my pages later on—this will keep my script working)
♠ Convert to Grayscale (if I'm accidentally in color?)
♠ Make the Color mode 8-bit (which will seriously reduce your file size and speed up your Actions. we only needed the extra depth for our adjustment stage)

And then, hit Ctrl-Alt-I to go to the Image Size dialogue.

Enter your size increase of 106 percent into the Percent box, and then follow that by the resolution change to 2400. You'll notice that the overall percentage at top will change when you enter the second value. (Photoshop is multiplying the first percentage you entered by the second percentage change implied in the resolution). Then change your Resample mode to Preserve Details and bring the Reduce Noise slider all the way to Zed, and away we go! Hit OK.


(Check out the top percentage box. Photoshop is sometime bad at maths...)

Now, while the Action is still running, do the following —

♠ Right-click the layer with the image and click "Duplicate Layer"
♠ Name your new duplicated layer "Sharpened"
♠ Make a new layer with the "New Layer" button at the bottom of the Layers pane
♠ Name this layer "Cleanup"
♠ Go to Adjustments-> New Adjustment Layer-> Threshold to make a Threshold layer that will preview what our image will look like once it's a bitmap
♠ Lastly, toggle this layer off so we can work without it for a while.

Now, we're going to sharpen our "Sharpened" negative layer. Why do we do this on a new layer? For a few reasons. One, we can easily check what we've done to the image overall, by clicking on or off the layer and seeing the changes. Two, so if anything goes wrong in the future, we can easily just duplicate this layer again and run a new script (or a portion of our old script) and retain all of our other work. In other words, it gives us flexibility.

Our actual sharpening methods are going to be almost identical to how we sharpened our original artwork (it's worth another read!) We will sharpen a round, with a radius close to 1 pixel, then knock out a tiny bit of our white/paper color, then sharpen one more time with the same radius.

I'm going to zoom in on one of the tiny-toned Cerebi and take a look up close.

I end up using a very strong dose of sharpening, with a very low Threshold (around 10) and a high Amount (around 250), then bringing up my Levels and backing down the white point just a bit, then running the Unsharp Mask filter one more time at the same settings. Here's the result. First, below, is the unsharpened, bitmap-flattened image.


And below is the sharpened image. In addition to the various tones now being sharper and more their correct values, a host of details have appeared, including the "motion blur" lines on Cerebus's face that haven't been present in any printing prior. Although they were captured in the photography, they weren't sufficiently contrast-adjusted to make it to the plate.



We can take this further by scanning the rest of the page with the Threshold adjustment off, and looking for very light/very gray/very fine lines or other information, and circling them with the Lasso selection tool, and then blasting these areas selectively with their own round of sharpening.

But first, turn off your Action! We're going to save it to use on future negatives.

As always, questions and/or gratuitous praise? Hit me in the comments!

Next week: Bad negatives, and the women who love them!

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

No comments: