Page One, Part 1: Technical Difficulties

First color panel...ever!

So lets get the technical goobledygook out of the way first.

 

Remember when I found a method of refining the edges on the magic wand that I said cured pixelization? Well the cure had a curse to it too. It leaves this area around my lines that I can't select. Why is that bad? Because it means any color I lay down with a paint bucket won't reach the lines and leaves sometimes noticeable white areas around every line. So I can't use the paint bucket or magic wand to select and fill areas with color, which is the quickest, easiest way to color something.



So how in the name of Lynn Varley do I color the thing? Well I have to do what real comic colorists do. You see, when coloring for the web, anything goes. What you see is what you get on screen (though the colors will vary slightly from screen to screen, but that's a frustration for another day).

Print is a bit more...complicated. You've got different color plates for CMYK you have to color on and you have to do it a certain way because the plates can slip. I don't really understand all of it, but since I'm doing this for the web, I don't have to deal with it.

But basically all this forces those coloring comics for print to color using multiple layers that you lay the inks over, and that's the idea that came in handy. If you've ever seen comics colored by hand, you've got some idea what I'm talking about, with the inks on one transparency that's laid over other transparencies with the different colors on them.



Here's the work in progress.

So that's what I had to do: delete to a transparent background on the inks and put color layers underneath it. Now there's definitely advantages to that. Instead of having to worry about coloring around a person to do the background color, you just color over him and color his face, clothes, ect on the layer (or layers) on top of that.



The downside to this is that I had to select larger areas manually by drawing around it with the lasso tool and I had  to color the finer details freehand with the paintbrush tool. I'm doing this with a mouse. It...takes...forever.



However, I will say I actually really enjoyed doing it, and wouldn't mind being a colorist based on my experience so far. But it does take up a lot of time, so hopefully I can find a way to refine how I'm refining the magic wand so I can use some of the quicker methods too.

I'm also finding the method of digitally cutting the word balloons and text boxes out of the old printed comic and laying them in the new version is also taking forever. It's just taking too long to erase the art around the balloons, and the zillion layers in Photoshop I'm having to deal with as I drag them in isn't helping either. So I'm probably going to have a friend transcribe the script and I'll have to re-learn Illustrator so I can re-letter it. It may slow me down at first, but it'll be a lot faster in the long run, I think.

Next: Page One, Part 2:  Now that's a damn fine peacock!

 



The Cover, Part 2: Me vs Pixelization

The original cover, as printed, back in 2000.

So the original version of the first issue, of which only a few dozen copies were produced at a local print shop, didn't print well. The line art ended up looking squiggly (like something out of Dr. Katz), probably from bad compression or low resolution.

This was back in the year 2000, when it was a challenge working with and storing files big enough for print and I didn't know what I was doing. Back then I had to fit all my art class projects on a zip disk that held 100 megabytes. Nowadays I have single projects that are 100 megabytes a piece. The technology has made quantum leaps, so big files are no problem now.

So when I redid the cover, like the rest of the comic, it needed to be rescanned from the original inks. I scanned in the printed comic too to pull the lettering from, since the lettering turned out okay in the comic. So all I had to do was scan in the cover, pull the word balloons and text, slap them together, add color and call it a day. Simple, no?

Well, I went through the entire process and produced a final PDF. I then double checked Comixology's submission guidelines to see if I got the size exactly right. I ended up stumbling on a bit on there that pages should not look pixelated if you zoom in to 300 percent. As I was working on it, I only viewed the cover in the size that matched how big Comixology's comics were on my computer. It never occurred to me to zoom in THAT much.

So I did...and the text was a bit pixelated. Well, ok, I'll just redo it and type the text, no big deal. But then I noticed the line art was slightly pixelated. Everything was scanned in at 300 dpi. That's print resolution. That's the resolution Comixology used in its examples. And it's...pixelated.

So at this, point I fight the urge to assume the fetal position, and again look up how to scan comic art online. I found most people recommended scanning it in at least 600 dpi. One guy suggested 1200 dpi, but my scanner crashed when I did that, so I decided 600 would be good enough.

It looked great, but then I tested it after I did background removal. You see, the scanner will get the texture of the paper, any left over pencils, etc. along with the inks so you got to get rid of that junk. One way is to scan it as text or line art, which ends up with a very poor image on my scanner. You can also try turning into a bitmap or using threshold in Photoshop but that also seemed to down the quality.

So I've been trying a technique where you select the black ink with the magic wand (by turning contiguous off, so it selects all the black, and anti-aliased off, so it gets the tightest line selection possible). Then you color those selected lines with default black (because they scan in faded and need to be darkened  up). Then I "select inverse" and  hit delete to get ride of all the background junk...and boom! All I got left is black lines and white background. This technique has created some nice looking art I've been using on my social media sites.

So I tried this technique on the cover at 600 dpi and then tested the image, busting it back down to 300 dpi as a pdf. It was better, but I could still see the jagged pixels at 300 percent. So I fumbled around for a couple hours trying different things when I found an option on the wand that lets me "refine edge." I tried it and...voila, smooth lines at 300 percent!

So at long last I think I have an effective way to scan the art work in that'll meet Comixology's criteria (and also give me high quality line art if I ever get it printed again). Hopefully, with a bit of fine tuning, this will work for the rest of the pages.