I’ve been asked to make a post outlining my process for creating a puzzle page. It’s something I’ve wanted to do for a while, so here goes. I’m using my most recent puzzle page, Mazie’s bedroom, as it’s a fairly typical example of my work.
The writing generally comes first, and that’s just something I’m doing in my head all the time. Sometimes, though, a drawing will come first. I use the Notes app on my ipad/iphone, which backs up to the icloud, to jot down ideas as they come to me. I have tons of stuff in there, enough “idea debt” to keep me busy drawing for several lifetimes.
Anyway, once I have the general idea for a spread sorted out, the first thing I do is start making some rough sketches to spew out some of the ideas in my head. These are really just very quick notes for my own benefit. Here are some of the things I scribbled down for Mazie’s room, with my wife Tina’s help:
Next I need to decide on the overall layout, so I’ll do some quick thumbnail sketches for that. For some scenes it can take a while to fix the idea in my head to a specific scene on the page. For Mazie’s Room, though, I already pretty much knew how it was going to look, so I didn’t need to spend a lot of time on that. Here is my thumbnails for it:
All, super rough, as you can see. Just enough for me to know what I am going to do. I work very fast and very scribbly.
Now I am ready to plunge into the main drawing. I usually build the puzzle pages up in several layers, generally starting by working out the background/overall scene, especially if there is some architecture/perspective that I need to get right.
In this case, I had already done the drawing of Pickle Farm on page 1, and while doing that made some notes and sketches on it’s layout, so I could establish it as a “real” place, and know what it was like inside, and who’s room was where. I knew that I wanted Mazie’s room to be an ‘attic’ room (a bit like my son’s bedroom at home), and it was to be right above the big archway in the farmhouse courtyard. But I realized I needed to know a bit more, so I made some more detailed notes on the house and it’s layout. You can see those here on my Patreon page.
Anyway, here is my drawing of Mazie’s room:
You’ll note that I draw in red pencil first, then go over it with a regular B pencil. In the past, artists used a blue pencil, which could be easily dropped out photographically, but I’ve found that it’s actually very hard to make blue drop out on the computer, it always leaves grey lines behind. I discovered that red pencil works much better – just turing off the red channel and then converting the image to greyscale makes all the red disappear like magic. Hooray! That is my #1 artisiic discovery to date.
With the red channel turned off:
…and a tweak of the levels:
I’ll often draw the main characters separately, especially if they are more prominent in a scene like this one. It usually takes a few goes to get them right, but in this case they popped up more or less fully formed. I then scan them in along with the background, and make a composite image in the computer, which I then print out so I can add more elements and continue building the image. Depending on the scene, I might have several rounds of doing this, building things up in layers – particularly for complex search scenes. In this case it was fairly straightforward to add the additional items to the scene. The trickier bit was placing everything in exactly the right place with regard to the eyeballs. Again, the computer is very helpful in playing around with the layout and figuring all this stuff out.
Here is my sketch of the room elements:
And the final composited sketch.
Up until now I have been drawing everything at the size it will appear on the finished page (with a bit of additional scaling done on the computer), which is an 11×17” spread. I now print the finished sketch at 135%, which requires 3 tabloid-sized sheets of paper, which I tape together. This is as large as I can make it to fit on a sheet of 19”x24” Caslon Vidalon Vellum, which is what I use to do my inking on.
Here are some of the pens and tools I typically use. Most notably for inking are the Pitt Artist pens, and special Japanese brush pen which I buy in bulk from Asia on eBay. I love these, which I discovered through my pal Brian Kolm. Before that I used to use a Windsor and Newton sable brush, but these brush pens are so much easier to use, yet still give me the nice cartoony, analog brush line that I want. The inking is a pretty fun part of the process, and I usually put on my Pandora playlist and just get into the zone for however long it takes. The most important thing in inking is actually is to be sure that the pencils were just right when you start. If they are too loose and ambiguous, it leads to too many mistakes in the inking as I try to resolve things. By the same token, if the pencils are too tight and rigid, then the inks become a bit dead as there is no room from expression. To some extent I still need to be able to invent and draw as I ink, but not too much! This is because I’m inking my own work. I think it’s a different story if you are doing pencils for someone else to ink, or if you are inking someone else’s pencils, but since it’s all me, I just know the balance I need to keep in order to work as efficiently as possible.
Here is a shot of my inks, as I stick them together, as the original is so bit it takes 3 passes with the scanner:
I then tweak these further on the computer, cleaning up errors and making other little fixes. As you can see in this initial drawing, I had kind of cheated because I wanted the view from the window to show the trees and the forest, as a visual cue for the next scene. I had figured it was Ok to use a bit of artistic license here for story purposes, but in the end I decided that it was more important to stay true the the world I was creating, so I made sure the view was what you could actually see through the window – i.e. part of the farmhouse wall. But sill, there’s always a bit license taken. The main thing is, though, that the hidden joke of this spread is that Wilma is actually floating up to the window, and Mazie will have to jump down using (SPOILER ALERT) an umbrella and a pillow.
Cleaned up inks:
Next is where I use some secret sauce to process the image, and this bit is more for you technical nerds out there. I have devised a script to separate the image out into layers. The usual way to process comic artwork is to convert the image to bitmap, and this is what all the how-to books will tell you to do, but I hate those jaggy bitmap edges at any size. Sure, at the correct size, you can’t see them, but what if you wanted to enlarge them. Early in my career I worked in a print shop, managing the “Desktop Publishing” desk. I learned a lot about print production techniques and was an early Mac user. As a result I know that you can actually print things at a low lower resolution than the textbooks tell you, but those jaggy edges would be a dead giveaway if you were enlarging an image too much.
So I developed a way to keep them a little anti-aliased, which means that I don’t need to work at a super-high resolution. Well, OK I do work at 405 dpi (because of 300 dpi x 135%), but you can actually get away with only 150 dpi before most people would notice, so I say you could print my images at nearly 3 times their original size before anyone would notice. OK, well, I did say that this bit was for only technical nerds!
My inks layer:
Traditionally bitmapped inks:
I could explain in more detail if anyone is interested, but basically I tweak the levels so they are almost fully bitmapped, but not quite. Then, by selecting the channels I copy just the anti-aliased black inks to their own, transparent layer, that looks like this:
But I also create a duplicate inks layer which is fully bitmapped. This is the layer I will color on, and I use a plug-in called BPelt Multi-fill to random fill the bitmapped area with colors. All that is part of the same single script, so with one click I end up with this:
You can read more about the BPelt plug-ins here:
It’s a very useful start to the color “flatting” process. Basically it helps you more quickly see where there are “holes” in the artwork that need to be closed before filling with flat colors. Color flatting is basically where the drudgery happens. The idea is simply to define flat areas of solid color. One day I’d love to employ and assistant to do this for me, but right now I just put on a good audio book and get to work, and it’s a pleasant enough way to pass the time. It’s no wonder those adult coloring books are getting so popular!
It’s not 100% necessary for colors to be accurate at this stage, but I try to get them more-or-less right, as it saves times later. I have a specific palette of color swatches that I have developed over the years. This is largely based on the colors found in Crayola crayons and Copic markers. Bright and cheerful, and giving my colors a signature look.
I have a Wacom Cintiq, which is basically a pressure-sensitive screen I can draw right onto. This is an invaluable tool. Previously I had just a regular Wacom tablet, which is better than just using mouse, as a mouse is almost impossibly tedious to use. I know that a lot of the kids coming up these days don’t even bother with pencils and inks and just draw straight into the computer using their Cintiqs. I don’t know, I feel like I spend enough time staring at screens as it is, so I prefer to keep to the old fashioned ways for the pencilling and inking. Plus lines drawn on a computer just still feel dead to me, but they do look a lot cleaner, I must admit, and I do do it sometimes.
The trick while doing the color flats is to turn anti-aliasing off, i.e. hard, jaggy edges formed by working with just the pencil tool and paint bucket, as this makes the next stage possible.
Once the color flats are done I run another script on the image, using another bPelt Plug-in called Flatten Pro. This removes all the black lines from the image, and expanding the color fills to replace them, resulting in something that looks like this:
I always find the way the colors look at this stage very appealing, like nice colorful cutouts. I’m always wondering about a way to eliminate my ink lines entirely, but I haven’t figured out exactly how I could do that and still maintain the cartoony look I am after. Anyway, at this point I hopefully get a good feeling about how my overall color scheme is working. So far I’ve kept my transparent anti-aliased lineart turned off on another layer, now I turn it back on again:
The second script I ran, also duplicated the color flats layer and added some layers for highlights and shadows. I keep one flat color layer under the other, using the lower one for area selection and the upper one to work on. I keep the ink lines on their transparent layer separate and on top.
The next part is the most fun. The hard work of sketching and thinking is done, as is the drudgery of color flatting, now I just let go and see what happens. I suppose the main trick is not to overwork the image. I use the magic wand tool to select areas and paint in colors with the various brushes I have developed over the years for the process. I’ll use the airbrush a lot for gradients, the trick with that being to select an area with the wand, and then deselect parts with the selection tool to get the right combination of hard and soft edges. That is more-or-less the way digital coloring was developed in comic books, and it works pretty well. I’ll also use a harder edged “Cartoon Brush”, with a variable line-weight, and a “Modelled Color” brush. All the brushes are pressure sensitive, so this is where the Cintiq really comes into it’s own. I also have a few favorite textured brushes that I have developed over the years. It’s worth spending some time figuring out some brushes that you like to use. I guess some artists have tons of them, myself, I like to keep things pretty minimal.
The fact that the linked line-art is still on its own, transparent layer means I can now very easily add color to the outlines, instead of just leaving them black. Mostly I like to keep them nice and black, but a few touches here and there helps to de-emphasize certain lines.
I’ll work away at the image for a while, using the highlight and shadow layers to broadly define things, then work on a more specific areas. Some artists like to keep lots of layers, but I find it’s more efficient just to merge layers pretty freely, and am generally not afraid to commit to things. As long as I keep my original “flat color’ layer intact, it’s easy enough to select and re-do areas. It’s possible that I might even go as far as merging the line-art layer to the color layer, but usually after I am pretty certain that I’m committed to everything, as there’s no turning back from this. Still, it can be especially useful once it gets into the final editing stage, particularly as its now so easy to work directly on screen with the Cintiq.
At some point I guess I should make a coloring video, to show the process in more deatl.
Anyway, basically that’s it! Here’s the final result:
Thanks for reading! if you would like to see more, please join the fun by visiting my Patreon page: