Josh asked that I go back through my files and share what I could of my process. I'll do so here as best I can, though the ugliest failures have been long swept away by the power of Command-Option-Z. When I begin working on a coloring project, I start with a color reference. In this case, I was starting off with a night-time city scene, but I wanted something that looked intense and noir at the same time. I began with this image.
As you can see, it's got a dynamic range of colors -- warm to cool -- and looks both action-packed and mysterious. From it, I sketched out some colors for the first cover using the brush tool, without flats, under the lines.
The colors as posted here are not exactly right because Wordpress shifts them, but you should get the idea. For the hot areas, I used the horse's head for a color source (coloring quick-tip: hold down Option with any painting tool over the color you'd like to use and left-click the mouse to select that color as a new foreground color). I used the various greens for areas of shadow and the reds from the emergency lights to make accents (superpowers and lamps).
I decided it wasn't right, though at this point the reasoning was purely instinctual. Back to the drawing board. I realized that I didn't know what was going on in the piece enough for me to start coloring willy-nilly. I decided to do the absurd.
I called Josh.
The question was this: "Josh, what is this cover supposed to convey emotionally?"
Josh told me that the cover was supposed to communicate liberation and freedom and hope and (so on and so on and so on...if you know Josh, you know he is garrulous when it comes to this project). I internalized his concepts as best I could and rethought the image. The first thing that I realized was that it shouldn't be a night scene. Next, I began thinking about the background -- the environment -- in terms of his concepts. I came up with the following image as color reference from those musings.
The Google Images search terms (I think) were something like "city awesome cylinder building" or something. Whatever the case, it's pretty damned awesome. In particular, I was attracted to how desaturated the color scheme was. When I was first brought on, Kav and Josh told me that they wanted a desaturated, dusty look to the coloring, and this seemed to fit nicely with their desires. I started working and ended up with this. Before I began, though, I removed the night sky background using the polygonal lasso tool.
First off, notice that the primaries (Kinetic and the building on the left) are still flatted in their original hues. This is because I had not totally given up on the first color scheme yet. The advantage of new computers and their power is that you can build multiple layers that can be turned on and off as you like. So cool.
Anyway, if you look closely, you'll see that I modeled the main building off of the colors on the cylinder building in the source image. I built a gradient that went cornflower-->rose --> very light gray --> darker gray --> sky blue (these all cherry-picked from the source image). This seemed to work well, especially when I started blocking in the colors on the foreground architecture at the bottom of the page. In the background I used the source colors, but opted not to copy and paste clouds into the image (I once saw this done in Jae Lee's Namor series back in the 90s to distracting effect and have not been able to bring myself to do it since*). Instead, I used a great ink-splat brush-set that I downloaded a while ago (I use it on literally everything) and some liquify effects to give it that " liberation and freedom and hope" thing I was shooting for. It was here that things started feeling wrong, but I couldn't put my finger on it yet.
Next I started working on the windows for the main building. I used, again, the source building as reference (royal blue --> olive --> light blue --> dark olive green --> light steel blue). It was here that I noticed some problems of perspective in the upper part of the building and thought that these were, perhaps, the source of my unease.
I set about to correct them.
Obviously, these cosmetic fixes (though nice) did not solve the problem (if your curious, though, the vanishing point is technically at Kinetic's left palm).
No, the REAL problem was in the background. If you look at the source image, you'll notice that the camera sits at basically ground level and is fish-eyeing nearly 50% of the forward scene horizontally. OUR image, however, is shot from worm's eye view. It turns out (I learned) that you cannot simply transpose a sky pattern onto a different viewing angle. The reason has to do with science.
In the source image, you'll notice that there are these very attractive rose tints on the bottoms fo the buildings. These are caused by the refraction of the sun's ray's as they travel to our eyes through atmosphere that is much broader at the horizon than if the sun is up in the sky. This is why sunsets and sunrises are so beautiful and why noon is so patently blue and boring. In order to get this color scheme (blue-white-gold), I needed to put my sun up behind Kinetic. But to do this would put the sun at near zenith -- where there would be no refraction and no yellow and rose skyline. It made the picture look totally false and I had to scrap the source entirely.
Sad, but necessary.
I next looked for source images of cities and buildings from a worm's eye view. I found the following one that stood out as dynamic.
This image stood out as supremely right to me (and still does). However, I now faced a problem of artistic will: could I really devote myself to coloring the majority of a building (and everything else) in shadow like this? Would it work?
I remembered my Rembrandt and went for it (in retrospect, it was this contrast idea that attracted me to the first image with the horse, probably -- even a similar hue for the hot spot!).
The key colors to note, here are these:
1) sky blues: there is a simple, uncluttered, gradient that move from a light sky blue to a slightly more saturated sky blue
2) the top of this grey building is colored a pale coral in the rose sunlight
I set to applying this on the page as follows.
Then I ignored the source image -- and my own mantras -- and tried to create interest by dropping a sun and clouded sky in behind my hero...to no avail because it a) cluttered the sky and b) would have forced me to make Kinetic a silhouette...which would be wrong here.
[I also worked on dropping a half-daylight moon into the sky hear without knowing that Josh had a no-moons mandate on the cover!]
Instead, I went back to that source image.
The way I achieved this effect was to paste the image into place over top of it. I used the Distort function under the Move tool (V) to make the original source building fit exactly over the top of the art. This, though, looked a mess. I needed the colors more saturated and blended -- like I dreamed them up with a Photoshop paintbrush on the fly. To achieve this, I used the Paint Daubs filter set at its highest levels.
I colored the windows using a gradient. To figure out the colors for it, I used a weird approach. I painted the whole window section the gunmetal gray that is at the center (cherrypicked from the source image). Then I clicked on the left-most window on the source image, then clicked on its middle image and from that noted a) the shift the selector made on the color field horizontally and b) which way the slider shifted vertically. Then, I went to the foreground color (gunmetal gray) and tried to shift the slider and selector in a similar way for both the right an left sides of the windows. Hope this helps.
When stumped, I start messing around with my gimmicks because I'll need to use them eventually and because it keeps me from falling into artist's block. In this case, I dropped in a concrete texture over the building and created a color-hold channel (shameless, obnoxious, and something I ALWAYS do. I invite you to too). While messing with the color holds, I realized that I was thinking about the foreground all wrong. It wasn't just some concrete and limestone. It was an architectural FEATURE. I set to work getting it well-rounded.
2) The brown lines were not done with color holds but on their own layer using the pen tool, stroked with a brush (this was done to make them glassy-smooth)
3) highlights and shadows on the gilding were done using Multiply and Screen settings under the brush tool set with the background blue as its foreground color.
At this point I'm starting to feel like I'm closing in on an ending to the piece. In turn, I start fishing through my folders for the Rapid City logo...and that's where the trouble begins:
Josh and Kav have agreed to put the logo at the bottom of the image!
This is unacceptable to me for a variety of reasons, but the main one is this: I just got done c0loring the bottom of the page for the last two days. No way I'm covering it with words. Those need to go up-top. The problem is, the page cuts off just above Kinetic's head.
I'll need to extend the piece vertically.
To do this, I will need to do some trial and error work. Key combos to know are Command-Option-I (Change Image Size) and Command-Option-C (Change Canvas Size) -- along with the classic Command-Option-Z (Edit-Undo) -- to get this reframed.
At this point, I realized that Kav had drawn the artwork to a non-standard cover size, but instead to the dimensions of a comic's interior art. This is generally not a big deal and can be fixed by simply enlarging the artwork to the appropriate dimensions. However, in this case we have no free real estate: Kinetic's fingers stretch all the way to the edge of the live space already...and the first no-no of comics is "never cut off body parts!" Here's what I did.
Kav's artwork was drawn at 2016 x 3020 px (that's 6.75" x 10" at 300 dpi to you and me). The safe zone (called "live space") on a comic page that won't get cut off during printing is only 1875 x 2925 px (6.25" x 9.75"). This isn't really a big deal yet, though, because printers use precision presses any more and you can push your artwork all the way to the trim line (where they plan to cut the page) without much worry, which is at 2025 x 3075 px (6.75" x 10.25"). Kav's art -- as is -- fits here nicely, but with no horizontal wiggle. On the vertical, which is where we need the space for the logo, he only has a quarter inch -- and I needed much more than that to fit in the Rapid City logo comfortably.
I would need to do actual drawing to make it work.
The stages of the process are illustrated in this animation. First, I shrunk the image down to a size that would allow for the logo to fit. I was willing to overlap the i and t in City a bit with Kinetic's hand, but no more. Nevertheless, I would need to build onto the original image.
I first built out the needed right-hand side of the cylinder building. This was seemingly not so difficult since I already had the vanishing point in place when I rebuilt the windows on top. However, that sconce/fixture thingy (just to the right of Kinetic's left calf) would need to be repeated again for the image to work...and this was trickier.
I used the Marquee tool to copy and paste the sconce onto a new layer by itself and moved it into the appropriate position. Then is used the Warp function in the Move tool in order to model it as a rotated 3d object. It wasn't quite right, though, so I forced the last bits of it using the Smudge tool set at 100% and full hardness. Once it was arranged the way I wanted, I crisped it up with the Levels applet. I finished up the surrounding areas manually using the brush.
Likewise, extending the bottom left of the building's molding was simply a matter of drawing out the lines using a stroked pen tool. However, to get the marble pattern that Kav put down to carry out, I simply copied and pasted the pattern and cleaned up the joint.
The building at the top left was much more complex to extend and this was largely a product of its lack of vanishing point. When sketching, this is not an issue, but cloning patterns requires precise angles and measurements. In order to make this work, I copied and pasted the building into an entirely new document. This .psd file I made very large so as to accommodate a distant vanishing point. Once I found it, I retroactively redrew all the lines of the building so that they would run to the new edge of the page. I used a pen tool for this, which is the smartest way, I think. Once I found the vanishing point, I put one path anchor there and then simply dragged the opposite anchor to whatever spot lined the path up to the ridge on the building. Once these guidelines were appropriately in place, I simply cloned Kav's textures outward.
The top molding was the easiest to clone: the lasso tool for selection combined with a bit of a nudge using the distort function of the Move tool was all that was neaded.
For the boxy molding further in, I worked from scratch to get the spreading caused by perspective correct. For the measurements of this, I used multiple layers and the shape-line tool. It's a difficult technique to explain and more involved than is really necessary here.
With all the organizational ducks in a row, the rest of the job was really just fun coloring. For the weathervane, I used some reference images of oxidized copper to work from. It's shaded using a hard brush and multiply, normal, and screen modes. If you look closely, you'll see that the orange oxidization blotches are really those same splatter brushes that I used for the sky oh-so-long ago. Shameless. I ended up moving the left sphere thing downward a bit.
Another cool and easy tidbit was the drop-shadow from the weathervane. All this required was selecting the flat for that fixture with the wand tool (anti-aliasing was on) and then moving the dancing ants a bit down and to the right with the arrow keys. I used a 10% opacity brush set to black that was larger than the area once or twice to fill it in.
To create the depth on the windows was pretty easy, but required perhaps more steps than is totally obvious. I simply made a duplicate of the linework layer and dropped it below the linework layer. I then used the wand tool to select the linework (I do this a special way: click the white, then invert the selection -- this gives you thicker lines than if you only click on the lines themselves). I filled the selection with the blue in the sky on the right of the building, then used the Move tool and the arrow keys to nudge the layer down and to the right. This created the highlights for the right side of the building. To do the ones on the left, I duplicated the layer I just moved and then changed the blue of the linework to the blue of the sky on the left of the page. Again, I used the move tool to nudge these lines to the left of the windows. Then, I cleared out all the unwanted linework using the eraser tool.
Phew! Now all the background is done. I almost always begin with my background before doing the main character. It's not so much because it creates the color environment in which that person will exist (which it does, I guess), but that I know I have bad habits from the olden days. If I begin with the super-hero, I'll spend a lot of time on him and then skimp on the background stuff. But enough of that, I was lined up to color Kinetic.
Before I began, I grouped all of the layers I had worked on so far into folders to avoid clutter. Then, I used the lasso tool to separate Kinetic entirely onto a new layer of his own. This is because I needed his hand to be in front of the logo, otherwise I would have worked in exactly the same layers. Once I had his lines on a new layer up-top, I made an individual flats layer for him as well that went beneath that.
There isn't much to say about how I colored Kinetic himself. It's pretty standard approach stuff: I use the polygonal lasso tool to select areas, then use brushes of varying hardness and opacity to color (I have no Wacom...yet!). I generally color base/local colors (gray, yellow, brown, and peach in this case), then use the multiply and screen modes to shade or highlight. I did all of the shaping in this whole piece using the two shades of blue in the sky, the olive green color at the center of the cylinder building, and the coral color at the top of the same.
Finally, the fun bit was left: Kinetic's power effect. I took a lot of liberty with his power effect, trying to make it as eye-catching and flashy as possible. I also intended for it to communicate the movement of Kinetic downward. I began by isolating Kav's original effects. These, of course, were long obliterated when I went from night to day. To retrieve them, I went back to a previous version (at this point, I had five saved .psd files (which is why I am able to recreate so much of this process for you). The only tricky part about this was that the night was turned to day before I rescaled the page. In turn, I had to resize it on the fly with the scale handles. Not a huge ordeal though.
Once I had the layer in place, I went through with the wand tool and selected each little sparkle thing that Kav had drawn. Then, I inverted the selection, pressed backspace, and had a layer with only the sparkles on it. I double clicked the layer to bring up the effects interface and put on a gradient overlay (yellow to orange radial) and put an outside glow in red.
To get the streamers to work, I needed them to be on separate layers so I did this process:1) duplicate the layer, 2) select the layer (command-A), and 3) delete everything (backspace). Each streamer was begun this way.
Next I would make the particular streamer. This was done in two parts. First I would create the path for the streamer using the pen tool and then stroking it with the brush tool. I would set the minimum thickness to 0%. Then I would create a new layer and, using the same path, stroke it with the brush using a minimum thickness of 100%. Afterward, I erased the second half of the 100% streamer so that it matched up with the 0% one and then merged the layers together so that it would be one glowing streamer. Using the eraser tool, I removed any parts of the streamer that should be behind Kinetic. Some of the streamers did not overlap correctly, but this could be fixed by changing the ordering of the layers. The streamer on Kinetic's left hand was a little trickier because it overlapped itself. To make the glow work correctly for this, I separated the streamer into different layers and kept them there.
These were the major steps my process went through on the cover for issue 1. The artwork was now ready for any lettering, logo placement or whatever from that point forward. I hope you enjoyed this.