______________________________________________________

ONE
I started with a few quick smudges to get down the basic composition and movement I saw in my head. At this stage I work in black and white, at a relatively small size (the document is small, maybe 1024 pixels, and I'm zoomed out to maybe 25%, so it's about 2 inches wide on the screen. I use the mouse, not the tablet at this point as well, and keep the brush edge very soft. All of this is to keep it broad and simple. If the picture does not have strong bones, it's never going to be able to stand. In some ways pictures only get weaker as you develop them, so they need to be really solid to begin with.

A lot happens in a short time with these early stages, which remain for me the most critical in terms of ultimately determining how good the picture will be. If you get these stages right, the rest of the picture paints itself.

______________________________________________________

TWO - about 3 minutes elapsed
Here I added another swish for the third figure, and a hint of a horizon line behind the main hill.

______________________________________________________

THREE - about 5 minutes elapsed
Now I'm starting to look more closely at the value arrangement, being very careful to preserve the gestures I've established. There's so much randomness in those soft, loose marks, that you often don't realize how critical some of them are to the growing illusion of atmosphere and space. At this stage it's a lot like the "cloud game" (where you look at clouds and try to see forms in them). You're half directing the process, and half simply responding to what's come out.
 

______________________________________________________

FOUR - about 8 minutes elapsed
Now it gets dicey as I have to start to forge actual figures, limbs, heads, features, etc. out of that mass, without killing the gesture...

______________________________________________________

FIVE - about 10 minutes elapsed
I didn't see it at this point, but here is where I went wrong. I've included this stage as an example of just how delicate the process is. Note how the minor changes to the shape of the main figure's head have transformed his direction of motion from coming out of the picture in step FOUR (above) to moving more sideways (left to right, parallel to the picture plane).  I broke it.

______________________________________________________

SIX - about 15 minutes elapsed
Still not realizing that I've slipped a bit, I continue to develop the main figure. I'm pleased with how I've been able to get more and more specific with him, and his gesture is still very strong.

______________________________________________________

SEVEN - about 20 minutes elapsed
Now both figures are starting to look like gnolls (hyena / human hybrids).  But I now realize that the main figure should be thrusting out toward the viewer more, vs. just moving laterally. And I've also lost one of the figures...

______________________________________________________

EIGHT - about 30 minutes elapsed
I shrink the two figures, move them to the right, and add a new main figure. This way the three figures sort of radiate outward, and I'm back to my original composition.  Ah, the miracle of digital.

______________________________________________________

NINE - about 30 minutes elapsed
Right here I feel like I've basically nailed the final image. The gesture is still strong and loose, and there is enough detail and specificity for me to move to more of a rendering stage. So I start to think about color...

______________________________________________________

TEN - about 45 minutes elapsed
I play around a lot with color... maybe too much. I always want to try new things, and remain open to whatever may come up "accidentally." I work in Lab mode (sadly, not supported by Painter, but only by Photoshop), because it organizes the color (and value) space much more like the way the human eye does. I use color layers, as well as adjustment layers (e.g. color balance) to experiment with different schemes. I know I want it to be basically warm / yellow / reddish, to imply a sort of savannah setting, but that's about it.

This is one of the schemes I explored.

______________________________________________________

ELEVEN - about 1 hour elapsed
Here I switch the image to a vertical "portrait" format (I'll have to make a landscape version as well for use in game). Fortunately that does not affect the composition much.

I change the layout, rotate the image a bit, and home in on the basic final color scheme.

______________________________________________________

TWELVE - about 2 hours elapsed
Now I start really painting. I switch to RBG mode, move into Painter and cut loose. I'm pretty much past the point where a single bad stroke can kill the image, so now I focus on things like costume and accoutrements, facial expression, and generally sharpening up the forms with harder and more opaque brushes. Unlike in the early stages, here a lot more time passes without the image really changing in any fundamental ways.

At this stage I tend to work on the focal points first, so that I'm sure to not overdevelop the areas that are not the main focus. For example, if I worked too much on, say, the grass in the foreground, by the time I got around to doing, say, the main figure's face, I may find out that the grass is rendered too tightly and clearly. Instead I work on the areas that must have a fair amount of detail, then I move onto the supporting areas if they need a bit more (and only up to the point that they need).

______________________________________________________

THIRTEEN - about 4 hours elapsed
I continue to develop the details and sculpt the forms of the figures. A lot of beginners think this is the hardest part, but really it isn't. If the underpinnings of the picture are solid, this part takes care of itself. Actually the only hard part is still that if you keep your focus too narrow, maybe work too zoomed in, you will find that you've gradually sucked the life out of the image. So adding the details and concrete forms is not hard in and of itself - adding them while preserving the gesture and composition is what's tough.

I continue to adjust the overall value and color scheme as well, and add the sun behind the main figure to give the image more punch. I rotate the figures a bit and shrink them down so they fit in the composition better.

______________________________________________________

FOURTEEN - about 5 hours elapsed
Now I fine tune the main figure's facial expression (make him look a bit meaner), add a few highlights on the helmet, a nice bit of drool from the main figure... some more detail to the background. I also tipped up the second figure's weapon a bit, and add a few items of costume to him. I like the way the second figure is working a lot, so I want to be careful not to overdo him, nor to take focus away from the main figure.

______________________________________________________

FIFTEEN - about 6 hours elapsed
For the final stretch I move back into Photoshop, switch to Lab mode and bring the saturation of the image down just a bit. I tweak the levels and add a little bit of texture (mostly to the sky and foreground) via an overlay layer. DONE!

Six hours is not a lot of time, but then again this picture happens to have gone fairly smoothly. I could easily spend another 5, 10, 20 hours adding more and more detail, without the picture changing in any substantial way. Often that final polish is critical, though it can multiply the time investment a lot. For concept work you're generally better off to produce three pictures at this level vs. one at a more polished level.