Process

The Speedpaint

I start of by blending 2 or 3 images or textures. I experiment with the various blending modes, adjustments, and may even paint or smudge the results. The point of this step is to explore and find an interesting color arrangement.

A few fast strokes later, and the idea arises. I also feel the need to extend the canvas a bit, since the robot will have fairly long legs. The proportions are all wrong, and there's not much sense of depht, but It serves the purpose of sparking my imagination. I'm ready to start the painting proper.

I start developing the idea a bit further. At this stage I'm trying to pay attention to composition, light and values, in order to get them right. I also introduce some design elements, but they are more mental notes than anything else.
During the whole process I use selections and transform parts, in order to get them into perspective, and enhance the composition. I also don't shy away from using adjustment layers, for quick color corrections and experiments. I try not to be too drastic when it comes to color changes, since I wan't to stick to my original palette.

The end of the speed-paint part of the work, clocking at about 30\40 minutes. The colors, values, composition and design are all where I want them to be. Of course, that's what I am thinking to myself at that point, but we will soon find out that nothing is ever set in stone.

Detailing

I start detailing the image, by integrating some photographic material.
One thing people may notice about my work, is that I don't have a particular "plastic" style. I don't like having a set process, since that can be one step to stagnation.
As such, I approach every image differently in terms o technique, depending on the image I'm doing. For instance, in this case, I chose to approach the detailing by photo collage, in a kind of matte painting style. The benefits I found from this aproach are that not only I can imply detail faster, with realistic results, and I can also find new shapes very easily. In this particular piece, this method was very usefull, especially for the "innards" of the robot, which would be a bit complicated to detail by hand.

The important thing here is to be somewhat resolute on what you are doing. Its very easy to get lost in experimentations, and the image will take forever to get done. So after placing about 6 new photos, I like to flatten the whole image. Of course, there may be lost oportunities by doing this, but I save different files, and it's better to make a decision, even if it's not the best of the best, than taking eternity to finish the image.

Curves, Color balance and Hue\Saturation are the tools I use the most to integrate the photos in the image. Also, blending modes can produce good results in some situations.

Obviously, I also paint while integrating the photos. I basically attack problems as I see them in the image. Flipping the image is a blessing.

Something in the image seemed a bit strange. I thought it was the right leg. In order to fix it, I duplicated the whole image, and distorted it a bit. Looking back, it may have been the worst decision I made, since it made the image more symmetrical, but I have to stick to it. The legs were that one thing in the painting that I started getting a bit unhappy about, and never seemed to get completely satisfied.

Another big change. Like I mentioned before nothing is set in stone, The flow led me to abandon the spherical shape of the main body of the robot, in favor of a more agressive feel. Also, I start being more confident about the main focal point.

I quickly paint, in overlay mode, the yellow stripes to break the bluish monotony a bit. they will quickly go away, but I make a mental note to add them later.

Finally, I start thinking about the guns, choosing to have a single weapon, to break the symmetry a bit. I like the fact that such a huge robot would have such a small weapon. Why is he so big then? What is he housing within that huge shell of his? That's where the viewer comes into the picture.
I also throw in a Curves layer to tweak the colors to a warmer palette.

The final version. I improved the design of the legs a bit more. I also thought the robot was a bit too big for my tastes, so I increased the size of the man standing beside it.

Looking back, I really think that the initial leg disposition may have been better, but we learn from our mistakes, and composition isn't really my strongest point at the moment.