Like so many other 3D artists, I came to this field being a long time lover of fantastic fiction, be it in the form of books, movies or comic books. In my case, I’ve always leaned more towards science fiction, although I do enjoy other genres within fantastic fiction. The point here is that there are two things that can’t be missing in a science fiction loving artist’s list of projects: starships and robots. I created my first starship design a year ago, so it was just a matter of time before I faced my first robot design.
I guess this has taken me so long because after my starship scene, I dedicated myself almost completely to rendering and lighting and the creation of procedural shaders, leaving very little time to build up my modeling skills. I wanted to change this and create a robot I could proudly present to the world, so I embarked two months ago on a project I have been constantly abandoning and reviving ever since.
My first big roadblock was to determine the design style. I started to model blindly, without any kind of reference or previous sketching, which led me to create a lot of different pieces and components that looked rather good, but were not what I had in mind. The only thing that was clear to me was that I wanted the robot to be mechanically believable, so I wasn’t going to build it as a continuous single mesh that would deform when animated. Instead of that, I had decided to build each part separately. This, coupled with the fact that I had trouble setting a definite style, caused most of the pieces to match awkwardly or not at all, resulting in a very poor overall look.
After many fruitless efforts, I had the first of two revelations that would help me set a viable workflow: start designing the head, and once you have one that you are happy with, let it dictate the rest of the design.
I started to make progress this way and designs looked good, although I still had problems to make my model match what I had in mind. Then I reached the second revelation that would enable me to model the robot I wanted: sketch your ideas before modeling. This may sound obvious to other artists, as this particular advice has been voiced a thousand times in different boards, books or tutorials, but I had skipped that step so far because I didn’t like how crude my sketches looked. However, this experience made me realize that they are just sketches and they don’t have to look good. Their purpose is to let you quickly determine what works and what doesn’t, and to help you lock on what you really want. Plus, the more you sketch, the faster you do it, and the less crude your ideas look. Here you can see, chronologically ordered from left to right, my first three heads (without previous sketching) and the final one, based on a sketch I refined until I got something that really matched what I was looking for:
From that moment on, the process from concept to model was a lot more fluent, although I still had some modeling challenges ahead, such as the hands. They had way more moving parts and interlocking components than rest of the robot, resulting in higher polygon counts and a more complex modeling process. Here are some shots showing the left hand in different poses:
Although some minor detail polishing may take place in the future, I consider the modeling phase to be over, after this long learning process. Next up is the creation of final materials that will replace the current placeholders, and the setting up of a more advanced armature that is easier to pose. Current one is a very basic armature that I set up just for presentation purposes. Still, I couldn’t resist the urge to toy a little with some basic shaders to get this render:
Final shaders will probably look very different, but I kind of like this provisional look.