Barbel- my Overwatch style character design- is finally complete. Two and a half months of focused work has left me with something I’m relatively pleased with. There’s always room to improve though, and one of my favourite ways of doing that is to review the latest project and think of how to get better results next time.
The idea for Barbel was born sometime in early 2017- I knew I wanted to make a 3d character in the style of Overwatch, but I din’t have any specifics beyond that. Then one day, the idea came to me- almost fully formed- which never happens usually and is not be relied upon. I wanted a giant catfish armed with a trident like one of those Roman gladiators. I could see him hunched there in my mind’s eye doing some kind of daft “play of the game” pose and making weird fish noises. He had a space helmet full of water though. I ditched that early on.
What I learned:
In concept art 3d is superior to 2d.
Sketching is obviously a very fast way of presenting ideas, but the fact is that some ideas just don’t occur to you simply because you’re thinking in a 2d way.
When you have a lump of 3d clay and you can quickly manipulate it without preconceptions- you begin to think in a more 3d way and consider how forms twist and turn as they travel through space- how the silhouettes look from all angles rather just one or two key profiles, or how things overlap. All this stuff is obviously possible in 2d- it’s just harder.
It’s also pretty tedious to redraw a character from the back or the side, just so you can see a few areas to decide how they’ll look. In 3d you have one model- and you can see any angle you desire and what’s going to go there. Playing around in Gravity Sketch and Oculus Medium led to a few happy accidents that helped me overcome design challenges I was having at certain points. Which leads me on to my second point…
VR is the best way to concept design in 3d.
Wherever I thought that the 2D concept was underdeveloped or not clear enough I would play around in Medum or Gravity Sketch and quickly come up with a solution. Arguably you could use Zbrush or Maya or something traditional, but the immediacy of reaching out and drawing a line in the air, of genuinely working with three dimensions of input rather than the two offered by the monitor screen, of being completely immersed in a world where only you and your work exist- is pretty damn special! VR may be a little clunky at the moment, but it more than compensates for it by allowing for a state of flow that surpasses traditional modelling.
It’s also really fast. The aforementioned three dimensions of input mean you don’t have to move the camera every time you want to do a simple operation such as moving an object along an axis that isn’t perpendicular to the viewport. True stereoscopic vision means you don’t have to move the camera to get an understanding of form either. On top of all this- you can move, rotate and scale objects all at the same time, even as you move your view around.
Colour and form should be designed together
Good colour design requires you to balance different key colours, secondary colours and accents. If you have a model ready to texture and you haven’t decided how to balance these colours, you may find some areas of colour or material are bigger or smaller than you’d like, and it’s too late to change it easily.
By contrast- if you deal with colour and form at the same time you’ll get a much more immediate impression of how your design is going to look when it’s finished and textured. If you deal with this at the concept stage you have a lot more freedom to play with the design so that the hierarchy of shape and colour works as you’d like.