I’ve mentioned in several of my previous posts that the absolute best way to get people interested in your game before they can actually play it is with art/visuals. I finally decided to put my money where my mouth is and I spent the last month learning Blender, so now I’ve got all kindsa stuff you can look at!
So here’s where I’m at after about 1 month of using Blender…
It’ll probably be changed plenty by the end, but it exists now, and that’s what’s important. One thing I accidentally did right, and will probably repeat for the rest of the trees, was that I made what will eventually be the tree’s drops (seed, stick, log, etc..) contained within the tree model itself. This saves me the time of making those models separately, and thus, begins my quest for the shortest route…
Next, to try something a little more complex, I decided to see if I could make Elle’s head. I was mostly concerned with establishing an artstyle, so I didn’t worry about polycount and the like. I figured I could always find a way to simplify it in the end. I ended up with this panther-fox-human hybrid thing…
Realizing I didn’t want to spend too much time adding a bunch of details I wasn’t even going to be able to keep, I decided to go ahead and see what it was going to take to simplify this down into a reasonable polycount. Because Elle is played from an overhead view, polycounts need to be as low as possible as they will be squeezed into a small amount of pixels, which isn’t really a problem, because excess details are not going to be visible from an overhead view anyway. That was simple enough to fix without any loss in apparent quality using Blender’s decimate tool and a little cleaning up. But the bigger issue I need to consider, as a solo dev who wishes to finish his game at some point, is finding a process that can easily be repeated to create dozens of characters quickly and efficiently.
Before I dealt with that, I spent a bunch of time experimenting with baking the fur into a normal map.
In the spirit of my quest, I ultimately decided I was adding too many steps and will just come up with a ‘one-texture-fits-all’ fur setup.
Not quite ready to revisit the search for an efficient head-creation workflow, I decided to see what I could come up with for making animated bodies, once again, as quickly and efficiently as possible. This time I planned ahead. I wrote down every kind of animal-person that will be in the game. I then wrote down what body parts can be shared between characters, for example, the mammals have paws, birds & bats have wings, etc.. After I had narrowed down the amount of necessary body parts, I went to modeling them. I took care to make sure that everything snaps together at the seams and can all share the same rig and clothing. So now my workflow for creating animated NPC bodies is pretty simple. Choose your body parts, choose some clothes, make sure they’re snapped together, and texture it! They all share the same rig, so all animations just have to be done once. I haven’t quite gotten around to utilizing this workflow yet, but I tested it by starting Elle herself:
After spending some time trying to think of a way to do something similar for the character’s heads, it dawned on me that I could use Blender’s shape keys and their sliders to change certain features of the head separately and gradually in what results in a rudimentary “character-creation” game. Perfect!
Now for each character, all I have to do is set a few sliders, choose a texture, and bam! a new character is born. (I haven’t gotten it entirely implemented yet, but the ears demonstrate the concept)
When you consider how much time is saved by planning out and simplifying your workflow, it’s obvious that it’s not just useful for solo devs like myself. Once I’m finished implementing all of these setups, I should be able to create an animated NPC in 10-20 minutes. Otherwise, I think they’d take me about 2 days each. If there are 60 NPCs in a game, that’s 20 hours vs. 4 months. That’s the power of design and planning.