90 days to model and rig a character...is this standard?
To start, I've been 3d modeling for going on six years now without the aid of any courses or education in a related field. My first complete human character models only started happening about 3 years ago and prior to that the only models that made it passed zbrush hi-poly sculpting and became rigged low-poly models were monsters and things of that nature.
When the first rigged human models started becoming complete it took around 6 months for me to finish one model working in my free time about 3 hours a day (this process was post design, hi poly sculpt in zbrush, clean-up in 3dsmax, re-topology, baking and painting textures and then weeks of rigging). In the last few years my speed has become twice as fast but seems to have peaked at around 90 days or around 270 hours to reach an end result of a low-poly model with full articulation down to individual fingers and minimal facial movement.
Where my question lies is whether or not this an average amount of time to spend on one model or not? How much time can you invest in something before you consider your pace less than optimal?
The models in question are actually being used for comic creation purposes and are of in game cutscene fidelity (although I need lots of work on facial rigging).
When the first rigged human models started becoming complete it took around 6 months for me to finish one model working in my free time about 3 hours a day (this process was post design, hi poly sculpt in zbrush, clean-up in 3dsmax, re-topology, baking and painting textures and then weeks of rigging). In the last few years my speed has become twice as fast but seems to have peaked at around 90 days or around 270 hours to reach an end result of a low-poly model with full articulation down to individual fingers and minimal facial movement.
Where my question lies is whether or not this an average amount of time to spend on one model or not? How much time can you invest in something before you consider your pace less than optimal?
The models in question are actually being used for comic creation purposes and are of in game cutscene fidelity (although I need lots of work on facial rigging).
Replies
We roughly do that level of detail and our characters do a lot of conversational dialog.
For us it roughly breaks down like this:
2-3 days for concepts
1 week for high poly sculpt
1 week for low poly + unwrapping
1 week for baking and materials (this includes wrinkle maps)
1-2 days for facial morphs
2 days for rigging/skinning (we use CAT and Biped plus a custom facial control board)
There are 1-2 days of padding in each of those tasks to account for fuck ups, fuck offs, sickness.
Typically each stage is completed by the end of the day on Thursday and submitted for approval, Friday is used to make any corrections or used to get a jump on the next task.
My personal work is another story. It seems like I never finish anything.
I'm not terribly good at rigging and the anticipation of rigging and trying to make sure I've done everything I could topology wise to make it straight forward I think causes me to spend much more time than I should.
Being I am working alone on this stuff, all of these time intensive practices of mine also tend to rob me of steam and enthusiasm since when I run into a snag I can't tap anyone on the shoulder and get feedback/help with trouble shooting.
It's difficult to put yourself under the gun so to speak and I guess I should be happy I ever finish any of these things at all. Even if the project is a bust it still contributes to my "10,000 hours".
Speed comes from doing quality work over and over again. If all you have is speed without the quality then it's a pointless exercise. No one is going to hire someone who makes bad art but does it really quickly. "So you're really fast at getting to a bad result... ok... next"
Hit the quality mark first then work on speeding that process up.
Rigging is naturally a pretty tedious process which leads to a lot of people learning scripting so they can automate a lot of the mundane heavily repeated tasks. Learning a bit of scripting helps speed up other things like modeling.
One thing you can do is to record yourself and then watch it at 2-3x speed, to see where you're spending a lot of your time then come up with ways to get to the same results quicker. A lot of people learn one way of doing something and never really explore any other options, don't be that guy...
If you see yourself constantly going from the viewport to a button on the other side of the screen, you might want to create a shortcut, or add that command to the quad menu, or find a way to move the button to a floating tool bar.
It takes time to try things out but once you do, you find out what works the best for that scenario, but the next time you face it you won't be spending nearly as much time experimenting.
usually the longest amount of time i spend for an ingame ready "currentgen" asset is around 3 weeks, usually lower, most stuff is done in around 60-80 hours, though i usually am not doing the rigging, only when it comes down to using biped i from time to time am doing the setup but those are very rare occasions, there are just way better bguys for that than me for rigging.
So maybe together with some advanced rigging it can ramp up quite a decent bit, i can't say my experiences are just not in that area.