This could probably go in Technical, but it's not really a troubleshoot question.
I'm curious as to what everyone's feelings/experiences are in regards to baking details vs painting them. For the sake of example, lets assume we're working on a hard surface object with a lot of surface detail. It can be a robot, gun, whatever.
Taking a bake centric route, you would need to actually model the majority of the detail, like screw heads, seams, etc. Then you'd have to create and map a low poly version. This process is time consuming, but because a good chunk of your texture is reliant on the high poly, the low poly asset (mesh and textures) can be corrected or tweaked very quickly, without having to repaint a lot of the detail.
A paint centric route takes the opposite approach. The high poly mesh (if there is one) is used for major shapes only, with details like screw heads and seams added with something like nDo. While initial production time goes down, changing anything can be far more time consuming, depending on how intact the UV's are.
In your experience, which do you find to be the most efficient? Have you ever found yourself having to redo a lot of work because detail was painted instead of baked? Or have you found it better to just get things out the door as fast as possible, and worry about corrections as they come up?
Replies
Its a lot better way of working than leaving it completely up to an artists painting style.
The way that the industry has gone I think proves the power of the baking approach.
One of the things to remember is that you won't become more efficient at something if you never do it. So if you never hipoly model then it will always be slower.
You mention placing bolts - it's just as fast to do this using a baked workflow. You shouldn't be re-modeling bolts every time and there's plenty of one-touch placement options such as scattering scripts or IMM brushes in zbrush.