Hi folks, sorry if this is an uber noober kind of question, but when you’ve finished sculpting and retopologizing, do you normally bake the normals of the sculpt into the new mesh or does it normally have enough of the detail from the retopo? I’m new to sculpting and so far my renders are pretty good (I havent made anything too complex yet and i’m going for the softer disney look so maybe no normals is kind of working in my favor?)
But if I can achieve a look more accurate to the sculpt, I guess I should take it.
If baking normals after sculpts is a thing, (again, sorry if this is a facepalm kind of question) could anyone point me to any tutorials or tipsets on it? I’ve never ever ever been able to consistently bake normals accurately… they always come out messy and I don’t know why.
It really depends on what you are trying to sculpt, how close to the form the retopo is to the original and where you are going to use the resulting model.
On one hand if you mostly sculpting hard surfaces or you don’t have really high frequency detail and you are rendering the resulting models in for example cycles then you may not need to bake normal maps.
If you quite a bit more detail in your sculpt that you can reasonably retopo and particularly if the target is for a realime engine standard practice is to include enough detail in your retopo to create a plausible silloutte and use a normal map to fake the high frequency details. (assets that need to be animated have a few additional requirements).
When taking a static asset to a realtime engine. I will usually create a copy of the retopo’d mesh use limited dissolve with an angle of 10-20 degrees and UV seams on. I then bake the highres mesh against this version.
Basically the best way to know is to do a test and see how different the asset is with and without a baked normal map.