Do I understand correctly that the usual topology that we strive for, and polygon count optimization don't matter whatsoever when it comes down to 3D printing?
As long as it's manifold, and with something flat to start the print, all is fair? Duplicate verts, millions of polygons, N-gons and all that? Or will the quality of the print still be affected by the quality of the topology as we would texture/shade/animate?
Is there a best practice?
Sorry, here's the flood of questions... lol.
It's always best practice to have a good topology,
Even if you're not going to animate or manipulate the geometry.
Think of it like not looking like a n00b.
Always give your best, as you learn more with your practice.
I agree 100% travezripley! And I definitely don't want to get into bad habits, or sound lazy.
But when downloading the models from printables through this course I've seen some pretty hideous topology. Unusable without proper retopo in the 3D world I'm accustomed to. However, if it does not influence the quality of the print, some of these downloads may serve a purpose.
And as far as polygon count... I'm curious about the limitation because let's say, a photorealistic sculpture can get pretty hefty, so we usually go through some extensive optimization workflows before texturing and animating. But if it doesn't matter for printing, then I wonder... would it make it for better resolution on the print?
Also, if we get something straight out of sculpt mode, being able to skip the retopology step would be a considerable time saver.
Yeah when it comes to 3D printing, the usual rules of topology go on vacation and anything goes. It's like when mom and dad are out of the house for the weekend and they left you the keys and you're like yeeeaaahh baby! So I say take the nice break, you deserve it from all that hard surface modeling.
So I've been doing some testing and I feel like too many verts make for a longer slicing process.
Also, I feel like the models I took care to actually get proper topology printed cleaner layers. Though I'm not sure if it's because I'm starting to tweak the slicing settings more thoughtfully too. It could be.
But I tested one tile with some text on it. V1 had the text as a curve, V2 had the text turned into mesh and not cleaned up, V3 had the text cleaned up.
I know when we export the slt file the modifiers are applied, but not sure how it handles curve objects.
They all ultimately printed, but oddly enough the curve printed the cleanest. The cleaned up text was second. I also tested making nesting it in the tile with a boolean, just intersecting the meshes, and doing a union boolean. The intercepting meshes were the cleanest.
Definitely not a scientific approach or enough samples for that, but I wonder if that is somewhat true across the board. I sure wish the booleans don't matter for quality... who has time for cleaning up that mess anyway?