Yes, I’m starting to feel guilty when posting another example in stead of the happy announcement of QuadRemesher for Blender we’re all waiting for.
True, but I did notice that the new ZRemesher algorithm can suffer from some really mangled topology in certain spots, in order to achieve good retopology on the rest of the surface. In fact, I find myself often switching back to the legacy algorithm for organic model retopo.
I wonder if ZRemesher 3.0 was created by Maxime as well, or if a different developer coded the upgrade for Pixologic.
Indeed, amazing results. I’m not into retopo in general, but I’ve tried out some basic solutions for demanding models with complex geometry. These are the best results I’ve seen by far. Very clever positioning of quads and ideal density variation on the most demanding parts of the model.
There is no other retopologizing tool that gives results of such a quality nowadays. It would be good to see a Blender compatible version as soon as possible.
Here is another test I did in 3DS Max. This one was a little trickier, as I had some smaller/dense cuts with curves, and non-uniform faces/density. I ended up breaking off the end cap, and doing 2 passes, to get a better result, but it ended up looking pretty good in the end.
I will say, it doesn’t seem to handle meshes that are too smooth/no hard edges as well. Or rather, meshes with collapsed Bevels or basically a collapsed highpoly mesh. It can do some strange things. It might be a matter of settings in the tool, but so far I haven’t had really great results. Or the best results have been with a mesh that has hard edges/clean cuts for better edge detection.
This is to be expected. I beta-tested QRM in Max and just like ZRM it operates best on hard angles, due to the algorithm using smoothing groups/hard edges. Collapsed high poly meshes with rounded corners will be treated as organic meshes and output poor results.
Max has its own native auto-retop modifier in development at the moment. Has some nice features, but still in the early stages. It’s beta, but i can mention it as they’ve already demoed at Siggraph.
Also testing HardMesh plugin for Max at the moment. Very nice tool, and a modifier too.
Hmm… The previous ZRemesher algorithm (2.0, now called Legacy in ZBrush 2019) is quite competent in retopologizing organic meshes. As I do a lot of cartoon character modeling, I’d be disappointed if QuadRemesher wouldn’t retopo those nicely.
@Kickflipkid687, could you please show some QuadRemesher results of a smooth organic mesh, like a cartoony character? I’m eager to see that.
Not the best test maybe, but I took the dog from Zbrush, which is already Quads it seems like. If I Run the Tool, with same Target Quad Count, and also testing the Adaptive Quad Count, this is what I get. Seems to do OK… but you lose the fine details, at least in this case.
BUT, you can paint Vertex color where you want more detail, so I can test that too…
Thanks a lot @Kickflipkid687, much appreciated. I think the retopology structure looks quite OK on the organic models. I’ll mostly be retopologizing non-complex, symmetrical, organic models.
Also interesting is seeing the dog’s ears being preserved nicely. Auto-retopologizers have a tendency to choke on thin areas. That also happens using 3D-Coat’s Autopo.
Metin, please tell the dude to “chop-chop” and give us the dang thing already! I don’t care if it occasionally crashes and whatnot… I NEED this mofo… like yesterday! I’m ready to pounce on a subscription the very minute he has it ready.
myself to sleep every night QuadRemesher hasn’t been released for Blender yet.
The previous ZRemesher algorithm (2.0, now called Legacy in ZBrush 2019) is quite competent in retopologizing organic meshes. As I do a lot of cartoon character modeling, I’d be disappointed if QuadRemesher wouldn’t retopo those nicely.
