This was a experiment to take a Fusion 360 model into a “game” engine. My ideas was to auto generate the lo-poly in Houdini and just have to do a little UV touch up manually.
The steps I took was: 1. Fusion 360 modelling 2. Houdini for auto generating lo-poly and “start UVs” 3. Modo to clean up UV’s 4. Substance Painter
I converted the Fusion 360 mesh to VDB in Houdini since there was alot of lose meshes, convertVDB to get it back as polygons and then a Polyreduce SOP, and the weak link here is the PolyReduce SOP, it generates wierdly triangulated mesh and it's hard to pull down the polycount to something more acceptable without getting even more artifacts. I did a group sop for edges which I used to generate a “base UV” as I called it which I then fixed up in Modo.
But, the newcomer, OnShape, has really came along nicely. With their latest release that introduces “FeatureScript”, which allows you to create your own features using the same Scripting language that the Dev's use is really awesome.
No more “feature” request, you can just make it yourself…and if its a really killer feature, you can sell it to others….similar to an HDA.
@Luiz Kruel, Yep tried not happy with it, there are 2 main problems for me. sometimes when “following” a line there will still be alot of splits along it, adding to me what is unnessarcy geo. The 2nd problem is the rotation of the triangles, which sometimes are shockingly bad making round holes not round etc.
@Daryl , i'll look into it but somewhat allergic to “real programs” running in webbrowser, can it export decent meshes ?
@Debris, there interesting defently have to look into ! thanks
MagnusL3D Doesnt look like Instant Mesh plays well with hardsurface models.
Eeeek! Yes, this looks a bit weird. But I was very successful with a bit simpler hardsurface models, where I kind of helped the app by combing the orientation field.
Yeah my Original Houdini method was also working alot better when it was simpler things like the attached images. as soon as it gets more complex thats when it starts failing.
Damn, that looks like it did a really good job with the small piece.
Obvious question, have you tried running the poly reduce on the small pieces to bring it down slowly and then combine it and do a polyReduce for your target poly count?
That might keep the edges a bit better and still get your desired count
lkruel Damn, that looks like it did a really good job with the small piece.
Obvious question, have you tried running the poly reduce on the small pieces to bring it down slowly and then combine it and do a polyReduce for your target poly count?
That might keep the edges a bit better and still get your desired count
Thank you, not sure what you mean, poly reduce and then combine and poly reduce again ? combine what ?
I've had luck in the past taking several sub objects, dropping them individually down to around 10-20k and then doing a merge on the chunks and doing a final reduction to my target count. In my case I had all of the parts come into Houdini as individual OBJs, but in your case you might just do a connectivity and not have to worry about combining it at all.
