Jump to content
3DCoat Forums

farsthary

3DCoat Developer
  • Posts

    276
  • Joined

  • Last visited

Posts posted by farsthary

  1. Hi :)

     

    Thanks for your feedback, and there's more coming :P

     

    @Haikalle and Artman:

    That is not actually an error, is a limitation we had imposed hard-coded on the amount of faces per-cell around 20K triangles , during subdivision, because Andrew and I thought artist don't need to go crazy with such high poly-count in large strokes anyway. This can easily be changed (or exposed to the artist in detrimental of the simplicity) or even completely removed but your RAM can suffer if go crazy with tiny polygons. But after each end of stroke, cells are rearranged so the counter is reset, meaning is not a real limitation on how many detail you can achieve, but on a single stroke continuously only.

    Anyway, we can increase it since performance now is so good that average artist PC has more room to spare :)

     

    As more good news I've improved further subdivision algorithm and now is even faster, robust and "topologically beautiful" under any circumstances. :)

    • Like 1
  2. Hi :)

     

    Well, count me in. I used Blender for many sort of things ranged from artistic work to development too so somehow I know a bit from its inner quirk and tricks and I must tell you is amazing, and most importantly its rapid development pace. So bigger general 3D apps long ago had a serious competitor in Blender. in the beginning it was regarded as a newbie oriented app and indeed it was, but within few years of crazy development it has catch up more than one of the biggest out there (cough, 3dmax ,  cough.... :P) code wise is a pool of new ideas and very easy to experiment with it.

    So I personally think that while Blender is a very powerful all-in-one 3D suite, combined with specialized tools like 3DCoat is an unbeatable combination for production work!

  3. Merge on pen is similar but this tool is more like 2D photoshop but in 3D, no need to clone entire objects, just parts you're interested in.

    Also the merging is more organic instead of the hard merging of booleans.

     

    This supports also arbitrary topology, for example, mesh with holes wich merge on pen does not support. though it will adopt the merge on pen interface eventually. I was showing only a first test, UI/workflow is not reaady yet.

  4. @Michalis

     

    You can actually clone without merging, like instancing or scattering across the surface (I will add random orientation and stuffs) so still work for foliage.

    Perhaps if you want an exact copy of hands, etc, booleans will do the job better, but cloning fish or dragon scales for example, may be better with this. This tools introduces cloning with slight variations, deformation so is not aimed at perfect symmetry  Look at Meshmixer, is currently a subset of its main feature but eventually it will get there.

  5. Hi :)

    Those explosions should be gone in coming update, they have being spotted and fixed already and yes, the cause was exposed when a mixture of LC and Surface tool where used, but, again, should be solved in next update.

    The performance improvements have been done in most if not all LC tools.I haven't touched CUDA versions yet, but eventually I'll explore better GPU acceleration, not only from NVIDIA, but is not as straightforward as may sound, there are lots of algorithms that does not take advantage of 2-4 cores not even mention hundreds of them for GPU due to their linear nature, but I will explore that path too.

    Is just Im constrained now to test, stability and minor optimizations :p:

    • Like 3
  6. #AbnRanger

    That version of SnakeClay while good, has pretty unsafe defaults (too high detail level more evident in smaller strokes) and internal code does not restrict you from some intermediate steps that can increased poly count to bring your system to its knees, that is hopefully improved now with coming update...tough for safety we have sacrified somehow performance in that tool and I will eventually find a better solution, but as you request, stability is first.

×
×
  • Create New...