Jump to content
3DCoat Forums

V4.1 BETA (experimental 4.1.17D)


Recommended Posts

  • Member

im making this meshes to add:

hand

foot

ear

nose

Haha, i was thinking about something else :) Hand, foot, ear etc. can be combined to human body, so there can be one Humanoid mesh. Other mesh can be for example Car for hard surface. Third can be Grapes. Why grapes? Because of its complicated structure:

Grapes_blue_02.jpgfca7f653-3397-461c-b02

 

Maybe some kind of Mobius String:

mobius.jpg

Link to comment
Share on other sites

I continue to work, but let me tell you short description of what happens. I was supposing this task to last 2 or max 3 weeks. But it took around 6 or 7 weeks that is enormously big for me. It is frustrating and I feel that so long time and so big efforts are not correspoding to importance of this task - there are more important tasks that will take less time (for example layers groups or axial symmetry, bugfixing and a lot more). But I can't drop it, big progress done and I feel that I am close to very stable and nice result. So please be patient a bit more time, I need your support to finally finish it.

  • Like 2
Link to comment
Share on other sites

  • Advanced Member

Ok, just had a spare couple of mins so I thought I would try out the new Autopo routines.  How long should the process take? I ask, because I've just tried to retopo my Ben Grimm model and it sticks on 'CreatingCutGraph' - doesn't go past 0%. 

Link to comment
Share on other sites

Ok, just had a spare couple of mins so I thought I would try out the new Autopo routines.  How long should the process take? I ask, because I've just tried to retopo my Ben Grimm model and it sticks on 'CreatingCutGraph' - doesn't go past 0%.

On some meshes it may hang. Ensure you are not using hardsurface retopology.
Link to comment
Share on other sites

  • Advanced Member

I got a hang up also, I wasn't able to do further testing, but it was an uncomplicated surface mesh of about 600k triangles.  There was a fair amount of self intersections because I used crease clay on it and then in the vox layer RMB menu, I used extrude function(Love this function, but seems to cause problems in the mesh)  twice.  Then I decimated the mesh to around 100K triangles.

 

Then I did autotopo and painted my density, I didn't put on any guides, I had x-symmetry on and then it hung at 0%.

 

I'm not sure what it was, but maybe the self intersections caused some issues.

Link to comment
Share on other sites

  • Advanced Member

I have a lot of deep detail in the sculpt (Ben Grimm - aka Thing - brick details) 12mill in surface mode. Didn't use Hard Surface, nor symm, nor decimate or voxelise. Used guides. Hangs at 0% CreatingCutGraph.  I'm so busy at the moment, but I will try to have another go as soon as I can.

Link to comment
Share on other sites

  • Advanced Member

I am having issues retopoing anything at all now.

 

basic shapes like primitives merged together as rudimentary forms are taking forever.

I even tried messing about with a sphere and retopoing. It has created several holes and an unusable mesh (result attached)

 

The old system was not perfect but it was certainly very quick. I am certain I must be doing something wrong as it used to feel effortless.

post-421-0-83823400-1379427298_thumb.jpg

Edited by Marc Wakefield
Link to comment
Share on other sites

Spiral-like topology is really hard to avoid.

 

Every test i do found it fixed :D

 

------------------------------------------

 

Lowering quads density decreases probability of spirals. Sometimes it is worth to make lower density at the beginning and then subdivide model afterwards.

ok, i understood now the decimate options. ty.

-----------------------

Any chance to save the modulator map? it's getting tedious to redraw it all the time.

 

This upgrade is going to be a piece or art :D

Link to comment
Share on other sites

  • Member

Am also having various issues - crashes, extreme slowness, offset mesh created, weirdly broken meshes, very low poly meshes created occasionally even though target is set to 5,000+, viewport slowdown to 6-12 FPS. Don't mean to sound negative, just reporting my findings...

 

The only success I'm having is by turning off 'create edgeloops with AUTOPO' and using the old basic method which is good enough for now. Keep cracking at it Andrew, seems potentially awesome but still a few rough edges to smooth off.

 

 

 

 

post-982-0-38457000-1379433908_thumb.jpg

Link to comment
Share on other sites

  • Advanced Member

Someone mentioned Zbrush....and of course Modo has it's rather expensive CAD loader tools which create effective hard edged meshes.

 

Would this Dynamesh/Live Clay approach of Vitaly Bulgarov be easier, and more  effective?

 

http://i.imgur.com/31MINd7.jpg

 

Posted with all due respects to the work of this very accomplished artist not for the purposes of copyright infringement but for the purposes of scholarly dialogue.

 

It's an excerpt from this book;

 

http://www.bookdepository.com/book/9781118066805?redirected=true&gclid=COKbtrPp0rkCFYw1QgodhVQArQ

http://www.bulgarov.com/blackphoenix_d7.html

 

 

And Modo CAD loaders

 

http://www.luxology.com/store/CAD_Loaders_for_modo/

 

"Not just a file translator...it's got an INTEGRATED TESSELATION ENGINE"

 

http://www.youtube.com/watch?v=RFYDrkUjcW0#t=30

 

 

I'm not saying this is what Andrew should do; I'm merely suggesting that his approach may be too automated, be too far beyond the capacities of the available algorithms or today's available computers..Maybe what is needed is a more integrated approach halfway between autopo and retopo by hand...and I'm not just thinking of the clumsy way we now can try to prep the autopo with wobbly hand drawn indicator splines. That's what Modo seems to be doing with imported STEP and IGES files...some of its retopo is automated and other parts are laying down sub d splines by hand on the original imported mesh.

 

Maybe if Andrew ported in Open SubD  from Pixar he could use its tools to this effect on voxel and Live Clay sculpts..

Edited by L'Ancien Regime
Link to comment
Share on other sites

  • Reputable Contributor

I continue to work, but let me tell you short description of what happens. I was supposing this task to last 2 or max 3 weeks. But it took around 6 or 7 weeks that is enormously big for me. It is frustrating and I feel that so long time and so big efforts are not correspoding to importance of this task - there are more important tasks that will take less time (for example layers groups or axial symmetry, bugfixing and a lot more). But I can't drop it, big progress done and I feel that I am close to very stable and nice result. So please be patient a bit more time, I need your support to finally finish it.

This is a MAJOR feature, and it NEEDED some improvement, because the way it was, I rarely ever used it. Spiral edgeloops and it not putting strokes specifically along hard edges, where one may put a stroke guide down. It's more of a viable option now...just needs to work correctly. So, it was/is definitely worth the time investment. As I mentioned before, it is critical for anyone considering 3D Coat as a modeling alternative. They might be somewhat turned off it they have to manually retopo everything after they constructed their model(s) in the Voxel Room....feeling like they'd have to basically model twice. If Auto Retopo works on many of their models, then that concern goes out the window.

You're just used to a fast pace, and it's ok to take your time and get it right, on something this important. Plus, you're very close already.

  • Like 2
Link to comment
Share on other sites

I continue to work, but let me tell you short description of what happens. I was supposing this task to last 2 or max 3 weeks. But it took around 6 or 7 weeks that is enormously big for me. It is frustrating and I feel that so long time and so big efforts are not correspoding to importance of this task - there are more important tasks that will take less time (for example layers groups or axial symmetry, bugfixing and a lot more). But I can't drop it, big progress done and I feel that I am close to very stable and nice result. So please be patient a bit more time, I need your support to finally finish it.

 

 

Andrew, I know you like to work on new stuff, and know you're super quick at programming what you're thinking. That itself is amazing! Really. For you spending that amount of time must feel forever, and excruciatingly painful. BUT, it is very needed! It gives 3DC the level of polish that people want and need. I've given demonstrations of 3DC and people say, "Hey that's cool. But, it crashes a lot. We can't use that in production." Or some variation of that comment.

 

The auto retopology tools are finally nearing a level of polish that makes it's useable and adoptable. Now, do this in every area of the program! :) Really, I'm not joking. Spend that long on each area, on each function. In one year's time, 3DC will be the best, top shelf application you can find on the market for retopo, 3D painting, UV mapping, the UI and digital sculpting.

 

Doing that level of refinement and polish on each area, with a build every 6 weeks sounds perfectly fine to me. I'm sure most would agree.

 

 

Your work doesn't go unnoticed! I'm talking about with folks a few times a week, at least. It always comes up at our 3D User Groups here locally, even at Foundry and autodesk events. :)

  • Like 1
Link to comment
Share on other sites

  • Advanced Member

Andrew, I know you like to work on new stuff, and know you're super quick at programming what you're thinking. That itself is amazing! Really. For you spending that amount of time must feel forever, and excruciatingly painful. BUT, it is very needed! It gives 3DC the level of polish that people want and need. I've given demonstrations of 3DC and people say, "Hey that's cool. But, it crashes a lot. We can't use that in production." Or some variation of that comment.

 

The auto retopology tools are finally nearing a level of polish that makes it's useable and adoptable. Now, do this in every area of the program! :) Really, I'm not joking. Spend that long on each area, on each function. In one year's time, 3DC will be the best, top shelf application you can find on the market for retopo, 3D painting, UV mapping, the UI and digital sculpting.

 

Doing that level of refinement and polish on each area, with a build every 6 weeks sounds perfectly fine to me. I'm sure most would agree.

 

 

Your work doesn't go unnoticed! I'm talking about with folks a few times a week, at least. It always comes up at our 3D User Groups here locally, even at Foundry and autodesk events. :)

My sentiments exactly...

Link to comment
Share on other sites

I continue to work, but let me tell you short description of what happens. I was supposing this task to last 2 or max 3 weeks. But it took around 6 or 7 weeks that is enormously big for me. It is frustrating and I feel that so long time and so big efforts are not correspoding to importance of this task - there are more important tasks that will take less time (for example layers groups or axial symmetry, bugfixing and a lot more). But I can't drop it, big progress done and I feel that I am close to very stable and nice result. So please be patient a bit more time, I need your support to finally finish it.

 

I agree with what everyone else is saying; half-working features help no one. Finish the feature so that you and everyone else who wants it can be satisfied with the results before moving on to the other areas that need attention. And that goes for any feature you implement in the future as well.

 

I'm glad to hear you mention layer groups, and hope the dedication you're showing to auto-retopology is reflective of the effort you'll put in once work is started on the Paint room.

Link to comment
Share on other sites

  • Member

Hi All,

 

I'm using the 4.07 version -  am having a ton of trouble with the UV'ng -  at least with one particular model -  I tried manually and with the uv path ....and for example ...the right hand glove -  makes the new section ...but the left,  simply refuses to make a section ...may have to switch back to main release because there's no working around it ...even tried new paths    - I also suspect a memory leak somewhere because sometimes I have to shut down the software and restart it to clean out errors that seem to carry over from re-loading a file. 

 

Also ,  on the auto topo -  if I keep the edgeloops checked ....it hangs - remove edgeloops option,  it works....wasn't the case with the prior version.

 

and I can also confirm the holes issues -  I was violently modeling ( lol )  and insteand of stretching,  the voxels would rip -  I tried fixing them with overlapping,  but when I tried to smooth it out or add clay ..the holes came back.....and the software would randomly crash when I tried to use the bridge function - I think it has part to do with the brush randomly shooting off to different directions of the mesh and making the bridge bigger than it was supposed to be.

 

Overall I love the software and commend Andrew on his amazing progress  -  bugs and all!

 

Cheers :)

Link to comment
Share on other sites

  • Advanced Member

You know I spent half of today going back and forth putting the autopo through its paces and finding out a lot about just how it works and has to be set up. I'm finding out a lot about not only autopo but what you have to do to prepare the voxel sculpt for autopo, particularly preceding your autopo with CLOSE INVISIBLE HULLS and then FILL VOIDS. The jury is far from being out on  7A. There's a lot of work to be done just on figuring out the best way to use it.

 

A lot of the nasty holes and surface  topology anomalies we're seeing produced here are being caused by improperly prepared voxels. You should never assume your voxel sculpt isn't full of bubbles and weird interior chamber caves that can't be spotted by the eye.

 

I was mainly working with 6A and after I solved my problems I found that 6A is frankly amazing in its own right. Now that I've got my own mistakes squared away I'm going to reinstall 7A and give it some close scrutiny..

Edited by L'Ancien Regime
Link to comment
Share on other sites

Are you all so kind and don't forget to post errors and example files (if possible) into mantis too, please. So Andrew can ask questions or post the error status individually.

 

Click -> 3D Coat Bugtracker

 

Of course you are invited to continue your discussion here. But it makes it easier to trace issues when you post them into our bug system.

 

Thank you very much

Chris

Link to comment
Share on other sites

  • Advanced Member

I continue to work, but let me tell you short description of what happens. I was supposing this task to last 2 or max 3 weeks. But it took around 6 or 7 weeks that is enormously big for me. It is frustrating and I feel that so long time and so big efforts are not correspoding to importance of this task - there are more important tasks that will take less time (for example layers groups or axial symmetry, bugfixing and a lot more). But I can't drop it, big progress done and I feel that I am close to very stable and nice result. So please be patient a bit more time, I need your support to finally finish it.

 

Dont underestimate the importance of this feature. Its a major feature. Retopo work is very time consuming and I avoid it unless there's no other choice.. If this new one will enable cleaner and better topology its a big deal!

Link to comment
Share on other sites

Updated to 4.0.07A

Auto-retopology improved, should work better and give more stable result. But it still requires work.

Please send me any problematic files (bad quality or hang) directly to support@3d-coat.com with subject "Autoretopo problem" - for ease of seeking.

This will help to reach stable point much faster.

Please be careful wit:

1) Enabling hardsurface on soft meshes will give wrong result

2) Symmetry is very important!

3) Too many guidelines may drop down quality. Mark only really important guides. But in most cases guides are not required at all.

Thanks for feedback and patience!

Link to comment
Share on other sites

  • Advanced Member

Thanks for the rapid upgrade Andrew.

 

Even on soft bodies, hard edge works fine.

 

http://i.imgur.com/YmDOsjO.jpg

 

http://i.imgur.com/lVFpu4b.jpg

 

 

No bad spirals, just clean mesh, and this thing runs FAST.

 

The only thing that's throwing me into a state where the program doesn't respond is when I paint out small areas for extra details

Link to comment
Share on other sites

Thanks for the rapid upgrade Andrew.

 

Even on soft bodies, hard edge works fine.

 

http://i.imgur.com/YmDOsjO.jpg

 

http://i.imgur.com/lVFpu4b.jpg

 

 

No bad spirals, just clean mesh, and this thing runs FAST.

 

The only thing that's throwing me into a state where the program doesn't respond is when I paint out small areas for extra details

Good!

Btw, try with symmetry.

Link to comment
Share on other sites

  • Advanced Member

Good!

Btw, try with symmetry.

 

 

That particular statue doesn't take symmetry, but on the big frame symmetry works well except for some artifacting at the symmetry line both in 6A and 7

 

http://i.imgur.com/TJ2V6xX.jpg

 

http://i.imgur.com/0xyue2A.jpg

 

 

Scroll down to the full post from Wednesday Sept 19

http://3d-coat.com/forum/index.php?showtopic=8588&page=4

 

It's nothing that can't be quickly trimmed by hand but still it's not quite perfect.

 

Autopo with edge loops seems to work best for me.

 

http://i.imgur.com/KkV8bdc.jpg

 

Here's a comparison between voxelization and decimation both with edge loops and otherwise identical parameters

 

http://i.imgur.com/bjs6GuY.jpg

 

http://i.imgur.com/LbvEyn0.jpg

 

By contrast the decimation worked far far beyond even what 6A could do and it did it fast, like under 5 minutes.

 

http://i.imgur.com/EmaFM1n.jpg

 

http://i.imgur.com/WdR9XcN.jpg

 

http://i.imgur.com/iaa6PC4.jpg

 

There's no way 6A could have resolved the leaves and that candle holding spike like that.

 

The one bug I am consistently coming up against in 6A and 7 though is with the tweaking of auto density by painting it on weak spots that call for more topology; this invariable throws my 8 Xeon cores with 8 gigs of RAM into a state of "not responding", even in small amounts..

 

http://i.imgur.com/aHfXGgc.jpg

 

 

One last test before bed. Without edge loops doesn't give quite as fine a result but it's really really fast, like under a minute.

 

http://i.imgur.com/6HkI74H.jpg

 

http://i.imgur.com/eZkOL8U.jpg

 

And of course it gives a more even mesh since it's not set up to give extra detail to more delicate areas.

Edited by L'Ancien Regime
  • Like 1
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

×
×
  • Create New...