Jump to content
3DCoat Forums

V4.1 BETA (experimental 4.1.17D)


Recommended Posts

  • Contributor

Note that in 3DCoat merging with the "on pen" option is available, which falls in the same category of tools as Insert Multi Meshes.

 

To me it appears a very odd and needlessly convoluted workflow to bake an image out of a 3D object (lossy in tendency) in order to transform portions

of a mesh to the shape of the original geometry – that is, when one can simply take the real (still editable) piece of geometry to perform that task as well :) .

 

I believe that this ear-sample created by the Mudbox-makers primarily was a proof of concept - intended to impress people. In this area that clip clearly was effective,

but I can not imagine anyone using that technique regularly – at least inside an application which allows merging in real geometry as well!

This is not intended as a talk against Vector Displacement – but I rather see its merits in use as texture-maps, derived from finished models.

Could not agree more .:)

Link to comment
Share on other sites

  • Member

I reported this problem to Mac programmer, he will contact you soon.

Excuse for this problem.

Hi Andrew, 

ok thanks - file retrieval not possible- is gone- data recovery failed. Please tell programmer to let me know if still is a problem in latest version or if something specific triggers this issue.

Link to comment
Share on other sites

Aaaarrrrrggggghhhh!!! :wacko: Booleans in Surface mode. Do it at your own risk....just like switching between Voxel and Surface mode used to be. Cause it was not multi-threaded. Nor are Booleans in Surface mode. It "might" take a minute...it might take all day. Hope you got another PC to work on something else all day. Going on 30min now and it's still chugging away on one thread. With 11 other threads left idle. :mad:

I don't know what you mean. It needs a only some seconds here...

I tried layer subtraction, layer merging and merging external objects. My testscene had about 2 million polys.

Link to comment
Share on other sites

  • Reputable Contributor

Note that in 3DCoat merging with the "on pen" option is available, which falls in the same category of tools as Insert Multi Meshes.

 

To me it appears a very odd and needlessly convoluted workflow to bake an image out of a 3D object (lossy in tendency) in order to transform portions

of a mesh to the shape of the original geometry – that is, when one can simply take the real (still editable) piece of geometry to perform that task as well :) .

 

I believe that this ear-sample created by the Mudbox-makers primarily was a proof of concept - intended to impress people. In this area that clip clearly was effective,

but I can not imagine anyone using that technique regularly – at least inside an application which allows merging in real geometry as well!

This is not intended as a talk against Vector Displacement – but I rather see its merits in use as texture-maps, derived from finished models.

I agree. It could be helpful in the Paint Room (MV mode), but I don't see it offering much more than what we already have in the Voxel room. But after speaking with Andrew yesterday, about the need for BAKING real vector displacement from the Retopo Room, he seemed open to giving it a try. So, anyone else who wants to see this happen...make sure to drop a note to Andrew (support@3d-coat.com).

Link to comment
Share on other sites

  • Contributor

Maybe we could solve that with a shader.

A shader dont need UVs.

 

 

Simple question.

Is ther a way to switch Y and Z axis? All of my other programms use Z for vertical movement.

Would be nice if i could keep that similar in all apps.

Edited by Malo
Link to comment
Share on other sites

  • Reputable Contributor

Maybe we could solve that with a shader.

A shader dont need UVs.

 

 

Simple question.

Is ther a way to switch Y and Z axis? All of my other programms use Z for vertical movement.

Would be nice if i could keep that similar in all apps.

You can swap the y and z on export, if you need.

Link to comment
Share on other sites

It is nice. But how to solve this in the sculpting rooms without having UVs before?

 

the original example use an alpha to sculpt geometry

 

In 3DC we can switch from sculpt room to paint room and fill all the surface object using fill tool with modulating type: custom and fill with freeze ON

 

Next when back to sculpt room... any extrude along normal axis do the trick

 

making this process more user friendly, automatic some steps and give the option do mapping trough stokes... will do the magic ?

Link to comment
Share on other sites

  • Contributor

You can swap the y and z on export, if you need.

Not realy what i mean.

I mean the axis in 3d coat. Like in the pose tool. X and Z are horizontal and Y is vertical.

If i click on "Place Gizmo along Z" it should place the gizmo vertical and not horizontal.

All of my apps and machines at work uses X and Y for horizontal movement and Z for vertical.

Link to comment
Share on other sites

  • Contributor

It is nice. But how to solve this in the sculpting rooms without having UVs before?

There is cube mapping option in Fill tool to be used instead of uvs.

But Uvs will produce better controlled results.

 

Most Zbrush noisemaker examples on the internet use Uvs.

Link to comment
Share on other sites

  • Contributor

So noisemaker alternative must be a Paint Room tool ?

Well,you can already use it to directly add depth/color ect in Paintroom wihtout filling with freeze before hand....so its already there,there is no need for an "alternative".

 

Solution is of course to allow import in surface mode that preserve Uvs/ vertex order...but it would be impossible to use LC brushes and keep those,so user would be only allowed default SF mode brushes.

Same thing with dynamesh in Zb,it pretty much kills any Uvs.

Link to comment
Share on other sites

  • Advanced Member

You wouldn't need UVs and doing it in the Voxel Room allows for greater fidelity than, say a 512x512 or 1024x1024 map, etc.  You could simply paint a freeze on the area you want to apply the noise to, but instead of it doing some kind  of VoxLayer or Cut & Clone, etc. it applies an even brush all around, or at least to the power of the freeze.  The frozen areas would be your UVs.  By the way, you can use all the NoisePlugs in ZBrush without UVs applied.

Link to comment
Share on other sites

  • Contributor

 By the way, you can use all the NoisePlugs in ZBrush without UVs applied.

Try it using tiled alpha maps like tiled fabric,metal rings ect... You will see that without uvs it does not look nice at all. :)

Cubemapping is much better in 3DCoat than in noisemaker anyway...

 

Also "the frozen areas would be your uvs" I dont see how that would give the surface any sorts of coordinates as fas as distribution of tiles...its just a selection you are talking about. Distributing fractal or procedural  noise over surface without uvs is easy,Fill tool does it already...but when distributing tiled images over a surface that has many curve/plane changes is very difficult to get nice distribution without uvs.

 

The specific picture Carlosan showed is using Uvs and would look awful without uvs in noisemaker.

attachicon.gifNOISEMARKER.jpg

Link to comment
Share on other sites

You can only import one OBJ FILE at a time, but if you gather all the objects you want to export from your host app, into one scene, you can select them all and when exporting all those objects are part of the file. It doesn't have to be one object per file.

 

Yea that's what I ended up doing, but it's a pretty terrible workflow :/

Right now for every update I had to do all these steps:

-Export the retopod pieces

-import the new reference pieces into max (from zbrush), and re-export as one obj

-import the new reference (group of) mesh(s), which resets the scene

-re-import the retopod pieces.

 

It really shouldn't require that much effort just to update one part of a reference mesh.

 

As a side note, why does the relax brush (shift) feature work in screen space? Is there any way to get it to relax based on the actual surrounding geometry instead of just whats facing the camera?

And I was fairly certain 3DC had an extend surface feature (for taking a border end extruding it), but I didn't see it while working this time?

Edited by PolyHertz
  • Like 1
Link to comment
Share on other sites

  • Advanced Member

Try it using tiled alpha maps like tiled fabric,metal rings ect... You will see that without uvs it does not look nice at all. :)

Cubemapping is much better in 3DCoat than in noisemaker anyway...

 

Also "the frozen areas would be your uvs" I dont see how that would give the surface any sorts of coordinates as fas as distribution of tiles...its just a selection you are talking about. Distributing fractal or procedural  noise over surface without uvs is easy,Fill tool does it already...but when distributing tiled images over a surface that has many curve/plane changes is very difficult to get nice distribution without uvs.

 

The specific picture Carlosan showed is using Uvs and would look awful without uvs in noisemaker.

attachicon.gifNOISEMARKER.jpg

 

You could distribute tiles with a slider.  As far as how it goes around curved surfaces, you could just use the new 3D lasso.  Whether or not ZBrush's Noisemaker looks good on a surface or not does not negate the fact that you're not required to have UVs for it to work.

Link to comment
Share on other sites

  • Reputable Contributor

Yea that's what I ended up doing, but it's a pretty terrible workflow :/

Right now for every update I had to do all these steps:

-Export the retopod pieces

-import the new reference pieces into max (from zbrush), and re-export as one obj

-import the new reference (group of) mesh(s), which resets the scene

-re-import the retopod pieces.

 

It really shouldn't require that much effort just to update one part of a reference mesh.

 

As a side note, why does the relax brush (shift) feature work in screen space? Is there any way to get it to relax based on the actual surrounding geometry instead of just whats facing the camera?

And I was fairly certain 3DC had an extend surface feature (for taking a border end extruding it), but I didn't see it while working this time?

Without being able to see what exactly it is you are trying to do (screen recording...many free options, like JING, VLC Player, etc), it's like trying to help someone in a college lab....blindfolded. Whatever you have in the Retopo room, you have different export options in the Retopo Menu (top of the list).

Link to comment
Share on other sites

  • Advanced Member

Note that in 3DCoat merging with the "on pen" option is available, which falls in the same category of tools as Insert Multi Meshes.

 

To me it appears a very odd and needlessly convoluted workflow to bake an image out of a 3D object (lossy in tendency) in order to transform portions

of a mesh to the shape of the original geometry – that is, when one can simply take the real (still editable) piece of geometry to perform that task as well :) .

 

I believe that this ear-sample created by the Mudbox-makers primarily was a proof of concept - intended to impress people. In this area that clip clearly was effective,

but I can not imagine anyone using that technique regularly – at least inside an application which allows merging in real geometry as well!

This is not intended as a talk against Vector Displacement – but I rather see its merits in use as texture-maps, derived from finished models.

I should have worded my original post a little better. The way I was looking at it was replacing greyscale bump maps with vector displacement ones instead. There is really no reason for it not to be just as universal by now. The dinosaur video AbnRanger posted was a good example. The ear video, which I've seen, is a different scenario to my mind and in that case I can agree with you (I merge often). I guess my thinking is that if you go to the trouble of building in the ability to generate them for use in rendering, how big of a step does it take to also use them as a brush like we see in the dinosaur video? Or are boolean operations better in terms of easier coding and more efficient use of a PC's resources? I'm not a programmer, so maybe someone who is can comment on this.

Link to comment
Share on other sites

Without being able to see what exactly it is you are trying to do (screen recording...many free options, like JING, VLC Player, etc), it's like trying to help someone in a college lab....blindfolded. Whatever you have in the Retopo room, you have different export options in the Retopo Menu (top of the list).

 

I uh...what? The two questions I asked in that post are rather straight forward I think (the last two sentences at the bottom), and what I said about reference meshes is just a statement/opinion that it should be improved so that sub-objects can be added/removed at will.

For export options, the only thing I said was I have to export whatever retopology data I have before updating the reference mesh because right now that resets the scene losing said data otherwise, and thus needs to be re-imported after the new reference is set up. Wasn't really presented as a question.

Link to comment
Share on other sites

  • Reputable Contributor

I uh...what? The two questions I asked in that post are rather straight forward I think (the last two sentences at the bottom), and what I said about reference meshes is just a statement/opinion that it should be improved so that sub-objects can be added/removed at will.

For export options, the only thing I said was I have to export whatever retopology data I have before updating the reference mesh because right now that resets the scene losing said data otherwise, and thus needs to be re-imported after the new reference is set up. Wasn't really presented as a question.

Describing what you are trying to do might seem straight forward to you, because it's your scene/model...but it could be a convoluted mess for others, trying to figure it out. Trying to help people "blindfolded" gets really frustrating after a while.You asked a simple question earlier. I gave a simple answer. Then you come back and say "that way sucks because of this, this, this, this, this, this, this and this."

 

BTW, merging all that stuff into the Voxel room could have saved you some time, but you must have your reasons for not doing so. Maybe someone else is willing to help.

Edited by AbnRanger
Link to comment
Share on other sites

I don't expect there is a better way then the answer you provided. What I stated about the current process being more lengthy then it need be was of a conversational piece or something for andrew to take note of for whenever he gets time to work on the Retopo room again (I'll probably add some things to Mantis later).

 

Thank you for your answer from before, I wasn't saying you were wrong, just that it would be nice if there ends up being a better way implemented at some point.

Link to comment
Share on other sites

  • Reputable Contributor

I don't expect there is a better way then the answer you provided. What I stated about the current process being more lengthy then it need be was of a conversational piece or something for andrew to take note of for whenever he gets time to work on the Retopo room again (I'll probably add some things to Mantis later).

 

Thank you for your answer from before, I wasn't saying you were wrong, just that it would be nice if there ends up being a better way implemented at some point.

Personally, I would never import using the Reference object option. Merge to the Voxel Room in Surface mode and the object comes in just like it was in was in the host application. Reference objects are pretty much legacy stuff, IMHO. Merging to Surface mode makes a lot more sense.

Link to comment
Share on other sites

  • Advanced Member

Updated to 4.0.14B

- Specularity blending modes introduced

- pose falloff applicable for lasso/rect selection inPose as well.

- different proxy mode problems solved

- camera shortcuts problems solved

- several mantis problems solved

 

Thank you Andrew, I'll be testing this out as soon as the download finishes, especially can't wait to play with the specularity blend modes! I've been hoping for that for quite a while and am very excited to see it implemented  :D

Link to comment
Share on other sites

wow

testing, ty !

 

specular channel is perfecttttttttttttttttttttttttttttttttttttttttttttt :rofl:

 

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

Pose tool

 

when switch from use free from to normal mode... the object jump to initial pose and then pose to the new position... to do undo, and redo the pose solve the jump

 

is a jump in history to initial state

 

* everything without left pose tool to another tool

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

//edit

 

this is cooooooooooool

 

post-10142-0-76698300-1391529610_thumb.j

 

  • Like 1
Link to comment
Share on other sites

  • Contributor

In 14B Downgrade/Restore Visible volumes to proxy mode seems really safe and reliable. :)

I used move tool with "through all volumes" on Decimated8X entire chracter a lot today,felt very safe ...I will test more.

Its great to get to aspect of sculpting workflow very solid.It can really help to do really  ambitious work in 3DCoat.

Link to comment
Share on other sites

decimate or reduce ?

 

its really needed the 2 ways ?

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

User refinement tweak question: This request helps actually ?

 

0000911: Moving Proxy Visualization to the bottom of the Vox Tree

http://3d-coat.com/mantis/view.php?id=911

 

I think yes, but can be wrong... any other opinion is welcome :D

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

×
×
  • Create New...