Bone Heat Weighting: failed to find solution for one or more bones #51250

Closed
opened 2017-04-19 15:08:50 +02:00 by gregus · 12 comments

System Information
win 7 SP1, GTX 670M

Blender Version
Broken: 2.78a and 2.78.4

I can't succeed in parenting a character to a rig with the "automatic weights" function. "Bone Heat Weighting: failed to find solution
for one or more bones" is always returned.
Character mesh seems okay, there are no separated parts and no non-manifolds vertex/edges.

  1. select the character
  2. select the rig
  3. CTRL +P with "automatic weights"

attached .blend file : AUTOMATIC WEIGHTS.blend

**System Information** win 7 SP1, GTX 670M **Blender Version** Broken: 2.78a and 2.78.4 I can't succeed in parenting a character to a rig with the "automatic weights" function. "Bone Heat Weighting: failed to find solution for one or more bones" is always returned. Character mesh seems okay, there are no separated parts and no non-manifolds vertex/edges. 1. select the character 2. select the rig 3. CTRL +P with "automatic weights" attached .blend file : AUTOMATIC WEIGHTS.blend
Author

Changed status to: 'Open'

Changed status to: 'Open'
Author

Added subscriber: @gregus

Added subscriber: @gregus
Author
[AUTOMATIC WEIGHTS.blend](https://archive.blender.org/developer/F551879/AUTOMATIC_WEIGHTS.blend)
Author

Hello, Here is a copy of the discusssion previously open on Discussion list to assist animation developers On Mon, Apr 17, 2017 :

https://lists.blender.org/pipermail/bf-animsys/2017-April/thread.html


Hello,
I can't succeed in parenting a character to a rig with the "automatic weights"."Bone Heat Weighting: failed to find solution for one or more bones" is always returned.
This method appear to work only with characters made out from cubes....

The characters I use are regular and imported from Poser or created with Manuel Bastioni software.
I read many posts about crossing geometry, non-manifolds, normals, cleaning up etc...I spent hours trying to clean a character, removing non-manifolds, normals, no extra point, no crossing geometry etc... in vain

I have some queries :

  with "Automatic Weights"  and during the process :
  1.   the parts of the mesh with poly/vertex problems should be left aside, but on the correct part of the mesh the script should go on and assign weights.2 .      the parts of the mesh with poly/vertex problems should be shown selected so that we know WHERE is the problem
    

I wonder why this algorithm is so unefficient where "envelope weights" always returns a weightmap zone which is usually pretty good but too wide.(spreading on opposite side of the character)Because of its name, would you answer but "Envelope weights" has no problems with non-manifolds or whatsoever. and does the job.

In conclusion, "Automatic Weights" should be less restrictive, more informative and assign weightmaps where it can instead of nowhere at all.

Thanks for your help.
Gregus

Blender 2.78a

Hello, Here is a copy of the discusssion previously open on Discussion list to assist animation developers On Mon, Apr 17, 2017 : https://lists.blender.org/pipermail/bf-animsys/2017-April/thread.html -------------------------------------- Hello, I can't succeed in parenting a character to a rig with the "automatic weights"."Bone Heat Weighting: failed to find solution for one or more bones" is always returned. This method appear to work only with characters made out from cubes.... The characters I use are regular and imported from Poser or created with Manuel Bastioni software. I read many posts about crossing geometry, non-manifolds, normals, cleaning up etc...I spent hours trying to clean a character, removing non-manifolds, normals, no extra point, no crossing geometry etc... in vain I have some queries : ``` with "Automatic Weights" and during the process : ``` 1. the parts of the mesh with poly/vertex problems should be left aside, but on the correct part of the mesh the script should go on and assign weights.2 . the parts of the mesh with poly/vertex problems should be shown selected so that we know WHERE is the problem I wonder why this algorithm is so unefficient where "envelope weights" always returns a weightmap zone which is usually pretty good but too wide.(spreading on opposite side of the character)Because of its name, would you answer but "Envelope weights" has no problems with non-manifolds or whatsoever. and does the job. In conclusion, "Automatic Weights" should be less restrictive, more informative and assign weightmaps where it can instead of nowhere at all. Thanks for your help. Gregus Blender 2.78a
Author

Answers :

Hi;

This issue happens often when a mesh is made of 2 or more unconnected
mesh parts in the same mesh object. For example when you create a set of
teeth for a character but modeled them all in one object. Or mesh based
hair strains, things like that. The weighting algorithm then can not
decide how to distribute the weights for some of the vertices.

The workaround is as follows:

1.) separate the Mesh object into parts ('p' in edit mode) where each
part contains one of the partial meshes
2.) Do the automatic weight from bones for all parts
3.) Join the parts back into one object

It should not be too complicated to improve the weighting operator: The
operator could run the weight algorithm for each unconnected mesh part
separately. Then the user would not need to bother with separating the
parts and later rejoining them.

Regarding the information about which vertices are affected: I have
looked at this a while back, but i could not figure out how to get the
list of affected vertices from the weighting functions. I looked into
this down to the function

bool EIG_linear_solver_solve()

But this function only returns either success or fail for an entire mesh
object. And i am not sure how i can get the list of affected verts from
that function. Maybe someone can give a hint or two here ? A nice way to
show the affected verts would be to automatically select them. So when
the user opens EDIT mode the affected verts shine in yellow.

cheers,
Gaia

On 18.04.2017 10:14, Sybren A. Stüvel wrote:

Hello Grégory,

On Mon, Apr 17, 2017 at 02:17:15PM +0000, Grégory Roger wrote:

I can't succeed in parenting a character to a rig with the
"automatic weights". "Bone Heat Weighting: failed to find solution
for one or more bones" is always returned.

Can you open a bug report at https://developer.blender.org/? There
you'll be able to attach a blend file that shows the issue. Please try
to make the file as minimal/small as possible to show the problem (so
not with a rig of 100+ bones).

with "Automatic Weights" and during the process :

  1. the parts of the mesh with poly/vertex problems should be left
    aside, but on the correct part of the mesh the script should go on
    and assign weights.

I don't know what that sentence means.

  1. the parts of the mesh with poly/vertex problems should be shown
    selected so that we know WHERE is the problem

That's a good idea.

Blender 2.78a

Before sending in a bug report, please try a more recent version of
Blender (either the last release or, even better, a daily build from
https://builder.blender.org/download/).

Kind regards,


Bf-animsys mailing list
Bf-animsys at blender.org
https://lists.blender.org/mailman/listinfo/bf-animsys

Answers : Hi; This issue happens often when a mesh is made of 2 or more unconnected mesh parts in the same mesh object. For example when you create a set of teeth for a character but modeled them all in one object. Or mesh based hair strains, things like that. The weighting algorithm then can not decide how to distribute the weights for some of the vertices. The workaround is as follows: 1.) separate the Mesh object into parts ('p' in edit mode) where each part contains one of the partial meshes 2.) Do the automatic weight from bones for all parts 3.) Join the parts back into one object It should not be too complicated to improve the weighting operator: The operator could run the weight algorithm for each unconnected mesh part separately. Then the user would not need to bother with separating the parts and later rejoining them. Regarding the information about which vertices are affected: I have looked at this a while back, but i could not figure out how to get the list of affected vertices from the weighting functions. I looked into this down to the function bool EIG_linear_solver_solve() But this function only returns either success or fail for an entire mesh object. And i am not sure how i can get the list of affected verts from that function. Maybe someone can give a hint or two here ? A nice way to show the affected verts would be to automatically select them. So when the user opens EDIT mode the affected verts shine in yellow. cheers, Gaia On 18.04.2017 10:14, Sybren A. Stüvel wrote: > Hello Grégory, > > On Mon, Apr 17, 2017 at 02:17:15PM +0000, Grégory Roger wrote: >> I can't succeed in parenting a character to a rig with the >> "automatic weights". "Bone Heat Weighting: failed to find solution >> for one or more bones" is always returned. > Can you open a bug report at https://developer.blender.org/? There > you'll be able to attach a blend file that shows the issue. Please try > to make the file as minimal/small as possible to show the problem (so > not with a rig of 100+ bones). > >> with "Automatic Weights" and during the process : >> >> 1. the parts of the mesh with poly/vertex problems should be left >> aside, but on the correct part of the mesh the script should go on >> and assign weights. > I don't know what that sentence means. > >> 2. the parts of the mesh with poly/vertex problems should be shown >> selected so that we know WHERE is the problem > That's a good idea. > >> Blender 2.78a > Before sending in a bug report, please try a more recent version of > Blender (either the last release or, even better, a daily build from > https://builder.blender.org/download/). > > Kind regards, > > > _______________________________________________ > Bf-animsys mailing list > Bf-animsys at blender.org > https://lists.blender.org/mailman/listinfo/bf-animsys
Author

For those who are not on the discussion list :

Hello Sybren,

I am going to open a bug report, the blend is about 6Mo, with only the character and the basic rig.

I was not clear in point 1, let me reformulate it :

  1. the parts of the mesh with poly/vertex problems should be left
    aside, but on the correct part of the mesh the script should go on
    and assign weights.

When the script has an issue in solving the assignation of a weightmap to a specific bone,it should leave this bone and jump to the next one so that the process can go on on the rest of the mesh which is correct.
Then, there can be an error message at the end of the process, but maybe 80-90 % of weightmaps are created and that's a big improvement.
The remaining "bad parts " can later be manually assigned.

Regards
Grégory

For those who are not on the discussion list : Hello Sybren, I am going to open a bug report, the blend is about 6Mo, with only the character and the basic rig. I was not clear in point 1, let me reformulate it : > 1. the parts of the mesh with poly/vertex problems should be left > aside, but on the correct part of the mesh the script should go on > and assign weights. When the script has an issue in solving the assignation of a weightmap to a specific bone,it should leave this bone and jump to the next one so that the process can go on on the rest of the mesh which is correct. Then, there can be an error message at the end of the process, but maybe 80-90 % of weightmaps are created and that's a big improvement. The remaining "bad parts " can later be manually assigned. Regards Grégory

Closed as duplicate of #45493

Closed as duplicate of #45493

Added subscriber: @RomboutVersluijs

Added subscriber: @RomboutVersluijs

I noticed this today as well on 2 different models. I havent done much rigging so i first did some research what it actually was. The error message isnt super clear. I was using 2.80 and that tried 2.79b and it showed the same. One model had quite some None Manifolds and separate meshes but the other didnt.

If it knows it fails, than it also knows where it fails. Why not use that info the possibly inform the user. Isnt this auto weighting done by some kind hitting rays from the bones and using that for weight painting. I believe i read something like that way long back.

I noticed this today as well on 2 different models. I havent done much rigging so i first did some research what it actually was. The error message isnt super clear. I was using 2.80 and that tried 2.79b and it showed the same. One model had quite some None Manifolds and separate meshes but the other didnt. If it knows it fails, than it also knows where it fails. Why not use that info the possibly inform the user. Isnt this auto weighting done by some kind hitting rays from the bones and using that for weight painting. I believe i read something like that way long back.

Added subscriber: @lauxmaux

Added subscriber: @lauxmaux

this is because of three vertices joining together. just delete that part.

this is because of three vertices joining together. just delete that part.

@lauxmaux That is a solution, but that not the issue im talking about. Im talking about the info the "Auto Weight Paint" returns when it know about where the error is. IF it knows there a 3 or more verts in close proximity, why no return that info to the user. The current error feedback is kinda useless for helping a user know where or what the issue is

@lauxmaux That is a solution, but that not the issue im talking about. Im talking about the info the "Auto Weight Paint" returns when it know about where the error is. IF it knows there a 3 or more verts in close proximity, why no return that info to the user. The current error feedback is kinda useless for helping a user know where or what the issue is
Sign in to join this conversation.
No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Interest: X11
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
No Milestone
No project
No Assignees
4 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#51250
No description provided.