Hair Curves: Create Guide Index Map: Group IDs over 2621 among other thresholds randomly delete a large amount of guide curves #116735

Open
opened 2024-01-03 09:32:44 +01:00 by SleepyEngi · 5 comments

System Information
Operating system: Windows 10
Graphics card: Nvidia RTX 2060 Super

Blender Version
Broken: 3.6, 4.0
Worked: n/a

Short description of error
The "Create Guide Index Map" node group for hair curves deletes guide curves at certain thresholds- the first one I found was 2621- over which, it will randomly delete a large number of guide curves- when they should in theory remain the exact same regardless of the number.
There may be other smaller thresholds- depending on the guide distance used. In my case, I was using 0.01, which's first major threshold was 2621.

Exact steps for others to reproduce the error
Based on the default startup or an attached .blend file (as simple as possible).
Change the "Group ID" input from 2621 to 2622 or anything over that. A large amount of curves disappears.

Details
Normally, the group ID is supposed to be used for parting: Curves will only get the guide_curve_index of a guide curve with the same group ID. So changing the group ID should not affect the guide curves themselves no matter what the value is.

But I'm experiencing minor random deletions on any change- from 0 to 1, it's unnoticeable- but from 0 to 100- you can see some shuffle around. The first major threshold is at 2621- right after going to 2622, a majority of the guide curves get deleted. This will happen again at other thresholds over 5000, 8000, etc (Not exact values). There may be smaller thresholds that are unnoticed if the guide curve distance is large enough to hide the fact that guide curves under that distance got deleted

Here, you can see the guide curves when the group ID is 0
Group ID: 0

Here we change it to 2621 - it is nearly identical
Group ID: 2621

But as soon as we change it to 2622, 66% of the guide curves vanish.
Group ID: 2622

**System Information** Operating system: Windows 10 Graphics card: Nvidia RTX 2060 Super **Blender Version** Broken: 3.6, 4.0 Worked: n/a **Short description of error** The "Create Guide Index Map" node group for hair curves deletes guide curves at certain thresholds- the first one I found was 2621- over which, it will randomly delete a large number of guide curves- when they should in theory remain the exact same regardless of the number. There may be other smaller thresholds- depending on the guide distance used. In my case, I was using 0.01, which's first major threshold was 2621. **Exact steps for others to reproduce the error** Based on the default startup or an attached .blend file (as simple as possible). Change the "Group ID" input from 2621 to 2622 or anything over that. A large amount of curves disappears. **Details** Normally, the group ID is supposed to be used for parting: Curves will only get the guide_curve_index of a guide curve with the same group ID. So changing the group ID should not affect the guide curves themselves no matter what the value is. But I'm experiencing minor random deletions on any change- from 0 to 1, it's unnoticeable- but from 0 to 100- you can see some shuffle around. The first major threshold is at 2621- right after going to 2622, a majority of the guide curves get deleted. This will happen again at other thresholds over 5000, 8000, etc (Not exact values). There may be smaller thresholds that are unnoticed if the guide curve distance is large enough to hide the fact that guide curves under that distance got deleted Here, you can see the guide curves when the group ID is 0 [Group ID: 0](https://i.imgur.com/rE6mOYc.png) Here we change it to 2621 - it is nearly identical [Group ID: 2621](https://i.imgur.com/tSF8SyO.png) But as soon as we change it to 2622, 66% of the guide curves vanish. [Group ID: 2622](https://i.imgur.com/SIK3Idu.png)
SleepyEngi added the
Priority
Normal
Type
Report
Status
Needs Triage
labels 2024-01-03 09:32:45 +01:00

I have looked at the node tree for quite long time, Still not quite sure how this works. Seems that Group ID somehow influences curve_guide_index value and this is then rejected when compared with index node.

In any case if I understand this correctly, it is broken so will confirm

I have looked at the node tree for quite long time, Still not quite sure how this works. Seems that Group ID somehow influences `curve_guide_index` value and this is then rejected when compared with index node. In any case if I understand this correctly, it is broken so will confirm

@SimonThommes Have use very odd way to implement this node group (add group to positions, merge nearest, move all back) (it was only way to achieve such effect, but i not sure if this is safe to use this thing at least without manual and notes about floating-point limitation) so floating point error is ocurve on such large value. This is expected, that this will be merge all due to it's relative positions is much smaller that group offset. So i'll close this as not a bug. But this can be changed today, via Groups to Instances node.

@SimonThommes Have use very odd way to implement this node group (add group to positions, merge nearest, move all back) (it was only way to achieve such effect, but i not sure if this is safe to use this thing at least without manual and notes about floating-point limitation) so floating point error is ocurve on such large value. This is expected, that this will be _merge all_ due to it's relative positions is much smaller that group offset. So i'll close this as not a bug. But this can be changed today, via `Groups to Instances` node.
Blender Bot added
Status
Archived
and removed
Status
Confirmed
labels 2024-01-06 15:30:45 +01:00
Member

The fact that this is expected behavior because a workaround had to be used at the time the nodegroup was implemented is not a reason not to fix it imo. However, using the Split to Instances node would still be a workaround and it alone does not fully solve the issue, since sampling the guide index map would still require a separate workaround. This should be solved with a Group ID input in the Merge by Distance and Sample Nearest nodes. That is something I have remarked as one of the common things that still need to be worked around in an ugly way (#114280) and the Split to Instances node does not fully alleviate the issue.

So instead of replacing one workaround with another one right now, I'd rather wait until there is the proper solution and go over all of the node-groups in the Essentials that use this workaround. This is also not the only issue that required a workaround in these setups. It's a shame that those are in there and inhibiting the functionality of the node-groups to some degree, but those are compromises we had to settle with for the time being.

The fact that this is expected behavior because a workaround had to be used at the time the nodegroup was implemented is not a reason not to fix it imo. However, using the `Split to Instances` node would still be a workaround and it alone does not fully solve the issue, since sampling the guide index map would still require a separate workaround. This should be solved with a Group ID input in the `Merge by Distance` and `Sample Nearest` nodes. That is something I have remarked as one of the common things that still need to be worked around in an ugly way (#114280) and the `Split to Instances` node does not fully alleviate the issue. So instead of replacing one workaround with another one right now, I'd rather wait until there is the proper solution and go over all of the node-groups in the Essentials that use this workaround. This is also not the only issue that required a workaround in these setups. It's a shame that those are in there and inhibiting the functionality of the node-groups to some degree, but those are compromises we had to settle with for the time being.

@SimonThommes Should this be reopened as known issue then?

@SimonThommes Should this be reopened as known issue then?
Member

Hm, yea probably!

Hm, yea probably!
Blender Bot added
Status
Needs Triage
and removed
Status
Archived
labels 2024-01-08 18:22:10 +01:00
Simon Thommes added
Type
Known Issue
and removed
Type
Report
labels 2024-01-08 18:22:21 +01:00
Iliya Katushenock added
Status
Confirmed
and removed
Status
Needs Triage
labels 2024-01-08 18:27:53 +01:00
Simon Thommes added a new dependency 2024-01-09 13:12:14 +01:00
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.

Depends on
Reference: blender/blender#116735
No description provided.