Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

RetopoFlow 3.4.3 unusable on Mac M2 Chip #1423

Open
HyperAlch opened this issue Jan 16, 2025 · 6 comments
Open

RetopoFlow 3.4.3 unusable on Mac M2 Chip #1423

HyperAlch opened this issue Jan 16, 2025 · 6 comments
Assignees
Labels
bug: performance v3 Applies to RF version 3x

Comments

@HyperAlch
Copy link

Here is a video of how completely slow and unusable RetopoFlow is on M2 Mac

Screen.Recording.2025-01-16.at.1.26.45.PM.mov

The model only has 7,872 faces, well under the amount that RetopoFlow is suppose to be capable of handling. The lag makes it completely unusable and I have to use a Windows laptop with a MUCH worse graphics card to get anything done. An M2 chip is objectively a better graphics capable chip than a 10 year old junk graphics card on a old Windows laptop. Clearly there is something very wrong here.

@jlampel
Copy link
Member

jlampel commented Jan 16, 2025

Thank you for letting us know! Is this the case across different versions of Blender (you can see which version of RF to use for which past versions of Blender here) or just the most recent version of Blender and RetopoFlow?

There is a chance we could find an issue on our end, but if it's an issue with how Blender handles custom drawing in the GPU module on M2 specifically then it's not something we can fix directly but, in that case, we can help report it to Blender.

@jlampel jlampel added bug: performance v3 Applies to RF version 3x labels Jan 16, 2025
@HyperAlch
Copy link
Author

@jlampel I just tried blender 3.6 LTS with RetopoFlow 3.4.3 and I have the same issue

Blender 4.3 was the original version I tried in the video I posted

@jfranmatheu
Copy link
Collaborator

Report related to issue: #1246

@HyperAlch
Copy link
Author

HyperAlch commented Jan 17, 2025

Report related to issue: #1246

Most likely not related. They had a Nvidia card and partially solved their issue via a Nvidia control panel setting. I'm on a M2 Mac, and do not have access to that panel or setting. The 2 pieces of hardware work somewhat different. One is a dedicated GPU, the other shares a chip with the CPU (aka it is integrated). So a solution for one doesn't automatically mean a solution for the other.

Not to mention that it works fine on my Windows machine with a much older graphics capable card.

@jfranmatheu
Copy link
Collaborator

They had a Nvidia card and partially solved their issue via a Nvidia control panel setting

By default, all settings in the control panel are in their the default values, so likely that user had set custom values that would make the performance issue worse 😅. I have an NVIDIA graphics card, and as stated in my post in that issue, with default settings in the control panel I have the same issues, and all impact seems to go to the tools logic.

You can also see MacOS M2 user in the issue thread with the same issue. And other Windows users with the same issue.

The related issues with this performance issue in tools:

The core argument to argue that the multiple named issues are related is that since the performance issues:

  • Occur across different operating systems (Windows, MacOS)
  • Happen with different GPU vendors (Nvidia, AMD)
  • Manifest with different CPU types (Intel, AMD)
  • Appear regardless of hardware capability
  • Show up with a relatively small number of faces depending on the system specs (2000, 5000, 7,872...)

The conclusion being drawn is that this must be a RetopoFlow performance bug rather than:

  • A driver issue
  • A platform-specific problem
  • A GPU backend issue
  • A hardware limitation

The supporting evidence includes:

  • Multiple GitHub issue reports showing similar symptoms
  • Consistency of the problem across diverse hardware/software configurations
  • The fact that some tools lag more than others (or affect in different ways), suggesting a software-specific pattern

@HyperAlch
Copy link
Author

@jfranmatheu Thank you for taking the time to clarify that for me. I understand your conclusion now.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug: performance v3 Applies to RF version 3x
Projects
None yet
Development

No branches or pull requests

3 participants