Categories
- All Categories
- 1.2K Cubism Editor
- 1K Help
- 56 Tips and Tricks
- 148 Feature requests
- 50 Other Live2D Products and Tools
- Cubism Viewer (for OW, for Unity)
- 3 Material Separation Photoshop Plug-in
- 47 Other Live2D Products and Tools (Other Photoshop Scripts, Sample Data, etc.)
- 265 Cubism SDK
- 238 Help
- 11 Tips and Tricks
- 16 Feature Requests
- 6 SDK for Unreal Engine alpha
- 3 Download
- 1 Manual and Update History
- 2 Bug
- Feature Request
- 181 Community
- 37 General Discussions
- 9 Feature Requests
- 26 Projects, Showcase & Critiques
If you have any questions, reports, suggestions, or requests about Live2D, please send them to this forum.
※We cannot guarantee statements or answers from Live2D staff. Thank you for your understanding in advance.
Live2D Cubism
Cubism Products and Downloads
Cubism product manuals and tutorials
Cubism Editor Manual Cubism Editor Tutorial Cubism SDK Manual Cubism SDK Tutorial
※We cannot guarantee statements or answers from Live2D staff. Thank you for your understanding in advance.
Live2D Cubism
Cubism Products and Downloads
Cubism product manuals and tutorials
Cubism Editor Manual Cubism Editor Tutorial Cubism SDK Manual Cubism SDK Tutorial
[About macOS Sequoia] (Updated October 22, 2024)
Live2D Cubism Editor 5.1.02 now supports macOS Sequoia.
Other Live2D Cubism products currently released are not guaranteed to work on macOS Sequoia.
Please refrain from upgrading macOS, as it may not operate properly.
Live2D Cubism Editor 5.1.02 now supports macOS Sequoia.
Other Live2D Cubism products currently released are not guaranteed to work on macOS Sequoia.
Please refrain from upgrading macOS, as it may not operate properly.
What could be causing performance variations & issues between hardware?
As I work with some large models, live2d performance is very important. Over the past year I have noticed a wide variation between computers and would be very interested what might be causing it?
I'm making this post now, as I was able to test in a more controlled environment with some interesting results.
I tested:
Live2d: 5.0.6/5.1
- Ryzen 6900HX (RTX 3080 ti) - 17fps
- Ryzen 6900HX (iGPU) - 25fps
- Ryzen 8700G (iGPU) - 25 fps
- Ryzen 7800X3d (iGPU) - 42fps
They all have 8 cores with similar max clock speeds. The 6900HX & 3080ti surprisingly scored the lowest results, with the 7800X3D & iGPU getting 2.5x the fps in the physics window (Screenshot below). The integrated graphics on the 7800X3D is very low performance, so I was quite surprised at the results. In comparison the 8700G has similar specs, but still scored lower with a higher spec iGPU. Hence I was wondering if something like V-cache could be causing a bottle neck? The V-cache on the 7800X3d is much larger than normal. Or maybe another factor that is different between the 8700G and 7800X3D?
Here is some tests with other riggers computers. The only computer to reach similar results to the 7800X3D with iGPU was the 14700f + 4070 ti super and 5800x3d + 4070 ti.
FPS test results using a high spec model (fps when moving the model in the editor and in the physics window):
Please note the tests are a bit rough, but hopefully help in getting an idea of the performance variation.
Other riggers computers:
My own computers:
In addition I would also like to note I am getting cpu spike/micro stutter like in the thread below. All tests have some degree of stutter. With more consistent spikes/stutter on larger models & the 7800x3d.
relevant thread:
https://community.live2d.com/discussion/1743/cpu-spiking-repeatedly-in-physics-window-in-5-0#latest
I'm making this post now, as I was able to test in a more controlled environment with some interesting results.
I tested:
Live2d: 5.0.6/5.1
- Ryzen 6900HX (RTX 3080 ti) - 17fps
- Ryzen 6900HX (iGPU) - 25fps
- Ryzen 8700G (iGPU) - 25 fps
- Ryzen 7800X3d (iGPU) - 42fps
They all have 8 cores with similar max clock speeds. The 6900HX & 3080ti surprisingly scored the lowest results, with the 7800X3D & iGPU getting 2.5x the fps in the physics window (Screenshot below). The integrated graphics on the 7800X3D is very low performance, so I was quite surprised at the results. In comparison the 8700G has similar specs, but still scored lower with a higher spec iGPU. Hence I was wondering if something like V-cache could be causing a bottle neck? The V-cache on the 7800X3d is much larger than normal. Or maybe another factor that is different between the 8700G and 7800X3D?
Here is some tests with other riggers computers. The only computer to reach similar results to the 7800X3D with iGPU was the 14700f + 4070 ti super and 5800x3d + 4070 ti.
FPS test results using a high spec model (fps when moving the model in the editor and in the physics window):
Please note the tests are a bit rough, but hopefully help in getting an idea of the performance variation.
Other riggers computers:
My own computers:
In addition I would also like to note I am getting cpu spike/micro stutter like in the thread below. All tests have some degree of stutter. With more consistent spikes/stutter on larger models & the 7800x3d.
relevant thread:
https://community.live2d.com/discussion/1743/cpu-spiking-repeatedly-in-physics-window-in-5-0#latest