When did it happen last time? The fix was rolled out 1 or 2 day after 732 release.
Could you provide your place file or invite to team create? if not is it ok if I bind you to a channel with more logging so you can upload logs?
When did it happen last time? The fix was rolled out 1 or 2 day after 732 release.
Could you provide your place file or invite to team create? if not is it ok if I bind you to a channel with more logging so you can upload logs?
yes, you can bind me to the channel please.
Yeah after updating it was working fine for a few times after opening studio, but now its no longer fixed and is unloading my particle textures again.
How easy is it to reproduce this issue? Are you able to share your place file?
Ill bump this since its happening to me as well, not over long periods of time but when I playtest the game 3-4 times consecutively in a short span of time it happens.
I am on Version 0.731.0.7310942 (64bit) no updates available rn
Creator Dashboard try downloading latest version here, the fix is in 732 release
On 732, I cannot even load textures when initially opening the place on studio. nothing is loading now.
Thousands of errors: MeshContentProvider failed to process, Unable to load plugin icon, Unable to load, Unable to generate, etc.
Now 80% of the loaded textures unload after 2 minutes of being in studio, I have not even booted the game
How do I downgrade studio? I fully cannot work now @rdev0_0
I can ask around, but my guess is that you could try running a different version located in C:\Users\[YourUsername]\AppData\Local\Roblox\Versions\[version-folder]\.
Can you share your place file so I can try to reproduce it? Alternatively, I can bind you to a channel with additional logs so you can share them there, or existing logs might also help.
It looks more like asset requests are returning ‘too many requests’ errors. But I need logs to confirm.
The texture unloading issue I was having with the previous build last night is fixed in the latest build.
However, mesh still is persisting as low LOD after a few game startups and shutdowns.
You can load into the place, it is fine to start and stop to test this which reproduces the issue https://www.roblox.com/games/15502302041/Pet-Simulator-99
Our game does not boot as a saved rbxl, so you must test in the dev place. I give you permission to do so.
You will want to mirror these studio quality settings to properly reproduce it.
Im a Principal Render Engineer looking at these Mesh issues. (LODs)
We have fixes in flight and I wanted to try them out on your project
You could either give me access or a place file.
And/or you could try out the fix in studio the flag is ‘RenderFixMeshBatchNullGeometry’
Please feel free to ping me. Part of the challenge here is that even if we have a fix it takes weeks to deploy.
Still happens with flag on
2026-08-07T02:45:07.366Z,0.366236,1011c,6 [FLog::Output] LoadClientSettingsFromLocal: "{
"RenderFixMeshBatchNullGeometry": true
}"
The repro is to boot the game and let it run for ~10 minutes, and then stop. Upon stopping the studio view freezes/locks up while meshes struggle to load back in. Then after it unfreezes, the meshes display as low LOD.
Feel free to login to our dev place. I assume staff can join any place.
I thought this bug was only happening to me. Had enough of it so decided to look around for a solution and found this thread. I’ve followed the advice, updated studio to the latest version (0.732), and this bug still happens.
For more information on my particular case - I run studio on both a windows device and M4 MacOS device. The windows desktop keeps things fine texture-wise (other issues such as major edit mode lag are an issue but not part of this), though my MacOS computer removes the textures pretty consistently still. It’s become a really annoying issue
Hardware:
Macbook -
M4 Macbook Pro
24GB Unified Memory
Windows -
9950X3D
RTX 5090
64GB Memory
Can you please share logs from your Mac?
Can you also share your place or invite to team create?
DM it to the Logs / Crash Dumps / Other Bug Files group.
I believe my issue might be linked as well?