Limit Pixi atlas size - #1817
Conversation
|
Hi @KTibow! Thank you for your pull request and welcome to our community. Action RequiredIn order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you. ProcessIn order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA. Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with If you have received this in error or have any questions, please contact us at cla@meta.com. Thanks! |
|
Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks! |
Summary
Closes #1815
Why
Issue #1815 reported severely pixelated imagery on an Ubuntu ARM/Snapdragon X system. The canvas/DPR setup was correct, and bypassing Rapid's atlas path with direct Pixi textures made imagery render sharply. Capping Rapid's atlas size fixed the visible pixelation.
The reduced repro points to large normalized atlas coordinates being sampled through Pixi's WebGL shader path with mediump fragment UV precision. mediump only guarantees roughly 10 bits of precision, which is not enough to reliably address large atlases such as 8192x8192 on true-mediump GPUs. Keeping atlases at 2048px avoids relying on precision that WebGL/GLES does not guarantee.
This trades some atlas packing efficiency for correct imagery on affected hardware.
Testing
🏗 Building data...
🏗 Building css...
👍 css built: 104.335ms
👍 data built: 4.148s successfully.
(And also manually previewing Rapid locally)
Disclosure
Generated by GPT 5.5 with human oversight.