Fix playback starvation on long sequences at high speed - #641
Fix playback starvation on long sequences at high speed#641mapillaryanders wants to merge 1 commit into
Conversation
Above PlayService.sequenceSpeed the graph switches to GraphMode.Sequence, where play() gated image prefetching on cacheSequenceImages$. That resolves only once every image in the sequence has its metadata cached, so on long sequences the request window never runs and playback falls back to appending one image per round trip - making a higher speed setting slower than a lower one. Gate on cacheSequence$ instead and run cacheSequenceImages$ alongside it, so the batching optimization no longer blocks playback. Also resolve sequence positions through a Map rather than per-frame indexOf scans, and skip the request window while the trajectory is still in another sequence, where indexOf returned -1 and images were requested from the start of the sequence.
|
Hi @mapillaryanders! 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! |
1 participant
Above PlayService.sequenceSpeed the graph switches to GraphMode.Sequence, where play() gated image prefetching on cacheSequenceImages$. That resolves only once every image in the sequence has its metadata cached, so on long sequences the request window never runs and playback falls back to appending one image per round trip - making a higher speed setting slower than a lower one.
Gate on cacheSequence$ instead and run cacheSequenceImages$ alongside it, so the batching optimization no longer blocks playback. Also resolve sequence positions through a Map rather than per-frame indexOf scans, and skip the request window while the trajectory is still in another sequence, where indexOf returned -1 and images were requested from the start of the sequence.
Motivation
(Write your motivation here.)
Have you read the Contributing Guidelines?
(Write your answer here.)
Contribution
(If this PR adds or changes functionality, describe it here.)
Test Plan
(Write your test plan here. If you changed any code, please provide us with clear instructions on how you verified your changes work.)