Repository navigation
medialink: a listener catches up after a late burst instead of staying late - #358
Merged
Merged
Conversation
…g late Sound arrives as fast as it plays, so whatever a late burst leaves queued at the speaker stays for the rest of the turn, and each hiccup made every word after it later, up to the speaker's buffer limit. When the queue hasn't come down to the prebuffer for catch_up_ms (2 s), the intercom now skips enough frames to get back to target_ms, and counts them in stats['trimmed']. Measured end to end between two ESP32-P4 boards by sending a burst and hearing it come back from the far speaker: sound alone 500 ms late before, 175-185 ms after; sound to a board sending pictures 385-445 ms before, 185-345 ms (median about 220) after.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Sound arrives as fast as it plays, so whatever a late burst leaves queued at
the speaker stays for the rest of the turn, and each hiccup made every word
after it later, up to the speaker's buffer limit. When the queue hasn't come
down to the prebuffer for catch_up_ms (2 s), the intercom now skips enough
frames to get back to target_ms, and counts them in stats['trimmed'].
Measured end to end between two ESP32-P4 boards by sending a burst and
hearing it come back from the far speaker: sound alone 500 ms late before,
175-185 ms after; sound to a board sending pictures 385-445 ms before,
185-345 ms (median about 220) after.