AudioInputStats
Defined in: audio-player.ts:138
An input’s counters and jitter-buffer state. Counters are cumulative; durations are microseconds.
Properties
Section titled “Properties”bufferedUs
Section titled “bufferedUs”bufferedUs: number;Defined in: audio-player.ts:161
The audio held now.
concealedUs
Section titled “concealedUs”concealedUs: number;Defined in: audio-player.ts:167
Silence played in place of frames that never came.
cutHeardAtUs
Section titled “cutHeardAtUs”cutHeardAtUs: number | null;Defined in: audio-player.ts:183
When the latest cut() is heard: the time, on the clock’s time source (performance.now() µs without a clock),
at which the first output rendered after the jitter buffer applied it reaches the speaker, so the cut audio has
stopped. null before any cut, and from each cut() until the buffer has applied it. How an interruption is
timed: from the talker’s request to here is how long its old audio kept playing.
cutUs: number;Defined in: audio-player.ts:176
Audio dropped by cut().
decoded
Section titled “decoded”decoded: number;Defined in: audio-player.ts:140
Frames handed to the decoder.
decoderResets
Section titled “decoderResets”decoderResets: number;Defined in: audio-player.ts:144
Times the decoder was reset or replaced: at a discontinuity, a timestamp jump, or after an error.
dryAtUs
Section titled “dryAtUs”dryAtUs: number | null;Defined in: audio-player.ts:159
The timeline timestamp where the buffer last ran dry while playing, until playback starts again; null otherwise,
and after a flush or a cut that empties the buffer.
errors
Section titled “errors”errors: number;Defined in: audio-player.ts:146
Decoder errors.
heldUntilUs
Section titled “heldUntilUs”heldUntilUs: number | null;Defined in: audio-player.ts:154
The timeline timestamp just past the newest audio held: the end of decoded audio. null when nothing is held.
jitterTargetUs
Section titled “jitterTargetUs”jitterTargetUs: number;Defined in: audio-player.ts:163
The jitter-buffer target now.
late: number;Defined in: audio-player.ts:142
Frames dropped on arrival: behind audio already decoded, or inside a cut.
playing
Section titled “playing”playing: boolean;Defined in: audio-player.ts:150
Whether the buffer is playing, rather than filling to its target.
playStarts
Section titled “playStarts”playStarts: number;Defined in: audio-player.ts:152
Times playback started: at the start, after running dry, and after a flush or a cut that emptied the buffer.
recoveredUs
Section titled “recoveredUs”recoveredUs: number;Defined in: audio-player.ts:174
Audio rebuilt for frames that never came from the next frame’s in-band FEC; counted as synthesizedUs is.
reports
Section titled “reports”reports: number;Defined in: audio-player.ts:185
Clock reports made.
synthesizedUs
Section titled “synthesizedUs”synthesizedUs: number;Defined in: audio-player.ts:172
Audio synthesized in place of frames that never came: libopus packet-loss concealment. Counted as it is decoded, less, from the next report, what the jitter buffer dropped because it had run dry by the time it came.
trimmedUs
Section titled “trimmedUs”trimmedUs: number;Defined in: audio-player.ts:165
Audio dropped to shed latency: held past what the target needed, or past the hard bound.
underruns
Section titled “underruns”underruns: number;Defined in: audio-player.ts:148
Times the buffer ran dry while playing and audio came back within the jitter-buffer ceiling.