Fix audio clock drift and adaptive jitter buffering
This commit is contained in:
+4
-2
@@ -177,8 +177,10 @@ Noise reduction does not gate speech.
|
||||
`TimeProvider` timestamps; tests inject a clock. Threshold changes are atomic; all
|
||||
processing state otherwise has one owner. Codec/DSP processing methods allocate no
|
||||
managed memory after initialization, verified across 1,000 combined cycles. They run
|
||||
on a managed worker, never the native real-time device callback. Native device rings,
|
||||
jitter, mixer, and audio scheduling remain later work.
|
||||
on a managed worker, never the native real-time device callback. Managed capture/playback
|
||||
rings compensate independent clock drift around a configurable 20/40/60 ms target. Receive
|
||||
jitter is duration-aware, reserves one codec frame for DRED/FEC look-ahead, and decodes each
|
||||
channel packet at its actual duration before the fixed 20 ms mixer stage.
|
||||
|
||||
## Initial managed server
|
||||
|
||||
|
||||
Reference in New Issue
Block a user