AudioClockSnapshot

One coherent reading of the selected audio output's presentation clock.

position is the source media timestamp estimated to be audible at hostTimeNanos, after device buffering has already been accounted for. Null means that no audio anchor is available, including while seeking, before the first device report and after teardown. Zero and negative positions are valid timestamps. This never substitutes a requested seek target or a video clock.

rate is media seconds per host second, zero while held. generation identifies continuous audio delivery and also changes on an audio track switch that leaves video running. It matches the generation delivered to AudioTap. quality describes the device timing evidence, not a guarantee about wireless latency or when a display actually presents a picture.

Read a fresh snapshot for every display frame. To estimate the media position after a known display delay, add rate * delay to position. A display callback from another clock domain must first be translated to this clock's domain. Do not subtract audio latency again.