Testing Philosophy

Measure the fault being fixed and preserve the blast radius. Do not modify buffering, routing, UI, diagnostics, and priorities in the same pass unless the evidence proves the same root cause.

For DYY9 audio, correct and fast operation outranks perfect diagnostic counters. A brief occasional click/pop is acceptable; a live page that fails to open or opens late is not.

Operational Scale Testing

4, 8, 16, and 32 zone simulations are used to identify architecture stress before user experience degrades, not to find crash limits.

Audio Standard Testing

DYY9 Stored-Message Validation

TestResult
Duration-cache fixTrigger-to-pipeline reduced from about 2.5 seconds to about 0.52 seconds; large multi-second underruns eliminated.
300 ms stored delayFour-zone stacked manual message latency measured LPA 342 ms, LPB 312 ms, LPC 316 ms, LPD 317 ms.
Stored All CallMeasured 325 ms trigger-to-pipeline latency with no large underruns.
Sequence pathSequence-generated messages still use shared stored-message path; first sequence message entered playback path without adding the manual command delay.

DYY9 Live-Page Validation

ScenarioExpected / accepted behavior
Same-zone takeoverLive page interrupts stored/sequence audio on the same zone immediately.
Cross-zone isolationStored messages on unrelated zones continue through a zone live page.
All Call live pageAll stored messages and sequences stop globally; All Call live mic takes priority.
First-tap live pageFirst activation must produce route-open, sender-lock, and first microphone frame without requiring a second tap.
Repeated start/stopLive PCM queues clear on stop; no delayed microphone audio plays afterward.

Evidence Files

Networking / Phone Testing

Hardware / Thermal Testing