Diagnostics

When a call misbehaves — one-way audio, a rejected register, a codec mismatch — the SIP wire trace and the quality metrics are your two primary tools.

SIP wire trace

Set the SDK's logging to Trace to capture the full SIP exchange, including SDP bodies:

new VoipConfiguration { LoggerFactory = loggerFactory };  // factory configured at Trace

The trace shows REGISTER/INVITE/200/ACK/BYE and, crucially, the c=/m=/a= SDP lines you need to diagnose media problems (advertised address, negotiated codec, a=crypto).

Secrets are redacted: SDES SRTP keys (a=crypto ... inline:) and ICE passwords (a=ice-pwd:) are masked as <redacted> in the trace, so wire logs are safe to ship to a central log system. The crypto suite name and codecs stay visible for diagnostics.

Quality metrics

Subscribe to QualitySnapshotChanged (or poll client.QualityManager) to watch measured jitter, packet loss, RTT and peer MOS live — see RTCP quality metrics. Rising jitter/loss localizes a problem to the network path rather than signaling.

Telemetry

client.TelemetryManager exposes diagnostic sinks/traces for wiring the SDK's signaling observability into your own telemetry pipeline.

A diagnosis checklist

Symptom	First look at
Register fails	Trace: the 401/403 and the Authorization retry; credentials/realm
One-way / no audio	Trace SDP c=/m= address vs. reachability; NAT
Choppy audio	Quality snapshots: jitter/loss; codec mismatch in SDP
DTMF not detected	RFC 4733 negotiated in SDP; peer's expected DTMF mode
SRTP call fails	a=crypto present; policy Required vs. peer capability

Reproducing

Capture a Trace log of the failing call plus the quality snapshots around the failure — that pair is usually enough to localize signaling vs. media vs. network. See Troubleshooting for concrete fixes.