Monitoring Meeting Quality and Connection Health

Good meetings depend on good connections. StretchMeet continuously measures call quality and exposes it as quality telemetry (QoS), plus an engine status view for the underlying media service. This article explains what's measured, how to read it, and how to use it to diagnose problems.

What StretchMeet measures

During a meeting, StretchMeet collects real-time quality metrics per participant, including:

• Round-trip time (RTT) — latency between the participant and the media service. Lower is better.
• Jitter — variation in packet timing. High jitter causes choppy audio and video.
• Packet loss — packets sent vs. received/lost inbound and outbound. Loss causes dropouts and freezing.
• Bitrate — available incoming and outgoing bandwidth.
• Frame rate (FPS) — how smoothly video is flowing.
• Connection and ICE state — the health of the underlying peer connection and how it negotiated (connection state, ICE connection/gathering state, candidate types).

These metrics let you tell the difference between "the app is broken" and "this person is on weak Wi-Fi."

View quality telemetry

1. Open a meeting and go to its Quality / QoS view.
2. Review the metrics for each participant.
3. Look for red flags: high RTT, rising jitter, or climbing packet loss usually explain a participant's poor experience.

Participants' clients report these measurements automatically while the meeting runs, so the picture stays current.

Reading the numbers

• All green, low numbers → healthy call.
• High RTT → distance or a congested network path; expect delay.
• High jitter / packet loss → unstable connection; expect choppy media. Often fixed by switching from Wi-Fi to wired, or closing bandwidth-heavy apps.
• Low available bitrate → not enough bandwidth; video quality will drop to protect audio.

Engine and readiness status

Beyond individual calls, StretchMeet exposes the health of its media engine — the service that carries the media. The engine/readiness view reports whether the core pieces (media routing, connection negotiation, and browser transport) are ready to carry traffic. If the engine isn't ready, native features like live control and recording may be affected. This view is mainly useful for administrators diagnosing a platform-wide issue rather than a single participant's Wi-Fi.

A realistic example

During an important client call, one attendee's video keeps freezing. The host opens the Quality view and sees that attendee's packet loss climbing and available bitrate dropping, while everyone else looks healthy. That tells the host it's a local network problem, not StretchMeet. They ask the attendee to switch off VPN and move closer to the router; the metrics recover and the freezing stops. The rest of the meeting proceeds normally.

Tips

• Check QoS before blaming the app. If one person is bad and everyone else is fine, it's that person's network.
• Wired beats Wi-Fi for anyone presenting or on a critical call.
• Close bandwidth hogs. Large downloads, backups, and other video apps compete for bandwidth.
• Watch jitter and loss together. They're the clearest signals of an unstable connection.

Troubleshooting

• One participant has poor quality, others are fine. It's their local network. Suggest wired ethernet, disabling VPN, moving closer to the router, or closing other apps.
• Everyone has poor quality. Check the engine/readiness status; a platform-level issue may be affecting the media service.
• No quality data is showing. Telemetry is reported while a meeting is live on the native engine. External-provider meetings report quality within that provider instead.
• Metrics look fine but audio is bad. Check the participant's actual microphone and speakers, and browser audio permissions.

FAQ

Do I have to turn quality monitoring on?
No — clients report quality automatically during native meetings.

Can I see quality for a past meeting?
Quality telemetry is captured per meeting and can be reviewed for the meeting it belongs to.

What's a "good" RTT?
Lower is better; consistently low latency with minimal jitter and loss makes for a smooth call. Sustained high values point to network problems.