Metrics Reference

CallMeter collects 90+ real-time measurements per endpoint per second during every test and probe execution. These metrics give you complete visibility into what is happening on the wire, from high-level quality scores like MOS down to individual packet timing measurements.

This reference documents every metric: what it measures, why it matters, and how to interpret it.

Metric Categories

CallMeter organizes metrics into eight categories, each covering a different layer of the VoIP stack.

Quality

The headline metrics that summarize call quality into actionable numbers. Start here when evaluating test results.

Network

Transport-level measurements that show what is happening at the packet level. Use these to diagnose the root cause when quality metrics show degradation.

Feedback

RTCP feedback messages exchanged between endpoints. These metrics are critical for understanding how endpoints react to quality problems.

Jitter Buffer

Receiver-side buffering metrics that reveal how well the receiver is coping with network jitter.

Audio

Codec-specific audio measurements including audio levels, silence detection, comfort noise, DTMF event tracking, audio bitrate, and packetization interval.

Video

Encoding and playback metrics including frame rate, resolution changes, keyframe intervals, freeze events, decoder errors, and discarded frames.

Call Timing

SIP signaling timing measurements captured once per call. Unlike the categories above, these are one-shot values rather than time-series data.

Impairment

Counters emitted by CallMeter's built-in network impairment simulator. They report what the platform deliberately did to the stream, so a degraded result can be attributed to the injected condition rather than to the network under test.

Time-Series vs One-Shot Metrics

Most CallMeter metrics are time-series measurements. The platform samples them every second throughout the call and records each data point with a timestamp. This produces a continuous timeline that you can chart, zoom into, and correlate across endpoints.

One-shot metrics are captured once per call and represent a single event or duration. Call timing metrics like post-dial delay and setup time fall into this category. They appear as single values in the results summary rather than as charts.

Send vs Receive Direction

Every metric has a direction: send or receive.

• Send metrics describe what the endpoint is transmitting. For example, packets_sent counts outbound RTP packets, and bytes_sent_rate_bps measures outbound bitrate.
• Receive metrics describe what the endpoint is receiving from the remote side. For example, jitter_ms on the receive side measures how much the incoming packet arrival times vary.

Some metrics are meaningful in both directions. Packet counts, byte counts, and bitrates exist for both send and receive. Quality scores like MOS and R-Factor are computed per direction based on the conditions observed in that direction.

Using This Reference

Each metric page in this reference follows a consistent structure:

1. What it measures — Plain-language explanation followed by technical detail
2. Why it matters — Business and operational impact
3. How CallMeter measures it — Measurement methodology
4. Thresholds — Good, warning, and critical ranges (where applicable)
5. Common causes of problems — What drives this metric out of range
6. How to fix it — Actionable remediation steps
7. Related metrics — Cross-references to metrics that provide additional context

Start with the Quality metrics for the high-level picture, then drill into Network metrics when you need to understand the underlying transport behavior.