Voice Quality Metrics — MOS, PESQ, POLQA, and R-Factor
Learn about voice quality metrics — mos, pesq, polqa, and r-factor
Why We Measure Voice Quality
You can't improve what you can't measure. In a NOC environment, "the call sounded bad" isn't actionable. You need quantitative metrics that tell you how bad, what kind of bad, and where in the call path the problem originated.
Voice quality measurement spans a spectrum from purely subjective (asking humans) to purely algorithmic (computing a score from signal characteristics). Understanding these metrics and how to read them in dashboards is a core NOC skill.
Mean Opinion Score (MOS)
The MOS is a number from 1 to 5 representing subjective voice quality:
| MOS | Quality | Description |
|---|---|---|
| 5 | Excellent | Imperceptible distortion |
| 4 | Good | Perceptible but not annoying |
| 3 | Fair | Slightly annoying |
| 2 | Poor | Annoying |
| 1 | Bad | Very annoying |
Originally, MOS was determined by having a panel of listeners rate audio samples — literally asking people "how does this sound?" on a 1-5 scale. This is subjective MOS and is the gold standard for quality assessment, but it's obviously impractical for real-time monitoring.
In practice, when you see "MOS" in a dashboard, it's always an estimated MOS computed algorithmically. There are two fundamentally different approaches to this estimation.
PESQ and POLQA — Intrusive Measurement
🔒 Full Course Content Available
This is a preview of the Telephony Mastery Course. The complete lesson includes:
- ✅ Detailed technical explanations
- ✅ Advanced configuration examples
- ✅ Hands-on lab exercises
- ✅ Real-world troubleshooting scenarios
- ✅ Practice quizzes and assessments
Preview shows approximately 25% of the full lesson content.