Current Location ：
MOS Scores and What They Hide About Call Quality
A mean opinion score compresses an entire calling experience into one number between one and five, and managers love it because it fits on a dashboard. The compression is also the problem. Two calls can report the same score while one is merely quiet and the other is full of dropouts, and the number cannot tell you which. This article explains where the score comes from, what it is genuinely good for, and what to measure alongside it so that a green dashboard does not conceal a real complaint.
Where the Number Comes From
The original scores came from listeners rating recorded speech. Modern systems estimate rather than measure, deriving a value from network and codec parameters. That distinction matters: an estimated score describes the conditions the algorithm models, not the experience of a person in a noisy room using a mediocre headset. Because the estimate is computed, it responds to loss, delay, and codec choice, and it remains blind to acoustics, echo, and device quality.
| Score band | General interpretation | Where the model is weakest |
|---|---|---|
| 4.0 and above | Generally acceptable quality | Cannot detect acoustic problems or poor endpoints |
| 3.5 to 4.0 | Noticeable but tolerable impairment | Averages over bursty loss that a caller notices |
| 3.0 to 3.5 | Users will complain | May still score acceptably with compression artefacts |
| Below 3.0 | Clearly poor | Often reported only after the caller has already hung up |
What the Score Cannot See
The model does not hear. It cannot tell that a caller is on speakerphone in a glass room, that a headset microphone is failing, or that one direction of the call is fine while the other is not. It averages across a call, which hides a ten-second dropout inside a generally good conversation, and it treats all loss as equivalent even though bursty loss is far more disruptive than evenly spread loss at the same rate.
- Report per-direction metrics rather than a single combined figure
- Track loss burstiness, not only loss percentage
- Pair quality metrics with abandonment and repeat-call data
- Alert on the tail of the distribution, not on the average
- Treat a complaint as evidence even when the dashboard is green
Metrics That Answer the Actual Question
The useful question is not what the average score was, but how many calls fell below an acceptable threshold and what those calls had in common. Grouping poor calls by site, by trunk, by endpoint model, and by time of day turns a number into a diagnosis. Where a specific carrier route or a specific handset model dominates the poor group, the cause is usually findable without any further measurement.
Agreeing What Good Means
Service levels built on a single score invite arguments about the model rather than the service. A better arrangement states the threshold, the proportion of calls expected above it, the measurement point, and the exclusion for problems outside the provider's control. That definition can be tested and disputed concretely, which is what a service level is for.
Redstone Approach
Redstone deployments expose per-call and per-direction quality data from the platform, and our monitoring guidance recommends thresholds grouped by site and by trunk rather than a single enterprise figure. Where quality is disputed, we work from endpoint-reported metrics and call detail together, so that the conclusion rests on more than one instrument.
Redstone Systems, Inc. founded in Delaware, USA in December 2002, has been the ODM vendor for many well-known communications companies, serving the Southeast Asian market. In 2020, Redstone Systems will begin to return to the North America market with its self-developed brand.
Redstone has a complete product line of intelligent voice gateways, providing IP-PBXs, analog VoIP gateways (FXS/FXO), digital VoIP gateways (E1/T1), border appliances, and session boundary controllers (SBCs).
With advanced technology in digital signal processor (DSP), speech coding and speech processing, as well as efficient operational tools such as cloud remote management, auto provisioning, Redstone gateways are widely used in markets of enterprise communications, cloud communications, call centers, operators’ IMS/SIP trunks, bringing users friendly, efficient and reliable communication experience.
For more information about Redstone, You can follow us on Facebook, Linkedin, and Youtube to be the first to get the latest news.
