R-Factor Score
Transmission rating factor from the ITU-T E-model — a linear 0-100 scale that quantifies voice transmission quality and maps directly to MOS.
R-Factor Score
| Property | Value |
|---|---|
| Key | r_factor |
| Unit | 0-100 scale |
| Type | Gauge |
| Direction | Send and Receive |
| RFC | ITU-T G.107 |
What It Measures
R-Factor (also called the R-value or Transmission Rating Factor) rates voice transmission quality on a scale from 0 to 100. It is the raw output of the ITU-T E-model calculation before it gets converted to MOS.
Think of R-Factor as the "engineer's MOS." While MOS compresses quality into a narrow 1-5 range, R-Factor spreads it across 0-100, making it easier to see differences between two calls that might both round to the same MOS score.
The E-model works by starting at a perfect score and subtracting impairments. It accounts for codec quality, packet loss, jitter, delay, and equipment factors. Each impairment reduces the R-Factor by a specific amount, making the model additive and predictable.
Why It Matters
R-Factor has two advantages over MOS for technical analysis:
- Linear scale. The difference between R=80 and R=85 is the same magnitude as R=70 to R=75. MOS is nonlinear — a MOS change from 4.0 to 4.2 represents a different quality gap than 3.0 to 3.2.
- Wider dynamic range. Two calls with MOS 3.8 might have R-Factors of 72 and 78. The R-Factor reveals a meaningful difference that MOS hides.
For SLA contracts, R-Factor thresholds are often more precise than MOS thresholds because the linear scale makes compliance boundaries clearer.
| R-Factor | MOS Equivalent | Quality |
|---|---|---|
| 90 - 100 | 4.3 - 5.0 | Excellent |
| 80 - 90 | 4.0 - 4.3 | Good |
| 70 - 80 | 3.5 - 4.0 | Acceptable |
| 60 - 70 | 3.0 - 3.5 | Poor |
| Below 60 | Below 3.0 | Unacceptable |
How CallMeter Measures It
CallMeter computes R-Factor every second alongside MOS. Both come from the same E-model calculation — R-Factor is the intermediate result, and MOS is derived from it. The inputs are real-time jitter, packet loss, and round-trip time measured from the live RTP stream.
Thresholds
| Level | Value | Meaning |
|---|---|---|
| Good | 80 or above | No action needed |
| Warning | 70 - 80 | Investigate impairment sources |
| Critical | Below 60 | Immediate attention required |
What Causes Low R-Factor
Because R-Factor is an additive impairment model, you can identify which factor is consuming the most "budget":
- Codec impairment (Ie) — Narrowband codecs start with a lower ceiling. PCMA/PCMU have an inherent equipment impairment factor that wideband codecs like Opus avoid.
- Delay impairment (Id) — Round-trip time above 150ms increasingly penalizes the score.
- Loss impairment (Ie-eff) — Packet loss has the steepest impact. Even 1% loss can cost 10-15 R-Factor points depending on the codec.
- Jitter impact — High jitter forces larger jitter buffers, which adds delay, which increases the delay impairment factor.
How to Fix It
- Quantify each impairment. Compare your R-Factor against the codec's theoretical maximum to see how much headroom network conditions are consuming.
- Prioritize packet loss reduction. Loss has the highest per-unit impact on R-Factor.
- Reduce end-to-end delay. Move workers closer to the SIP infrastructure or reduce intermediate hops.
- Consider codec upgrades. Switching from G.711 to Opus can recover 10-15 R-Factor points under identical network conditions.
Related Metrics
- MOS Score — Derived from R-Factor, used for non-technical communication
- Jitter — Contributes to delay impairment in the E-model
- Packet Loss Rate — Highest-impact impairment factor
- Round Trip Time — Contributes to delay impairment