Q-Model is inspired by E-model for audio. We are calculating aggregated distributions for several metrics and tune the performance based on objective and subjective (user feedback) quality. The Q-model returns a value of 1, 2, or 3, where 1 corresponds to bad, 2 to OK, and 3 to Excellent.
Articles in this section
- What is Automatic Diagnosis?
- What can I learn from the objective quality values?
- What is in my ICE Data table?
- What is objective quality and how is it measured?
- What are the transport metrics shown on the dashboard?
- How do you determine when a user drops out / rejoins a call?
- Can I find out which conferences are peer to peer and which ones go through a TURN server (i.e. are relayed)?
- What metrics does callstats.io collect?
- Explain the terms throughput, jitter, and latency?
- What is meant by churn?