Available for download: New edition of ITU-T P.863 voice quality assessment
ITU-T has published a new edition of Recommendation ITU-T P.863, a standardized algorithm for the assessment of voice quality in fixed, mobile and IP-based networks. The revision reflects deployment best practices identified by the standard’s implementers.
Recommendation ITU-T P.863 “Perceptual objective listening quality assessment” was first released in 2011 to meet new demands on voice-quality analysis arising from the rollout of technologies such as HD Voice, 3G, VoLTE (4G) and VoIP. In particular, the standard was developed to enable the assessment of superwideband speech in addition to narrowband.
On average, the voice-quality prediction scores produced by the new edition of ITU-T P.863 remain close to the values of its predecessor, however, significant differences will occur for certain individual measurements, especially for cases where the previous version failed. ITU-T P.863 now includes an updated Appendix III (ex. Amendment I of ITU-T P.863), which describes the standard’s application to narrowband acoustical measurements.
ITU-T P.863 describes an objective method for predicting overall speech quality as perceived by the listener, from narrowband (300 to 3’400 Hz) to superwideband (50 to 14’000 Hz) telecommunication scenarios. The standard presents a high-level description of the method, advice on its application and a sample of results from a five-year benchmarking exercise. Annex B provides all essential parts of the model, and Annex A contains conformance-testing data to validate the accuracy of alternative implementations of the model.
For more information on ITU-T’s standardization work on performance, quality of service (QoS) and quality of experience (QoE), see ITU-T Study Group 12.