scope:

The present document provides an instrumental model for the signal-based prediction of echo artefacts that may occur in telecommunication devices of all kinds. This prediction model assesses audible degradations of a virtual end-to-end communication scenario, including partial masking via sidetone as described in ETSI TS 103 801 [1]. The predicted Mean Opinion Score (MOS) corresponds to the Degradation Category Rating (DCR) according to Recommendation ITU-T P.800 [i.1].

The prediction model described in the present document is applicable to handset, headset or hands-free terminals and is not limited to a certain audio bandwidth, i.e. applications from narrowband to fullband are supported.

The following use cases are out of scope for the prediction model described in the present document:

• Ambient noise at the device-side (near-end).

• Ambient noise at the reference-side (far-end).

• Talker at near-end (double talk).

• Network impairments other than constant delay.

The prediction model assumes monaural listening at the reference-side under silent condition and high-quality handsets or headsets with low leakage. Assessment of binaurally perceived echo artefacts at the reference-side is out of scope of the present document.