ITU-T P.1203.2 PDF

This Recommendation describes the short term audio quality estimation module which is an integral part of the ITU-T P.1203 series. [ITU-T P.1203] describes a set of objective parametric quality assessment modules. Combined, these modules can be used to predict the impact of audio and video media encodings as well as Internet protocol (IP) network impairments on the quality experienced by an end-user of multi-media streaming applications.

The addressed streaming techniques comprise progressive download as well as adaptive streaming, for both mobile and fixed network streaming applications over transport control protocol (TCP) or other TCP like protocols which are not affected by transmission errors.

The model described is restricted to information provided to it by an appropriate packet- or bitstream-analysis module. The overall ITU-T P.1203 model is applicable for the effects due to audio- and video-coding as well as initial buffering and re-buffering (which are both perceivable as stalling of the media) as the typical degradations associated with progressive download. As final output, the ITU-T P.1203 series models target integral audio-visual media quality scores.

This Recommendation describes only one audio quality module. With regard to the required input data, this audio module corresponds to mode 0 of [ITU-T P.1203].

The same, purely header-based/bitrate-based audio quality module is also specified in [ITU-T P.1201.2]. Using a large number of subjective experiments, it was validated that this model also leads to accurate predictions within the scope of [ITU-T P.1203].

The audio module predicts mean opinion scores (MOS) on a 5-point absolute category rating (ACR) scale (see [ITU-T P.910]) as a per-one-second MOS score.

During the development of [ITU-T P.1201], explicit short-term audio quality tests were carried out in order to validate the stand-alone use of the audio module for the estimation of audio-only quality. It could be shown within the scope of [ITU-T P.1201] that this is possible.

It must be noted however, that since the subjective tests conducted for [ITU-T P.1201] included packet loss degradations, some range-equalization and other bias considerations may need to be considered (see for example [b-Zielinski_2008]) if the module is to be used stand-alone within the scope of [ITU-T P.1203].

This model cannot provide a comprehensive evaluation of audio transmission quality as perceived by an individual end user because its scores reflect the impairments due to audio coding only. Furthermore, the scores predicted by a parametric model necessarily reflect an average perceptual impairment. Note also that the model was developed and validated for one specific encoder and decoder implementation. If a different encoder and decoder pair is used in a monitoring situation the scores may not reflect that.

Effects such as audio level or noise (and corresponding similar audio factors) or other impairments related to the audio signals are not reflected in the scores computed by this model. Moreover, the scores predicted by a parametric model (i.e., without access to payload information, such as the audio signals) necessarily reflect a somewhat simplified representation of the perceptual impairment of the considered stream

However, presuming that it is applied in an appropriate manner, according to this Recommendation, the model still enables estimation of some coding quality related information and thus valid and in most cases accurate predictions.