Related Works

In [29], the authors study the effect of both loss and jitter on the perceptual quality of video. They argue that, if there is no mechanism to mask the effect of jitter, the perceived quality degrades in the same way as it degrades with losses. In [134,58,75,146,147,130], the effect of audio synchronization on the perceived video quality is analyzed (for instance, by quantifying the benefits of audio synchronization on the overall quality). The main goal of [53] is to study the effect of the FR for different standard video sequences on the overall perceived quality. A related work is [119] where the effect of FEC and FR on the quality is the subject of the study. The work presented in [23] is a study of LR effects on MPEG video streams. The authors consider the effect of LR on the different types of MPEG frames. In [153] and [102], a study of the effect of BR on the objective quality metrics (PSNR, NVFM, and MPQM) is presented. The effect of CLP on video quality is analyzed in [58]. The authors of [82] study the effect of packet size and the distribution of I-frames in the layered video transmission over IP networks. In [24] the analysis goes deeper: the authors present a study of the effect of motion on the perceived video quality. In [61], the effect of both LR and PI on speech quality is presented. The study is based on an automatic speech recognition system which is based on hidden Markov models instead of the usual subjective quality tests. The effect of LR on three speech codecs is evaluated in [28]. The codecs under consideration are Adaptive Differential Pulse Coded Modulation (ADPCM), Sub-Band Coding (SBC), and Adaptive Predictive Coding (APC). The authors concluded that there are two effects due to lost packets. First missing of fragments of the speech. Second, the adaptation logic is disrupted after the lost segment (in the decoding part). Pure delay effects on speech quality in telecommunications are evoked in [80]. It is mainly for conversation sessions. They argue that round-trip delay in the range of 500ms is annoying. In [146] the authors studied the variation of speech quality for different ways of error concealment (silence, waveform, or LPC repair). Additionally, they evaluated subjective speech quality against loss and they compared the result for redundancy and no redundancy cases. The impact of cell delay varition on speech quality in ATM networks is given in [86].