Transcoding of MPEG Bitstreams

Gertjan Keesman, Robert Hellinghuizen, F.W. Hoeksema, G.H.L.M. Heideman

    Research output: Contribution to journalArticleAcademicpeer-review

    163 Citations (Scopus)
    183 Downloads (Pure)


    This paper discusses the problem of transcoding as it may occur in, for instance, the following situation. Suppose a satellite transmits an MPEG-compressed video signal at say 9 Mbit/s. This signal must be relayed at a cable head end. However, since the cable capacity is only limited, the cable head end will want to relay this incoming signal at a lower bit-rate of, say, 5 Mbit/s. The problem is how to convert a compressed video signal of a given bit-rate into a compressed video signal of a lower bit-rate. The specific transcoding problem discussed in this paper is referred to as bit-rate conversion. Basically, a transcoder used for such a purpose will consist of a cascaded decoder and encoder. It is shown in the paper that the complexity of this combination can be significantly reduced. The paper also investigates the loss of picture quality that may be expected when a transcoder is in the transmission chain. The loss of quality as compared to that resulting in the case of transmission without a transcoder is studied by means of computations using simplified models of the transmission chains and by means of using computer simulations of the complete transmission chain. It will be shown that the presence of two quantizers, i.e. cascaded quantization, in the transmission chain is the main cause of extra losses, and it will be shown that the losses in terms of SNR will be some 0.5 ¿ 1.0 dB greater than in the case of a transmission chain without a transcoder.
    Original languageUndefined
    Pages (from-to)480-500
    Number of pages21
    JournalSignal processing
    Issue number6
    Publication statusPublished - 1996


    • METIS-111818
    • IR-14849

    Cite this