Home > Research and Market Data > Technology research > Research > Enhancing Spectrum Efficiency > Future Performance of Video Codecs

1. Purpose

The objective of this study is to establish the likely future efficiency gains in video compression technology. The study should result in a summary forecast of video codec efficiency for the likely major applications such as HD, SD and mobile-type definition. The forecast should be supported by clear explanatory text setting out the assumptions underlying the predictions.

This study should provide a high-quality insight from an expert viewpoint into the issue of video codec efficiency, of direct use to Ofcom in supporting its duties for efficient management of the radio spectrum, and also to the wider public domain and broadcast community.

2. Background

Ofcom has in interest in codec performance to understand the impact on spectral efficiency in wireless communications, for example enabling input to predictions in demand for radio-spectrum for video purposes. Whilst our interest is in all types, we are particularly interested in video codecs since potentially significant efficiency gains remain in this area. Video is also an application where there is potential for increased demand for radio spectrum to support wireless traffic.

Improvement in the efficiency with which a video signal can by communicated can arise from two principal aspects:

1. The genuine improvement in compression efficiency that results from advances in technology over time – for example the progression from DPCM to transform coding systems, and the introduction of inter-frame motion compensation to allow advantage to be taken of the correlation between successive video pictures.
2. Differences in the acceptability of video compression artefacts between different applications.

For example, for production interchange between studio centres the tolerable level of artefacts may be very low if the headroom for subsequent studio processing was not to be significantly impaired. Typically, for such an application, the user might be looking for an impairment level that would be judged as “imperceptible” or “just perceptible” in formal subjective testing, implying high bit-rates of perhaps 20 to 25 Mb/s.

In contrast, the level of artefacts that can be tolerated in the broadcasting of digital signals to the public is significantly higher, and may be different for different broadcast applications such as home video, versus mobile video. For example, many of the signals received on the average digital multiplex at statistically multiplexed bit-rates in the region of 2.5 Mb/s might have impairments that could be judged under similar circumstances as “perceptible but not annoying” or even “slightly annoying”. This difference of up to 10:1 can easily dominate the benefits to be gained from different compression technologies.

This study should address both aspects in forming conclusions regarding future codec efficiency.

Initial contact was made with the BBC R&D department regarding this study, who suggested they have a number of published papers which could be useful. Collaboration with this department by the consultants who undertake the study is encouraged (on the assumption that the BBC is happy to participate).

3. Description of Work Items

3.1. Essential Work Items

• A short collation of the key literature, collating the evidence regards current state of the art, likely trends and other relevant published information which will provide key inputs to the forecast exercise.
• Interpretation and distillation of the key factors and assumptions which will form the basis of forecasting video codec efficiency over a 10 year interval and a 20 year interval.
• A forecast of video codec efficiency against time for the likely major applications such as HD, SD and mobile-type definition. The forecast should be supported by clear explanatory text setting out the assumptions underlying the predictions. Predictions should cover a 10 year and 20 year forecast.

• Future Performance of Video Codecs [pdf]

Back to top