Information Technology Reference
In-Depth Information
success of Sony's PS3 game console which supports Blu-ray playback. As a con-
sequence of the demand for higher quality and higher resolutions, the transit to
television broadcasting in high definition (HD), with a resolution of at least 1280 x
720 pixels has started to take shape. This evolution requires fundamental changes
to existing file-based television recording, production and distribution processes.
Due to the increase in resolution, the bandwidth and storage capacity requirements
grow significantly. For the same reason, the computational resources needed for
operations such as format conversion and post-processing increase as well.
An entirely different, but equally important challenge is the selection of a suit-
able format for recording, production and distribution of HD material. While the
basic signal properties, such as resolution, frame-rate and scanning format (inter-
laced or progressive), are fixed for standard definition television, this is not the
case for high definition television. A full HD television signal consists of 50,
59.94 or 60 progressive frames per second, each frame having a resolution of 1920
by 1080 pixels; this format is typically denoted as 1080p50/59.94/60. However,
the bandwidth and storage requirements associated with full HD production and
broadcasting exceed the current capacity of distribution networks and off-the-shelf
media production and storage systems. To solve this problem, two different alter-
natives have been put forward: 720p50/59.94/60 (1280 x 720 pixels per frame, 50,
59.94 or 60 progressive frames per second) or 1080i25/29.97/30 (1920 x 1080
pixels per interlaced frame, 25, 29.97 or 30 interlaced frames per second). In sec-
tion one of this chapter, the benefits and drawbacks of both formats will be dis-
cussed in detail.
For digital production of SD content, relatively simple, intra-only video com-
pression schemes such as DV25 [1, 2] and D-10 [3] were used. Since such
standards are exclusively defined for SD material, new compression technology
suitable for HD television recording and production must be selected.
For HD acquisition and production, the European Broadcasting Union (EBU)
recommends to not subsample the luma component horizontally nor vertically,
and to limit chroma subsampling to 4:2:2 [4]. For mainstream production a bit-
depth of 8 bits per component is advised and for high-end productions a bit depth
of 10 bits per component is suggested [4]. These requirements cannot be met by
currently used HD compression formats such as HDCAM and DVCProHD, since
both formats apply horizontal sub-sampling of the material prior to compression.
Therefore, the transition to more advanced compression techniques is warranted.
The final selection of a suitable compression format should take into account sev-
eral different and sometimes contradictory requirements. First of all, to minimize
the impact of the transition to HD on the bandwidth and storage requirements, the
selected compression scheme should deliver state-of-the-art compression perform-
ance. Secondly, the quality degradation associated with recompression (decoding
the compressed material, editing it, and recompressing the modified result), which
can occur multiple times in a typical production chain, should be minimized.
Moreover, to reduce the number of recompression cycles, the recording format
and the production format should ideally be one and the same. Additionally, for
optimal edit-friendliness and frame-by-frame random access capability, so-called
long-GOP formats, which make use of temporal motion-compensated prediction,
