Choosing the right broadcast encoder/decoder combination ensures both picture quality and the functionality to support the total broadcast system. Shown here is Scientific-Atlanta’s Continuum DVP D9030 MPEG-2 encoder.
The broadcast market for encoders in digital transmission has developed considerably since the introduction of linear video codecs at the start of the 1990s. It began with contribution and primary distribution systems for in-studio/production-to-studio applications and continued on to studio and transmitter locations. The efficiency of encoders increased dramatically during this time, and this has contributed to the success of new formats such as HDTV and compression standards such as H.264. Many players, however, have entered the encoding market in the meantime, making the selection of the right compression solution a complex decision.
The need for encoders in broadcast environments can be split into three areas: contribution, primary distribution and distribution-to-the-home.
Larger broadcasters need to transport content between studios or live events in multiple locations — for example, a DSNG satellite connection or a telecom network could be used for this kind of transport. In general, these types of application will always require both an encoder and a decoder.
The core requirements in contribution are usually high video quality and short-end-to-end delay. Due to the nature of MPEG-2, this means that a high bit rate is required, which is usually available at a reasonable cost on fixed networks based on fiber or microwave systems. The high cost of the satellite segment does, however, usually drive the bit rates down.
Another key requirement is 4:2:2 encoding in order to maintain a high color resolution for later editing. In addition, audio/video synchronization must be considered when looking at encoding and decoding. Just a 20ms loss of audio sync with video (lip sync) is equivalent to half a frame and will be visible to both professionals and most consumers; hence the requirement for less than a 5ms A/V lip sync difference.
Another important criterion is the quality of the analog video input. Analog composite video, for example, must first be digitized in order to allow it to be compressed. This A/D conversion is critical for the consequential video compression and overall quality, and the proper implementation should be accessible as an integrated part of the encoder.
Synchronization of the video frame rate within the studio is another topic that is easily solved by using genlock functionality. This is a feature on some studio decoders that allows the decoder to be locked to a studio reference for frame synchronization. More importantly, genlock functionality can solve problems when transmitting content using a synchronous digital hierarchy type of connection for video stability over fiber-optic cables.
Audio requirements for contribution are usually MPEG Layer II or uncompressed linear audio. For multichannel audio contribution such as surround sound, Dolby E compression can transport up to eight mono tracks in high quality, so a key feature for any encoder/decoder combination in such cases is the ability to adjust the delay of the audio transport compared to video.
Aspect ratio handling is an increasing issue for broadcasters. Currently, content is mainly created in 4:3 and 16:9. While most encoders support these formats, it is also important to support the signaling for these formats so that a mixer or recorder at the decoding site is able to detect which format is being used. This is usually done by generating video index information signaling, a studio standard that signals aspect ratio within an SDI stream. There is also a special technique to ensure error-free transmission in contribution networks, particularly for the path from decoder output to production facilities, which, for some broadcasters, is quite a distance. To help troubleshoot the system, it is important that the decoder generates error detection and handling information on the output to make it easy to verify if the error occurred during MPEG-2 transmission or after the MPEG-2 signal was decoded.
Broadcast primary distribution
Broadcasters today are distributing content either via direct-to-home (DTH) DVB-T (digital terrestrial) or by analog to the home, or a combination (simulcast). In primary distribution, the goal is to transport video, using compression, to the transmitter site. The decoded video is then fed to the analog transmitter. The requirement is typically for 4:2:0 compression, which should be either ready for, or upgradable to, full DVB-T functionality. Using open DVB-ASI interfaces at transmitter site locations is important. In DVB-T systems, it is usually only necessary to insert a decoder at the transmitter site and decode selected programs from the transport stream feeding these digital transmitters. Many of today’s analog systems are, however, working on either single or multiple programs per E3 link. Even so, the network interface of the decoder should be ASI, and a separate network adapter should convert PDH/ATM to ASI. The benefit is that this will enable multi-service network adapters, which besides video, could also carry radio distribution and IP-services, with the latter being used for transmitter site management, for example. The result is lower investment and reduced operational costs.
Broadcast distribution, satellite and terrestrial
Cost and availability of bandwidth for transmission-to-the-home systems are key parameters, while minimizing bandwidth consumption is vital. Conversely, every broadcaster or programmer puts great effort into the appearance of programming and wants to avoid any loss of quality during distribution to the consumer. So what are the critical encoder technologies to look for in these applications? Compared with typical contribution systems, encoders used for satellite and terrestrial broadcast distribution must be capable of statistical multiplexing, sharing bandwidth dynamically between each other. To optimize this process, an encoder requires an efficiently-working dual-pass system to provide the best video quality and simultaneously provide a low overhead data rate on the total multiplex output, ideally in the range of 1 percent. A critical part of the dual-pass algorithm must be active filtering of the video, allowing noise caused by analog transport, or grainy movies, to be removed so that the picture is sharp instead of blocky and with artifacts. This filtering approach also helps improve quality when encoding at low bit rates, even with source video that is perfect and fed directly by the studio mixer.
Again, aspect ratio handling is an important issue for satellite and terrestrial distribution. This can be solved by the active format descriptor (AFD). With the help of AFD, the encoder can signal to the set-top box (STB) how it will provide the optimum aspect ratio for the TV set connected to it. To this end, STB functionality is vital. If a broadcaster chooses to launch MHP applications, there would be a need to transmit time code along with the video. This can be done with vertical interval time code, which can be read by the STB.
Because many compression systems in a satellite uplink or DVB-T playout also need to distribute radio programs in quantities, a practical and cost-saving feature for any encoder or decoder is the ability to separately and simultaneously compress radio programs along with the TV service.
There are also new formats currently entering the market that offer broadcasters new applications, or more cost-effective ways of solving their current challenges. The emergence of H.264 encoders/decoders will drive transmission costs down in all bandwidth-restricted distribution systems. It will also enable new business such as contribution over E1 lines, which could be used for some broadcasts, or even surveillance, and could prove to be quite cost-effective. High definition will also require new investments by broadcasters, allowing native production in HD from the start, and later, enabling this content to be distributed.
If an installed base of MPEG-2 STBs is already in place, it will take a serious economic argument to switch the entire population to alternative, more cost-effective transmission formats. So new compression formats will mainly be used in greenfield applications such as xDSL-systems, new DVB-T systems and maybe HD broadcast via satellite. Regardless of the format, the need to get the right encoder and decoder combination is essential for all broadcasters, as it serves to provide them with greater functionality, reduced cost, better use of available bandwidth and higher quality video for the viewer. Some markets will benefit from the new compression formats, but the technologies building the application around the compression core will remain as important as ever.
Lars Lastein is the market manager for broadcast and satellite at Scientific-Atlanta Europe.