It was the switch from analogue to digital broadcasts that allowed the industry to explore higher quality programming. Codecs were born as a result of this shift, and have significantly improved over the years to reduce bitrates to levels previously unheard of. With 4K around the corner, and the massive amount of data it will entail, can HEVC save the day?
Codecs are software or hardware-based ‘compression/decompression’ algorithms, that can also be defined by ‘encode/decode’. There are many different types of codecs and they vary based on the purpose they serve. There are codecs for video, audio and even for data.
The purpose of a codec is to encode or decode a particular stream of media into a format that is compressed and ideal for transmission, reception and storage.
Keith Wymbs, VP, marketing, at Elemental Technologies says: “Video compression seeks to reduce or remove redundant information from a video stream so that assets can be stored or sent over a network as efficiently as possible.
The algorithms used to eliminate data make up the process of encoding and the method used to play back the compressed asset and return it as closely as possible to its original state is known as decoding. The interoperability of the compression and decompression processes forms the basis of a codec.”
Some common codecs that have been used in the past include MPEG-2 and H.264. Most recently the codec that is being tested and is seen to be the successor to H.264 is HEVC. This codec is expected to reduce the average bit rate by 50% compared to H.264, and is hence touted as the perfect codec for data intensive 4K files.
The need for compression is growing especially with a boost in different technologies that will continue to be filmed and broadcast in qualities surpassing the current levels of HD.
Stan Moote, VP of business development at Harris Broadcast says: “Without having compression the broadcast industry would still have analogue distribution. It was the switch to digital compression technologies that made the million-channel universe available.
Compression is used in various areas of video transmission and storage. For example – a backhaul takes video from a stadium venue back to the main plant. Backhaul compression needs to maintain high quality and also not have too much delay (latency), so these types of compressors do not have a high compression ratio.
Distribution of video to cable/satellite headends also needs to be of higher quality, but not as high as from the original stadium and delay is less of an issue. The final distribution can be a much higher compression ratio.
Article continues on next page ...
“The key area of concern is cascading compression. The more compression/decompression steps, the video quality goes down. Keeping at a low compression ratio until final distribution helps to maintain quality,” continues Moote.
So how have standards improved in the last few years? Ian Trow, senior director of emerging technology and strategy at Harmonic says: “Every iteration of MPEG/ISO standards from MPEG-2 through MPEG-4 AVC (H.264) to HEVC has the principle goal of delivering a 50% compression improvement over the predecessor standard.
Additional objectives have existed for each new standard beyond outright reduction in bit rate. For MPEG-4 AVC it was to enable video streaming for VoD, catch-up and mobile applications common in mobile, multiscreen and connected TV environments.
For HEVC it is to enable Codecs to be realised on standard processor platforms instead of bespoke hardware platforms.
Additionally HEVC was conceived to cater for resolutions up to and including Ultra HD (both near 4K and 8K resolutions are covered by the standard). That said, HEVC has compression gains for all platforms on which video is viewed from mobile all the way up-to Ultra HD.”
Industry pundits insist that video compression is virtually lossless. Technology has advanced to address quality, according to Wymbs. He says the H.264 codec was architected to deliver comparable quality at half the file size. H.264 supports a broad range of applications - from low bit-rate Internet streaming applications to HDTV broadcast and digital cinema - with nearly lossless coding.
A codec must be efficient, and as more codecs come about, the need for efficiency is high up on the agenda.
Trow explains: “In traditional linear broadcast applications the greater the compression scheme the cheaper the rate card for transmission bandwidth. Usually this equates to less bandwidth per channel to enable more channels to be offered to the viewer within a program bouquet.
In certain markets, like OTT, there is a need to increase the Quality of Experience (QoE) to match that of traditional broadcast television. Increasing the QoE is as much a network issue as it is a video quality issue.
Any standard that can improve the quality and reduce the network congestion associated with the carriage of IP video is welcomed by operators keen to manage OPEX.”
Article continues on next page ...
Moote agrees and says that better codecs give fewer artifacts at higher compression ratios, which result in a fewer number of bits being transmitted or stored. With the advent of 4K, the industry will be dealing with files four times the size of HD, which are already quite large. Hopes are pinned on HEVC.
Wymbs says: “With more prolific 1080p HD resolution content and 4K/UltraHD on the way, the bar is rising for compression codecs. 4K Ultra HD images have four times more pixels than full HD images.
Today, 4K is simply bandwidth prohibitive for distribution networks at 18Mbps to 20Mbps required for H.264 compressed 4K content. HEVC shrinks bitrates required for 4K resolution potentially to under 10Mbps, making UHD much more accessible within current network bandwidth limitations.”
According to Trow, initially due to the complexity of algorithms, codecs were almost exclusively hardware based.
He says: “The advent of faster processors, large memory capacities and better porting of algorithms to commercially available processor platforms has led to codecs implemented in software for both live and off-line applications. With the increasing demand for VoD content much of the encoding for multi-screen applications is performed offline on server farms.”
Moote from Harris explains how the manufacturer’s Selenio platform is capable of compressing many different video signals. He says there are different bit-rates for various types of distribution, and there are also different types of redundancy strategies to assure on-air reliability.
“To understand how video compressors work is very detailed and complex, however it can be simply stated – video compressors focus on removing information that is repeated, such as repeated information video frame to video frame in areas where there are no moving objects.
Additionally video information that is not easily viewed is thrown away. An example of this is reducing the bandwidth of the colour information as our eyes are not so sensitive to observe colour details – the luminance component is noticeable and the colour is just essentially an added detail.
For final distribution some of the high frequency information is also removed. The key is to remove as much data as possible, yet still assure the video has a good representation of the original,” explains Moote.
All in all, codecs are created to harness the power of an evolving broadcast industry. “The introduction of a new compression standard is particularly welcome for broadcasters and network providers alike, but only when it can be adopted.
Many broadcasters have just completed HD refreshes based on MPEG-4 AVC and so are unlikely to undertake adoption of a new compression scheme like HEVC in the short-term. SD distribution applications are dominated by MPEG-2 which is firmly established in viewers’ homes and as such is unlikely to be replaced in the short-term.
A huge amount of inertia behind these legacy standards is associated with hardware infrastructure.
Where new codecs like HEVC will achieve initial success is in applications which are either “Greenfield” applications like OTT, applications where programmable infrastructure is in place making a new codec a viable proposition or utilised in smartphones where the replacement cycle is reasonably quick and facilitates the adoption of the compression benefits available,” concludes Trow.