In today’s ever-changing digital landscape, video quality drives viewer engagement and user experience. People want to enjoy video anywhere, anytime and this drives demand for higher quality content, improved bandwidth and faster processing. And with that, video codec quality, performance and reliability have never been more important.

Nicolai Otto_Headshot

MainConcept director product management Nicolai Otto

Today’s video landscape
Today, AVC/H.264 video dominates the professional video market due to its rich feature-set and maturity. It is used in a wide variety of industries ranging from medical and surveillance, physical discs to broadcast, and consumer video delivery via cable, satellite and internet.

However, AVC/H.264 is reaching a turning point where compression efficiency and speed can no longer be significantly improved. As such, many vendors of AVC/H.264 solutions are investing in pre-processing or layered compression add-ons to keep up with customer demand for premier visual quality.

In order to meet these demands, several new video codecs have entered the market, such as HEVC/H.265, AV1 and VVC; each promising to deliver superior quality with unprecedented performance. But the question is, are any of these emerging technologies ready to take the torch from AVC?

Achievements of HEVC
HEVC/H.265 is ITU’s official successor to AVC/H.264. And despite a bumpy start due to ambiguous licensing terms, we can consider 2019 to be “Generation HEVC”. HEVC is supported in most consumer devices today. OTT services, conferencing and surveillance systems are all switching to HEVC formats. In addition, major broadcast companies now extend their production workflows with HEVC.

In fact, Japan is already pushing 8K/UHD live production in preparation for their hosting of the 2020 Summer Olympics. Viewers will be immersed in ultra-high-definition live content, on nearly any screen from large scale video walls to mobile devices. Consumers have the ability to enjoy their favorite content in 4K HDR quality in more ways than ever before. On the other side of the spectrum, HEVC/H.265 enables professional-level video conferences - even over limited connections - and delivers sharp live content in high resolution.

Raising the bar with HEVC and GPU
Each new codec generation is tasked with increasing compression efficiency to deliver the same visual quality, while using less bits to describe the picture. This, of course, increases the amount of work a video encoder has to perform, and it consumes more computational power. In order to compensate for this requirement, it seems logical to tap into a mostly under-utilised resource: the GPU. The massively parallelised computing power of graphics cards seems predestined for encoding tasks. Repetitive video frames, which are easily dividable into smaller chunks (i.e. macroblocks), seem like a natural fit for such parallelisation.

Except that the implementation is not as easy as it may seem. All MPEG-based codecs (including AVC/H.264, HEVC/H.265 and VVC) make use of similarities between consecutive pictures. This means each encoding block needs to know about subsequent or previous images, as well as neighboring blocks. So, instead of simply chopping up an image into small chunks and sending the chunks to multiple parallel processing units, there must be a logic that keeps track of bitrate, temporal and spatial prediction and controls the encoding instances. Any shortcut taken here will invariably lead to image quality degradation, and any over-engineering in the control mechanism will sacrifice the benefit of parallel processing.

Thomas Kramer_Headshot

MainConcept VP product management Thomas Kramer

The solution to this challenge is using the best of both worlds - software and hardware - and it’s called “hybrid”. A smart codec implementation will split its processing pipeline into separate modules with ultra-fast data transfer connections. While encoding decision management and rate control is most effectively done in software on the host CPU, the actual picture encoding is running on the GPU. Offloading the heavy lifting to the GPU frees the host to enable pre-processing subsequent frames. Good Hybrid GPU encoding on standard server hardware can balance between CPU and GPU processing while still maintaining the expected high quality.

Is Hybrid GPU Encoding the ultimate answer? As every use case is different, there is no single best answer. However, Hybrid GPU accelerated encoding delivers faster processing, allowing for more live channels per server, with less demand for CPU power when encoding. But, the performance of the encoder depends on a variety of factors such as the desired encoding profile, resolution, server system, GPU model, etc. In today’s world of ever-increasing quality and performance requirements, especially for live video, GPU Hybrid offers flexibility to get the most out of your investment.

Driving innovation
The benefits and rising opportunities of CPU versus GPU encoding equip the market with the potential to innovate with new services and offerings around high-quality video, without the need for costly investments in hardware. Existing installations will be able to deliver more live streams per server. Furthermore, 8K live encoding is possible using standard server hardware.

What’s next?
HEVC/H.265 sets the stage for an era of immersive high-quality content viewing on any device from nearly anywhere in the world. As media heavyweights, like Netflix and Amazon, continue to invest in original content programming on a massive scale, and in UHD quality, consumers will experience episodic and live content in impeccable quality. The impact of HEVC on emerging technologies, like VVC and AV1, will undoubtedly set a high quality and performance bar for these next-generation codecs, which will continue to feed consumer appetites for new viewing experiences delivered without delay and in superior quality.