The Need for Advanced Video Transcoding

Contributed by the widespread use of smartphones and tablets, the technological advancements of video streaming and mobile gadgets have rapidly transformed the way people consume their entertainments. The market of subscription VOD (Video-on-Demand) and live streaming on social media has shown a continuous rise in terms of the historical high number of subscribers and viewers. It is reported that there are more people watching their subscribed videos on smartphones, laptops and mobile tablets than living room TVs. Therefore, VOD service providers have to make their bit rate streaming adaptive to various network conditions and protocols, as well as the bandwidth for the client devices, in order to enable high-quality video contents without compromise. In addition, compression has to be performed “intelligently” in order to trim the data involved while maintaining the viewing quality. Therefore, streaming and video transcoding have to re-programmed with artificial intelligence.

A.I. in Video Transcoding – Dynamic Adaptive Streaming

A suggested method is to implement artificial intelligence in the transcoding process that can boost streaming and compression performance based on the currently used ABR (Adaptive BitRate) technique. The legacy ABR was a decent algorithm to stream multimedia contents over computer network and adjust the video quality based on bandwidth, until the era of HEVC, and buffers and pixilation have been often encountered. Therefore, to reduce the stutters caused by frequent buffers, the advanced version, Dynamic Adaptive Streaming over HTTP (also known as MPEG-DASH) is introduced. This advancement implements artificial intelligence to select segments and intervals in the multimedia contents with highest bit rate possible that can be downloaded for playback without buffering. The result is, less data used during the streaming. The artificial intelligence makes streaming adaptive to changing network conditions, such as bandwidth changes, so that video quality can remain without frequent rebuffering.

As discussed, Dynamic Adaptive Streaming over HTTP enables high quality streaming of multimedia contents over the Internet and it has become widely adopted by leading VOD service providers. Indeed, as subscribers are demanding HEVC standards for what they consume, service providers have to offer high-quality streamed videos with less data in the process in order to perform sophisticated compression without stalls.

Besides VOD, Dynamic Adaptive Streaming are also applied for OTT services, live broadcast (sports and news events for instance), and pay TV services.

Recommended Hardware for Advanced Video Transport

To properly facilitate Dynamic Adaptive Streaming, it requires a top-notch, open architecture hardware platform that can perform live video transcoding and adaptive bit rate streaming without latencies. One example is Lanner’s NCA-6210. The powerful hardware offers future-proof performance made possible by dual Intel® Xeon® processors (6th Generation, with up to 56 cores), DDR4 memories up to 640GB, eight (8) modular expansion slots, and Intel® AVX-512 instruction/Hyperscan/DPDK technologies. This feature set can ensure MPEG-DASH running smooth on NCA-6210.

Since adaptive streaming is highly graphic intensive especially upon the delivery of 4K video contents, it is recommended to accompany NCA-6210 platform with the module NCS2-VT02A, a 4K video transport NIC module empowered by Intel® Xeon® E3-1515 v5 processor. NCS2-VT02A supports 4K ultra-HD resolution and H.265 compression. To further enhance MPEG-DASH capability, the module is built in with Intel® Iris Pro Graphics GT4e technology to reinforce graphic intensive applications in OTT, VOD and live broadcasts.

Featured Product


NCA-6210

High Performance Network Appliance for Network Security and Virtualization

CPU 2nd Gen Intel® Xeon® Processor Scalable Family (Cascade Lake)
Chipset Intel® C621/C627

Read more
 

NCS2-VT02A

4K Video Transport NIC Module Powered By Intel Xeon E3-1515 v5 CPU

CPU Quad-core Intel Xeon E3-1515 v5 CPU with 8MB cache
Chipset Intel CM236

Read more