Blog

Behind the Science: H.264/H.265 Encoding Solutions

As video technology continues to evolve to newer digital formats and high definition video transitions into 4K and 8K resolution, video compression (or video encoding) remains an essential technology for applications such as digital television, disk formats like DVD and Blu-Ray, mobile TV, video conferencing, internet video streaming, and video transmission for military and commercial security and ISR (intelligence, surveillance, and reconnaissance) systems. Everyday video consumption like watching live sports or streaming Netflix wouldn’t be possible without the codec technology that is used today. 

The Basics of Encoding and Decoding

Encoding is another word for compressing, or reducing the size of the data making it easier to transmit or disseminate to multiple endpoints of a network. An encoder converts video into a compressed format and a decoder converts compressed video back into an uncompressed format. The encoder and decoder each rely on each other to function and therefore are often packaged together as a single package called a Codec.  

H.264/H.265

H.264 (AVC) and H.265 (HEVC) are Codecs designed specifically to compress data for video streaming and/or recording. H.264 (also known as MPEG-4 AVC) is the current industry standard for video compression and has been well vetted over time. H.264 has been around for well over a decade resulting in many available H.264 codec implementations and most running highly efficiently. 

While H.264 encoding is efficient for HD streaming, it does not have the bandwidth for 4k and 8k streaming. Most H.264 codec implementations max out at Level 5.1 (~240Mbits/sec) resulting in no 8k support and 4k support only at 30fps. While some implementations have started supporting higher levels (levels that support 8k and 4k60 resolutions), most are not intended to be used for live sources and can not keep up with the throughput in real-time.

H.265 (HEVC) is a more modern and advanced codec, offering higher compression and hence reduced output size, utilizing less bandwidth for streaming. H.265 utilizes a more complex algorithm that can target at a finer grain what areas to compress and what areas to leave untouched. In addition, H.265 is designed from the ground up to be more effective at 4K and 8K, and also adaptable for future higher resolutions. Because H.265 compresses data so much more effectively, it can drop your bandwidth and storage requirements by roughly 50% with no impact on video quality compared to H.264. 

What’s the Science Behind it? 

H.264 and H.265 analyze the motion, color, and entropy of a video stream to determine what information or areas require more or less visual fidelity. This encoding methodology deduces which areas will not change from frame to frame and acts to compress accordingly. It can also predict the motion of future frames to determine where to more effectively compress data. Essentially, these codecs reduce the number of full frames, known as “keyframes”, that are stored while only storing “change” data for the intermittent frames. H.264 and H.265 help to balance the visual quality of the information to appropriate output sizes, for cost-effective bitrate management.

High-end Encoding Solutions

Most software implementations are great for pre-recorded data, where latency doesn’t matter since it will be viewed at a later date. In situations where the captured video must be displayed in real time, the encoding hardware needs to be “high-end”, as it demands extremely low latency. In defense, security, C4ISR and surveillance applications, with increasing camera resolutions,  the amount of data that is collected and processed grows significantly. The encoder must be able to keep up, in real-time, with the raw video data coming in. Using HEVC/H.265 hardware-based encoders is the optimal solution for data-intensive applications. 

One such hardware encoder that primarily used in the the C5ISR applications is the Tyton VS2X, a powerful HEVC/H.265 video encoding and streaming solution that is designed to encode raw data with exceptionally low latency and low power consumption. This stand-alone rugged system supports a latency of 60 ms or less.  The product has 8 dynamics encoding engines that can be configured to encode in H.265 or H.264 formats at various bit rates. The product can take 4 1080p60 (HD) video inputs simultaneously. Each of those inputs, under software control, can be assigned to one or more of the encoding engines such that each video stream cane be simultaneously be encoded as H.264 or H.265,  at a high bit rate or a low bit rate.

There will always be tradeoffs between size, power, quality, and speed. The lower the bit rate used for encoding,  the smaller the size and the worse the video will look. However, when higher quality encoding is desired,  higher bit rate must be used, which requires more processing resulting in a higher latency, size of data and power. There will always be some amount of latency. There is simply no way around it. Only a dedicated customizable hardware encoder optimized for high-efficiency with low power consumption can be used in intensive applications.e

Share
Facebook
Twitter
LinkedIn

View Latest Press Releases