What is multi-channel encoding?

Using multi-channel encoders and/or streaming media servers provides multiple advantages.

Since different video streaming protocols deal with different problems, it makes sense that multiple different protocols are sometimes required to get video from media sources, like cameras, all the way to many simultaneous consumption nodes like smartphones, tablets, PCs, media players, and game consoles, and over very large distances to a disparate base of viewers. This often necessitates the use of cloud services or the public internet.

For example: “continuous” streaming protocols, like RTMP, can help maintain certain aspects of video performance while minimizing latency.

HTTP-based protocols, like HLS and MPEG-DASH, package video streams into fragments to better borrow the massive interoperability of networks and software applications by behaving like all other network traffic. They rely on TCP transmission to provide error correction, and on HTTP to traverse firewalls without requiring special instructions. However, these protocols require huge amounts of buffering to make this all work which injects significant latency. These solutions are perfectly acceptable for on-demand streaming workflows. But the market is working very hard to continue to compress latency for live streaming applications.

So having multiple protocols and the ability to change protocols for different segments of your workflow allows you to maximize both reach and performance by providing you with the ability to have some more advanced nodes capable of maintaining a low latency and very high video performance while also assuring that everything else is compatible with your streaming delivery setup.

This applies at a local level just the same as it does over the internet.

At a local level, an encoder running on a decent network can feed directly into a decoder and provide high-resolution video with minimal latency. If it is a multi-channel encoder, the same encoder can provide additional streams that work with standard players and browsers on lower bandwidth parts of the network, including wireless devices. Whether or not your encoder supports multi-channel encoding, it is also possible to use a streaming media server on your network to multiply the streams and/or change the protocols to suit your applications.

Some manufacturers of encoders also provide hardware and/or software decoders–minimizing complexity to have everything work together seamlessly.

“Recording” for on-demand streaming can also be fairly mission-critical in order to avoid losing a keynote speech or important moment during a network interruption. Sometimes multi-channel encoders and/or encoder and streaming media server combinations provide a local cache of what’s being recorded while simultaneously recording on cloud services. Or recording and simultaneously live streaming captured video sources may be the desired application. Here too, different protocols may be called into service such as FTP for an MPEG-4 file recording and a live RTMP H.264 stream.

The same applies to cloud/internet where-by multi-channel encoding enables the use of the right protocols for the right segments of the video streaming workflow.

By leveraging the appropriate protocols it is possible to have a mix of very high-performance nodes and very easy-to-access nodes. Protocol flexibility also allows you to mix old/legacy compute equipment with much more modern equipment. This means it is possible to pursue continuous improvement and evolution of your video streaming infrastructure instead of requiring large overhauls and revolution of your infrastructure.

Many cloud streaming architectures currently use a low-latency protocol, such as RTMP from the video source to the cloud and use more broadly compatible HTTP-based protocols for mass distribution.

The multi-channel load can be placed on the encoder or the streaming media servers being used or a combination of the two.

Another instance of a change in protocol is when streaming from a stored file rather than a live source. A perfect example of this is Video on Demand (VOD) services. These providers must store the video content in a container, and when a user initiates a viewing session it then converts it from a stored file to a video stream which is sent to the viewer over the internet. This could be handled by a multi-channel encoder or streaming media server. The protocol that they will use to communicate with the viewing device (such as a SmartTV) will help inform them of the bandwidth availability and reliability of the network, which allows them to select the appropriate resolution stream to create/send from the stored file.

One of the most important variables that affects the bitrate of live streams is the resolution of the video being streamed. Multi-channel encoding deals with this problem as well.

Delivering streaming video is a balancing act between visual acuity and stability of the stream. In the early days of watching videos from the internet, users often experienced the frustration of buffering. Many videos were simply un-watchable.

Significant progress has been made to deliver optimal experiences that account for how much bandwidth is available and how much information can be carried in the video streaming payload to each node. (Higher resolutions require more information.)

Today, adaptive bitrate streaming technology automatically detects users’ bandwidth and computer processing availability in real time and provides a media stream that fits within these constraints.

Transcoding in the cloud is something that creates latency and requires paid-for services. For this reason, many organizations that generate a lot of private (corporate) video content are balancing the load by either sending multiple different resolutions from multi-channel encoders from each captured video source in their organization, or using adaptive bitrate encoders that can be leveraged by certain compatible multimedia players that have the ability to switch between the different bitrate segments and offer the maximum quality (often includes resolution) that optimally suits the compute power and network conditions of that player node.

In enterprise and media and entertainment encoding, this basically means that video sources are often sent at their maximum quality and resolution profile but the local encoder and/or streaming server also create additional stream copies of the source in reduced settings.

This “scaling” of video sources through multiplication of the streaming video profiles is very useful to instantly accommodate all destination types. A 4K source, for example, can be kept in 4K and decoded at an appropriately powered viewing node. But the same 4K source can comfortably supply the same source content onto tablets and smartphones. These devices often have a lower resolution screen anyway and the corresponding reduced resolution stream is served to match what the wireless network and processing power of these wireless devices can handle.

One of the most important variables that affects the bitrate of live streams is the resolution of the video being streamed.

There are different motivations for changing the codec of your video assets.

Here is a simple example:

Assume an organization has added new equipment capable of generating very high resolution, such as 4K. When these new assets are captured at full resolution, using codecs that produce a small-enough bandwidth might be enticing. But the codec and/or encoding profile used directly from the source to mitigate its bandwidth use may not match what is the optimal codec or encoding profile for content distribution at large.

Using HEVC (H.265) to encode 4K content may appear to shave off some bandwidth and help assure the stability of the stream from its capture point to its stream re-distribution point on a network or on the internet. But HEVC tends to drain battery on handhelds more than H.264 and many older devices do not have hardware implementations of HEVC. Media servers and other tools are therefore still extensively used to turn new HEVC sources into more convenient H.264 streams for many applications.

Conversely, some installations have legacy MPEG-2 sources. In this case, a transcoding effort could mitigate distribution bandwidth ‘and’ augment downstream device compatibility.

It should be noted that there is a big gap in performance between encoders. Some highly-optimized H.264 encoders can produce bitrates that are superior to some early or basic HEVC encoders. The same applies for other encoding performance metrics such as latency or image quality.

But over time there are transitions in the market for resolutions and codecs. At some inflection points it sometimes makes sense to use different technology from the source-side encoder to the content delivery infrastructure versus the content delivery network to the final consumption nodes. Archiving the highest resolution content is sometimes a good enough excuse to move to less established technologies to mitigate storage costs. But mass distribution always requires well-established technologies for maximum compatibility and reach.

Transcoding can be expensive. It makes sense to study what can be achieved to minimize transcoding burdens on a video distribution infrastructure. When a video library is archived in a highly compatible format it may still be the better compromise to use a well-established codec, like H.264, right from the get-go. Some emerging standards falter or get skipped. And some well-established standards continue to generate more evolved implementations and have a very compelling mix of performance and broad compatibility.

But whatever your workflow requires, multi-channel encoders and transcoding software and services can often assist with moving between codecs and encoding profiles and helping you reach your viewers.

All three previous sections above combine to demonstrate how supporting multiple protocols simultaneously, how transcoding and transrating, and how producing different resolution and quality streams to deal with different bitrates and decoder/players justify multi-channel encoding.

We also reviewed different methods of multi-channel encoding including: multi-channel encoders that produce multiple streams right at the source, adaptive bitrate encoders which produce multiple profiles for compatible destinations to choose from, and transcoding media servers–which are software and services that let you manipulate and multiply your source video streams to suit your application.

Hybrid environments that fully-leverage one or more of these multi-channel encoding technologies allow organizations to serve streaming content in the best ways to all points factoring in considerations of security, network bandwidth, number and type of decoders/players, and more.