Delivering Broadcast-Grade Low-Latency Live Streaming with Lumen® CDN
As cord-cutting continues to rise and more viewers around the world consume television over the top (OTT), the need for broadcast-quality streaming could not be clearer. That’s especially true for live content, where both the quality and timeliness of the video delivery must be able to match that of cable TV. And with live streaming experiencing 99% YoY growth in 2020, the need to meet user demands for exceptional live video has never been more important.
Whereas cable and satellite TV standards typically have 4-5 seconds of “glass to glass” delay between the actual live event and the rendering on the television sets, when done over the internet, the glass-to-glass process can sometimes take anywhere from 20 seconds up to a full minute. When this happens for a popular live stream such as a big sporting event or a breaking news broadcast, the streaming viewers often can have the “live” aspect of it ruined by app alerts on their phone, social media posts, or simply from the reactions of neighbors in the next room who see the video feed well before they do. Hearing your neighbor shout “Goal!” 20 seconds before you see the play on your OTT device is just not going to be acceptable.
Several years ago, the predominant method for live streaming was based on Adobe Flash Player and the RTMP protocol, which typically provided very low latency experiences with sub-5 second delay. As the industry migrated to HTTP-based streaming protocols such as HLS and MPEG-DASH, reliability and quality improved significantly, but we took a bad hit on latency. The new low-latency HLS and MPEG-DASH protocol extensions address this issue, bringing us back to the 5-second latency target while preserving reliability and quality.
A critical moment came with the 2020 release of iOS 14, which included low latency specs for HLS that have quickly become the industry standard for low latency implementation, as well as the increased usage of LL-DASH for non-Apple platforms. With these specs now being adopted by many companies specializing in streaming media and online entertainment, Lumen’s ability to support them is vital to our mission of helping companies deliver incredible user experiences around the world.
How Lumen supports low latency
Within the ‘Deliver’ phase of the capture-to-consumption process, standard video streaming protocols have traditionally required the video buffer to be full before the CDN pushes a segment out to the viewers. This is done to provide a cushion that reduces the risk of rebuffering in case the stream suffers an interruption. Segments are typically about 10 seconds each and are sent out after three of them fill the buffer, resulting in 30 seconds of latency added to the capture-to-consumption process (3x10s=30s).
Using this process, latency can be reduced by simply sending smaller video segments to the CDN (e.g. shortening them to 2-3 seconds). The danger, however, is that it increases the likelihood of rebuffering instances because there’s now very little margin for error; 2-3 seconds is typically not enough to recover from a performance interruption without it affecting the end-user experience.
To get the latency down to the broadcast standard of 3-5 seconds, those segments must not only be shortened to 2-3 seconds, but a hybrid push/pull process must also be enabled to ensure the continuous flow of video segments. This method, known as Chunked Transfer Encoding (CTE), is an advanced HTTP feature that allows the bits to be sent out before the full 3-second chunk is taken in.
In addition to supporting CTE, Lumen has also decided to not favor certain formats over others. By being format-agnostic, we can support both HLS and DASH, and the wide variety of devices they serve. By embracing this flexibility, our goal to is empower our customers to use whichever protocol is best suited for their needs, as well as make it easier to support other low-latency protocols as they emerge, such as HESP.
Lumen has also prioritized support for the Common Media Application Format standard, which allows media fragments to be shared across HLS and DASH protocols, thus removing the need for video providers to double the amount of data that must be sent, and doubling cache efficiency.
Use cases for low latency streaming
Source: wowza.com
In referring to the common latency continuum, Lumen has prioritized support for “standard” low latency, as it brings the streaming industry back in line with the ~3-5s standard of broadcast TV. With live sports now back in full swing – and with many stadiums still partially empty due to COVID restrictions – the demand for live streaming on OTT platforms is more relevant than ever. Providers who can offer those live sports in low-latency streams can call themselves a truly viable alternative to cable TV, and therefore hold a distinct advantage over competitors whose streams still come with the excessively high 20-60 seconds of latency.
We also see use cases for which low latency drives success in the online gaming and e-sports industries, where synchronization is critical to the end-user experience. This also applies to any platform in which users interact with or respond to video content in real time as they watch it. That includes live video- sharing platforms (e.g. Twitch, Facebook Live, etc.) in which the host interacts with viewers through the comment section, live online polls as an event is unfolding, or virtual gambling/betting platforms. Excess latency in instances such as these can disrupt the experience by ruining the “live” aspect of those interactions.
As more use cases arise, Lumen will be ready to fulfill the growing need for low-latency live streaming as it continues to grow around the globe.
Next week we’ll be diving deeper into the technical specifications behind Lumen’s support for low latency, as well as showcasing our solutions for all the different phases of the capture-to-consumption process.
Learn more about the Lumen CDN.
This content is provided for informational purposes only and may require additional research and substantiation by the end user. In addition, the information is provided “as is” without any warranty or condition of any kind, either express or implied. Use of this information is at the end user’s own risk. Lumen does not warrant that the information will meet the end user’s requirements or that the implementation or usage of this information will result in the desired outcome of the end user. This document represents Lumen’s products and offerings as of the date of issue. Services not available everywhere. Business customers only. Lumen may change or cancel products and services or substitute similar products and services at its sole discretion without notice. ©2021 Lumen Technologies. All Rights Reserved.