Companies that use live streaming or online virtual events to bond with their overseas clients know how important latency is when a live event is happening. The slightest lag has the potential to disrupt a entire virtual event causing the audience to lose interest.
However, lowering the latency is not as easy as it sounds, especially when you’re implementing procedures such as HTTP Live Streaming (HLS). Luckily, there’s a new process being rolled out known as Low-Latency HTTP Live Streaming (or, because that is a bit of a mouthful… LLHLS). LLHLS is basically an extension of HLS that could solve all your problems.
Here we will dive deep into why LLHLS is a great choice, the features it offers and how you can use it. So, let’s start!
What Is Latency?
Latency is the time elapsed between the user’s action and the resulting response. Network latencies refer directly to delays that occur within the network or on the Internet. For example, during a live event, if an attendee hears something a few minutes after the speaker has said it, that delay is the latency.
Although data on the Internet is moving at the speed of light latency cannot be eliminated. However, it is essential to minimize latency for better performance, especially in the case of online events.
Why Is Low Latency Important?
High latency is something that no one desires, but to what extent is low latency significant? In most streaming scenarios, a typical 30 to 45 second delay is no problem. Most of the broadcasters surveyed in the 2019 Video Streaming Latency report showed that these services experienced latency in the range of 10-45 seconds.
So, what is Low-Latency HTTP Live Streaming?
LLHLS is simply an extension to the existing HLS protocol that’s used for reliable on-demand video streaming services. However, the key difference lies in latency. While HLS focuses more on the reliability side and isn’t specifically known for its low latency, LLHLS performs a lot better in this regard and delivers promising statistics in terms of latency.
The new delay mode reduces video drops on social networks in the normal range of television broadcasts. Back-end production tools and content delivery systems should use new rules to allow for lower latency streaming.
What Does LLHLS Offer?
Here’s a peek at some of the features that LLHLS offers:
· Create specific media segments
LLHLS offers the same media distribution channel on the live edge of the Media Playlist, which separates the media into a more significant number of smaller files, such as CMAF (Common Media Application Format) chunks. These tiny files are called HLS Partial Segment. Because each incomplete component has a shorter duration, it can be compiled, published, and added to the media playlist earlier than its Parent Component.
· Provide Delta Updates Playlist
Clients can move playlists frequently via LLHLS. They can also request servers that can provide Delta Updates playlists that reduce this transfer cost. These updates change the bulk of the client’s playlist already with the new EXT-X-SKIP tag.
· Block Playlist Reloading
In support of adequate client information for new Media and Partial Segments, LLHLS introduces the ability to block application reload playlists. When a client releases HTTP GET to request a Media Playlist update, it can add unique query parameters called Delivery Directives to determine whether it wants the playlist response to include a future segment.
The server then holds the application (block) until the playlist version that contains that segment is available. Block playlist reloading eliminates playlist voting.
· Preload Hints and Blocking of Media Downloads
Eliminating unnecessary overhead time is essential when transporting low-latency streams worldwide. Servers use the new tag, EXT-X-PRELOAD-HINT, to notify clients of upcoming sub-categories and media launch phases. The client can issue a GET request for a predetermined source; the server responds to the request as soon as the media is available.
· Provide Rendition Reports
When using low latency, the client can change the output with a small number of renditions to perform bitrate adaptation. For this purpose, the server adds Rendition Reports to the Master Playlist in each Media Playlist. The EXT-X-RENDITION-REPORT marker owns the Rendition Report and provides details such as the latest Media Sequence number and part currently in the Media Playlist for this translation.
A Detailed Comparison of LLHLS Vs Other Protocols
Let’s have a detailed walkthrough of some ways in which LLHLS is significantly better than its predecessor.
· Flexibility and Adaptability
Speaking of which protocol has the best flexibility, LLHLS is the clear winner. As the most widely used media streaming protocol, HLS is supported by various devices and browsers and is fully compatible with a content delivery network. Another benefit of HLS is its support for closed captions and subtitles, metadata, digital rights management (DRM) and ad placement.
These are the skills people will need, but it will take time for the industry to work within the environment. However, no one can ignore this mechanism for long. This demand will push the development work, so skills like these will not be far behind.
Depending on the quality, LLHLS gets the edge. HLS is common in ABR video, and that includes LLHLS as well. Multiple renditions allow playback on different bandwidths. The media server then sends the highest quality streaming on each viewer’s device with connection speed. The delivery method that automatically adjusts the video streaming quality between multiple bitrates and resolutions is better than the one with only a one-bit rate.
If the solution needs real-time peer-to-peer streaming, other protocols such as WebRTC may be the protocol streaming services want to use. However, if they need to scale their audience to more significant numbers, they will need to think twice about achieving it. Protocols such as WebRTC cannot manage larger audiences and cannot scale in size. However, they can resolve this issue through different workflows. The right choice will depend on the use case and the desired size of the audience.
How can we (Onlive) help?
Onlive supports delivering streams in LLHLS so when the standard is adopted in video players and browsers it will be readily available. As a result, Onlive’s users can enjoy all the benefits associated with LLHLC and deliver content to their customers throughout the world with high scalability but minimal latency.
LL-HLS shows many opportunities to deliver low latency streaming in a single format across all platforms. While the entire ecosystem is still developing and vendors are working on support, there is a very high potential for LLHLS to be a leading protocol that brings low latency ranges of 2-8s across all platforms.