Low Latency Video Streaming: The Complete Guide

immanuel vinikas headshot
Immanuel Vinikas
Updated January 31 2021
Low latency streaming
immanuel vinikas headshot
Immanuel Vinikas
Updated January 31 2021

You want to check all the boxes to deliver the best video streaming experience to your audience. That’s why you invest resources to produce high-quality visuals, sound, and content. Another important factor that will influence your audience’s experience is latency. The lower the latency, the closer to real-time your stream will reach your viewers’ devices. But lower latency also comes with tradeoffs. Let’s take a look at what low latency streaming is, whether you need it, and how to choose the right service for your video stream.


What Is Low Latency Video Streaming?

Latency refers to the time delay between input and output, otherwise known as lagging. In video, it is often expressed as “glass to glass.” It measures the time it takes a live video stream to travel from the front glass of the camera lens to the glass of the viewers’ computer screens.


Latency is the culprit when a game is on and you hear your neighbor cheer for a goal, while you’re still waiting for the player to shoot the ball. Latency is responsible for those three awkward seconds a reporter at the scene stares into the camera before answering the news anchor’s question.


The latency of HTTP streaming is around 30-40 seconds, compared to 5-10 seconds in broadcast TV.


Why Is Video Latency an Issue?

For video-on-demand, this latency is not an issue. It’s when we enter the real-time communication or live streaming arena that this delay becomes problematic. Virtual events, online education, webinars, and all-hands meetings all depend on the latency being as close to real-time as possible.


According to ABI research, live streaming will grow 10% to 91 million subscribers by 2024. For many OTT services that wish to get a foothold in live streaming, achieving low latency remains a challenge.  Also, virtual event organizers are shopping for low-latency solutions to enable audience engagement and interaction. A 30 to 40-second delay would render real-time feedback impossible.


In an increasingly competitive market, low latency has become a major selling point. Offering a low latency service to your audience is an essential strategy to keep ahead of the game.


What Causes Video Latency?

The duration of video delay depends on which low latency video streaming protocols and encoding formats you use. Between video recording and display, several processes can influence video latency. Video encoding, CDN buffering, connection type, bandwidth, and the player itself can significantly burden data transmission. Technologies like buffering and adaptive bitrate streaming enhance the viewing experience, but they also increase latency.


However, the root cause for video latency is segment length. Video files are heavy. To make streaming possible, they first need to be broken up into segments that are more ‘digestible’ for the network. These segments are typically between two and ten seconds long. Delivering these segments to the viewer creates a latency that is at least as long as one segment. Often though, several segments are buffered before the video even begins to play.


Do you need a low latency streaming solution?

Why is there even high latency? Why not just make everything low latency from the get-go? Because low latency comes with tradeoffs. Video quality is one of those tradeoffs. Hence the low image quality you get on communication apps like Skype or WhatsApp. With low latency, you also introduce a higher risk of buffering issues on the viewer’s end, because the servers need to ingest the live feed in a shorter time. That’s why another tradeoff is that you’ll need more servers and higher bandwidth to pull off low latency.


Your streaming format might not require any interaction with the audience or maybe your content isn’t time sensitive. In those use cases, latency becomes a non-issue, and you can shoot for quality, scale, and reliability, without any compromise.


Conversely, if interaction and timeliness are essential, then other factors like 4K video quality will have to take a backseat. In video streams that require live interaction, getting everyone in your audience synchronized is key. You want your viewers to get the content at the same time so they can interact with it and with each other in a natural, flowing way. Think Q&A sessions and discussions during your all-hands meeting or a live interactive keynote during your virtual event.


In Video chat/conferencing applications such as WhatsApp and Zoom, low latency, and even ultra-low latency, is crucial. These platforms wouldn’t be able to function properly with delays. That’s why they put a lot of effort into optimizing their data transmission processes to assure uninterrupted conversations between their users.


Choosing the Right Low Latency Streaming Service for Your Needs

As mentioned earlier, many factors influence the latency of your video stream. The main factors being segment length and the streaming protocols that are applied in the transmission. Low latency streaming services address both factors to ensure smooth delivery of high-quality video at scale.

So, when choosing a suitable low latency streaming platform, look at the protocols they support.


Standard video latency (+20 seconds)

Choose standard HLS (HTTP Live Streaming) or DASH (Dynamic Adaptive Streaming over HTTP) protocols if a higher latency won’t compromise your audience’s streaming experience. Great for VOD or live events to large audiences without interaction. This type of streaming allows for HD or 4K video quality and massive scale.


Reduced video latency (20-5 seconds)

You can further reduce your standard latency via HLS or DASH protocols to under 20 seconds, by optimizing your streaming chain. RTMP (Real Time Messaging Protocol) will get you closer to the 5-second mark. Look for tuned HLS or DASH protocols for streams that are time-sensitive but don’t allow any interaction with the audience.


Low video latency (5-1 seconds)

To reduce latency even further, you need to add and combine multiple solutions.

Enter the low latency protocols: LLHLS, LLDASH, SRT, and RTP/RTSP. Great for any stream that would need to match up the 5-second content delivery of cable tv.


Ultra-low video latency or near-real-time (-1 second)

Ultra-low latency is considered real-time video streaming, meaning you no longer suffer delay. WebRTC (or web real-time communication) is the protocol you will need if you want to deliver competitive live streams with sub-second latency. This is a must for two-way communication situations such as online meetings and virtual classrooms. WebRTC is open-source and is supported by all the popular browsers, though most CDNs are not yet compatible.


Because of the tradeoffs that come with low latency streaming, you should ask yourself how your video streaming will benefit from nearing real-time.


For any high-interaction, time-sensitive live stream, such as virtual classrooms, video conferencing, online meetings, virtual events, and webinars, low latency video streaming is no longer about competitive edge. In 2020, tech companies pushed the envelope to meet the technological needs of remote workers and social distancing. Today, audiences have come to expect real-time, uninterrupted live streaming.