Ultra Low Latency Streaming: When You Have the Need for Speed
June 27, 2018 by
You have the need for speed—at least, when it comes to your live-streaming video app or service. Whether you’re a gamer looking for help beating the level; you’re hosting an online auction; or you’re having a video chat with a friend halfway around the world, you need streaming that’s as close to real time as possible, without delays, buffers or drops.
Today’s industry-leading streaming technology wasn’t designed with low latency in mind, and therefore can introduce 30 seconds or more of latency—which can result in poor user experiences, missed opportunities to engage and monetize audiences, service interruptions and lag. To combat this, streaming professionals have had to choose between low-latency streaming or high-quality delivery at scale.
That is, until now.
What Is Considered Ultra Low Latency?
When we talk about “ultra low latency,” we’re talking as close to real time as possible. That means less than five seconds of end-to-end latency (the time between when a frame of video is captured and when it appears on a viewer’s screen).
The Wowza Streaming Cloud™ service with Ultra Low Latency delivers with less than three seconds of end-to-end latency for audiences as big as you can assemble. Available now, this revolutionary service provides a consistent user experience across devices and apps—so you can deliver interactive streaming experiences that connect viewers and broadcasters like never before.
Here’s a look at how the Wowza Ultra Low Latency capabilities compare to other leading technologies:
For Time-Sensitive Streaming, High Latency Kills the User Experience
Latency is a crucial factor in attracting and engaging viewers. Just an extra half-second generating Google’s search page causes a 20 percent drop in traffic. And Wowza clients report that every second of end-to-end latency past three seconds dramatically impacts viewer engagement. For interactive content providers, any single-digit latency rating is a victory—but three seconds or less is ideal.
High end-to-end latency is especially disruptive in use cases with high viewer-to-broadcaster interaction, such as two-way chat or game streaming. If the time between one person performing an action and that same action displaying on-screen is 45 to 60 seconds, real-time interactivity is impossible.
For use cases such as financial trading, auctions and public safety, where users rely on streaming video to make informed decisions, real-time delivery is critical, and every millisecond of latency counts. Not only do these users need to receive video and data in real time, but they also need to get it at the same time.
The takeaway: where money and lives are on the line, high latency will kill the user experience.
Use Cases That Rely on Ultra Low Latency Live Video Streaming
Now, let’s explore some of the use cases for ultra low latency streaming in greater depth:
Financial Trading, Auctions, E-Commerce and Gambling Platforms
For apps and services in this category, time is quite literally money. An extra 100 milliseconds in latency costs Amazon 1 percent of sales, and being just 5 milliseconds behind competitors can cause a broker using an e-trading platform to lose $4 million per millisecond in revenues.
Financial traders need up-to-the-second information on stock market quotes, global financial market information and breaking news. Real-time data is crucial to make informed decisions, adhere to regulations and avoid insider-trading scenarios. Being even milliseconds behind can result in catastrophic losses and missed opportunities for traders and their clients.
Auction platforms give buyers from around the world the chance to purchase rare and collectible items they might not otherwise have access to. But if your app comes with high latency, participants will quickly become frustrated, since even the slightest delay can result in one bidder losing out to another. You need to offer streaming video delivery that can keep up with the auctioneer’s rapid clip.
E-commerce apps and services are also beginning to emerge in which viewers can purchase from directly within the live stream. For example, Amazon hosts a weekly live-streaming show called “Style Code Live” on its video platform, in which viewers can purchase the products being discussed through their Amazon accounts while watching the show.
Other apps, such as Busker and Live.Me, allow any user to create a live stream for selling products, where viewers can engage with them as if they were salesperson at a brick-and-mortar store. Viewers can ask questions, see different views or colors of a product and, ultimately, purchase directly through the app or service. If latency gets in the way of an authentic interaction, users will abandon the platform.
Finally, in online gambling platforms, participants from around the world place bets based upon on-screen action. If some viewers are seeing streams before others, it puts them at an unfair advantage, hurting the user experience—and the wallets—of those watching higher-latency streams.
Security, Public Safety and Remote Monitoring Platforms
The most serious implications for high-latency streaming are in security and public safety, where first responders and officers must respond to video and data feeds to address problems as they unfold. Even milliseconds of latency can mean life or death for those in harm’s way, and/or massive losses in property, goods and information for businesses, government facilities and private homes.
Whether it’s a store break-in or a large-scale event such as a natural disaster, those monitoring feeds from security and public safety platforms need split-second information so they can decide where to dispatch responders, how many people to send to the scene and what equipment or vehicles are needed to help. These solutions also need a global footprint, so RTMP streams or IP camera feeds can be pulled from dispersed geographic locations.
Remote monitoring solutions, such as drones and networked pan-tilt-zoom (PTZ) cameras, need the same split-second streaming delivery to ensure remote operators can keep up with what’s happening on the ground. If there is a delay between the video feed reaching the operator or control center and the action on the scene, it can mean crashed drones or security incidents.
Gaming and Esports Apps
Game-streaming services have become hugely popular in recent years. Engagement is naturally high on these platforms—unlike many other markets, it’s common for users to spend up to eight hours in a given session.
That’s largely because these online services are more than just destinations where viewers passively watch someone play a game. They’re collaborative communities where streamers broadcast their gameplay, and viewers not only watch the stream, but provide suggestions to the streamer and discuss the action with other viewers in real-time chat conversations.
Unlike professional-sports fans, most viewers on game-streaming platforms are regular players themselves. While your average sports fan wouldn’t tell a star quarterback how to pass, your average gamer has a wealth of tips and suggestions for the streamer—and is eager to participate.
Of course, all of this interactivity wouldn’t be possible without very low latency. If there’s a 10-second delay or longer between the streamer completing an action and the viewers seeing it on the screen, by the time they react, it will already be too late.
In an interview, Stu Grubbs, CEO of Lightstream—a game-streaming service that delivers to popular gaming destinations such as Twitch and Mixer—describes the impact of latency to the viewer and broadcaster experience:
“The success of video game live streaming depends on authenticity. In most cases, the ability to interact with the audiences, while sharing their gameplay experience not only increases the authenticity of the broadcaster, but the viewer experience. As you increase stream’s latency, the lack of interactivity or ability to engage in near real-time immediately impacts the viewer’s perception of the game or quality of the streamer’s content. That’s why latency is critical for game streamers.”
What’s more, streamers make money (and may even make a living) off paid viewer subscriptions. Viewers can also give monetary tips to their favorite streamers. Some platforms, such as Mixer and YouTube Gaming, offer extra incentives to streamers when viewers watch without ad-blocking. High latency on these streams translates to poor user experiences, lost viewership and fewer subscriptions. In turn, this means less money in the streamer’s pocket—and fewer percentages of subscriptions, tips and ad revenue for the host platform.
Indeed, delays of mere milliseconds can result in lost customers. Even two seconds of latency when loading a gaming transaction can cause abandonment rates of up to 87 percent. First-person shooter games suffer from as little as 100 milliseconds of latency, since action happens so quickly. Role-playing or turn-based games can afford slightly longer delays, but interactive capabilities may still be impacted.
For platforms featuring esports—professional gaming competitions where the upper echelon of gamers face off to entertain massive online and in-person audiences, and win big prizes and sponsorship opportunities—near real-time streaming is especially important. The host organizations can’t afford to disappoint the enthusiastic fan bases that turn out for these events.
User-Generated Content Apps With Live Video Streaming
Social media and user-generated content (UGC) apps with live video streaming, such as Periscope, Facebook Live, Instagram and YouNow, allow anyone to broadcast stories and snapshots of their life. Live and ephemeral video (video that disappears from the host platform as soon as the live broadcast ends) on these apps is made possible by the fact that users almost always have a camera on their person, and can capture unique moments while on the go.
In turn, other users can experience those moments in real time wherever they are, thanks to their own mobile viewing device. Globally, there are 2.6 billion smartphone users, and a shocking 87 percent of them are literally never away from their phone at any time of day.
This phenomenon leads to new expectations for the UX: Since users want to be able to stream and watch any given moment at any given time, they expect their apps and devices to keep up. They also expect streams to load immediately and play without interruption. Perhaps most importantly, many platforms rely on direct interaction between viewers and broadcasters—whether through two-way video chat, or through one-to-many interactions, where broadcasters take questions and feedback from the online audience via real-time comments.
For these interactions to be authentic and valuable to both viewers and broadcasters, low-latency streaming is key, enabled by a real-time streaming protocol such as WOWZ™. As with e-commerce apps, UGC platforms must also be able to scale rapidly to accommodate massive numbers of viewers watching the same viral stream or participating in a live broadcast from their favorite celebrity. The Ultra Low Latency Service offers all these features, along with a robust set of APIs and a mobile SDK for building real-time streaming apps to reach any device.
If an app or service can’t provide a top-notch experience, users quickly lose faith. And with so many UGC distributors to choose from, they’re likely to abandon platforms that don’t meet their expectations. To learn more about how to build live-streaming UGC apps consumers will want to use, read our complete report.
Global Live Video Streaming in 3 Seconds or Less, for Any Size Audience
Our Ultra Low Latency service meets the needs of any use case that relies on real-time streaming, offering sub-three-second delivery on a massive scale. Deployed on Microsoft Azure data centers, with the most global POPs and nodes, it shortens the first and last mile, reducing congestion that can slow down streams. This ensures that every user sees the same information at the same time.
Scalability is another key factor to consider. Live-stream viewership can spike because of popularity or regional demand, so your app must be able to handle a sudden surge of peak concurrent viewers. That’s why our Ultra Low Latency service offers auto-scaling capabilities: It intelligently adjusts streaming instantly and non-disruptively, so it can handle any size audience.
Intelligent load-balancing and auto-scaling capabilities are built in, so you always have the resources you need to deliver reliable, high-quality streaming. Wowza Streaming Cloud with Ultra Low Latency Service also provides a consistent user experience across devices and apps, so viewers can place more accurate bets, make smarter trades and act on up-to-the minute intelligence.
As a managed, cloud-based service, it even eliminates resource strain and saves you the hassle of managing your own network. Best of all, you pay for only what you need—no expensive infrastructure costs.
Low Latency Meets Enterprise-Level Stability and Uptime
When every second counts, your streaming just needs to work. Intelligent monitoring and optimization capabilities identify issues before they start and automatically adjust streams, reducing interruptions and maximizing uptime. To ensure availability, our service also offers self-healing and auto-scaling capabilities.
With these tools, you get complete visibility, insight and control throughout your entire streaming workflow: from ingest to playback. This allows you to anticipate potential problems, prepare for spikes in traffic and optimize your workflow to deliver top-quality, reliable streaming to your end viewers. And since it’s built on Microsoft Azure, the leading cloud computing infrastructure, enterprise-level redundancy and stability are available to everyone, everywhere.
The Wowza Streaming Cloud service with Ultra Low Latency meets the needs of users who need a reliable, managed streaming service in industries where every second counts. To add Ultra Low Latency streaming to your Wowza Streaming Cloud account, or sign up for a new plan, visit https://www.wowza.com/products/streaming-cloud#ultra-low-latency.