Hardware codecs in FFmpeg like those from Intel and NVIDIA deliver significant performance gains over x264/x265, but have a reputation for lower quality and use a completely different command set. This session benchmarks the performance and quality of these codecs and detail the FFmpeg command syntax. Attendees learn when and how to use these hardware codecs in your VOD and live encoding workflows.

CMAF is a format that can create one set of media files deliverable via DASH and HLS (and other ABR formats), potentially saving big dollars for encoding, packaging, storage, and delivery. Decoupling manifests from the media segment formats and being able to deliver the same media segments anywhere is a huge win. With DRM encryption requirements on the rise as well, CMAF offers the best path forward for reach and capability while reducing the formats and versions necessary to hit every platform. Attendees leave knowing the pros and cons of CMAF and with a high-level road map for deploying it.

Machine learning as a service (MLaaS), a subset of artificial intelligence (AI), has come a long way over the last couple years. Now is the time to really make MLaaS work for you and enhance your media discoverability and insights. These enhancements can range from video searchability and discoverability, automating transcript and translation generation, or even pinpointing specific faces or people within your live surveillance streams in an efficient manner. The key to harnessing this power is to learn to determine your needs and code against the numerous ML service APIs that are out there to make them work for you. Attendees walk away with an understanding of how to harness the power of AI machine learning to increase the ROI of their video libraries and build some compelling and innovative intelligence from their libraries of dark media data.

The next great codec battle will be between the Alliance Open Media’s AV1 codec and MPEG’s Versatile Video Coding (VVC) technology. In this talk, Bitmovin codec engineer Christian Feldman discusses the coding tools and performance of both codecs, and the lessons that he’s learned implementing AV1 in Bitmovin’s encoding stack for cloud and on-premise deployment. Learn the pros and cons of both codecs in this 45-minute technical dive.

Companies on the cutting edge of video engineering and development share the latest technologies that are pushing the industry forward.

When news is breaking, broadcasters need fast, on-demand ways of getting their content from the field to production, even from unexpected locations worldwide. By leveraging the SRT protocol to deliver media across the global Microsoft Azure fibre network, the SRTHub helps broadcasters and video service providers easily build live and file-based content routing workflows on-demand, with security and reliability.

For streaming providers delivering content over multiple infrastructures to multiple makes and models of consumer devices and pay-TV set-top boxes, optimizing customer satisfaction requires advanced QoE metrics that encompass every step of the video delivery chain. These include accurate monitoring to maximize performance of large-scale video distribution networks, as well as efficient problem identification, localization and resolution. Traditional methods operate independently at different points along the video delivery chain, reporting partial, inconsistent and fragmented measurements. This session will discuss how unified end-to-end quality monitoring—instantaneous and correlated measurement at every stage of delivery—is fundamental to understanding the final effect on the viewer’s quality-of-experience.

The Twitch video team has been working on adding multiple codec capabilities to Twitch’s video streaming platform, and VP9 is the first codec being added in addition to H.264. As other codecs such as AV1, HEVC, VVC, and MPEG5 are showing promising compression efficiency, we aim to be ready when the time comes to add any of these technologies and deliver better video quality at much lower bitrates. This presentation will cover the process it took Twitch engineers to change the platform from a single codec to multiple codecs. A wide range of technical topics will be covered, from codec choice to player support and validation

Developing a video player is no small task. Many different technologies and specifications must come together to create an accessible and cohesive experience. When you are a small team this can feel overwhelming, especially if the player requires custom functionality. In this talk we will walk through getting a core video experience up and running quickly. Then we dive into making our player unique and adding additional functionality such as closed captions and HLS streaming. All the libraries being used are open source, and include React, Video.js, vtt.js, and hls.js.

The details and scope of analytics can be daunting and unexplored for many online video deployments. In this session, learn how to gather critical audience participation data from your live webcasts. Both open source and commercial services are explored and demonstrated.

HLS and MPEG-DASH are the current standards for HTTP-based live streaming, but these designs are inherently slow and add delays to live feeds. Sub-second latency is critical for scenarios such as gambling, auctions, interactive communications, VR, sports, and gaming. WebRTC is touted for its sub-second latency but couldn’t scale to the volume needed by CDNs and couldn’t reach Apple devices. In this session, learn about a novel WebRTC-based solution that would match current CDNs in terms of reach (all devices), quality, cost, and scale, while providing sub-500 milliseconds latency.