August 2018 Summaries

The new Real-Time Streaming Dashboard is designed to help video platform operators quickly identify and resolve critical issues that may occur during real-time streaming events. It provides essential metrics for engineering and ops teams in one real-time view, allowing them to spot anomalies and failures before they escalate. The dashboard displays metrics as trends, enabling users to see abnormal activity instantly and investigate further if necessary. Additionally, it offers visibility into the impact of platform changes, helping inform infrastructure decisions. With this tool, video platform operators can ensure a superior viewing experience for their viewers.

Mux Video is an API that simplifies video publishing by combining encoding, storage, and delivery into one platform. It replaces the need to build a video streaming platform from single-function APIs or hand over control to third-party OVPs. However, it does not replace players, CMS, or workflow management. Mux Video is designed for developers building video functionality into their software applications and supports both live video and video files. It is not suitable for real-time video conferencing or individual content creators without development resources. The team behind Mux Video consists of experts from YouTube, Oculus, Twitch, and Brightcove. To try it out, sign up for a free account and make two simple API calls.

The text discusses the support of animated GIF creation from video using Mux's Image API. It explains how to create high-quality GIFs with adjustable parameters such as frame rate, width, and duration. The author also suggests that incorporating GIFs can increase user engagement by providing visually compelling previews of content. Additionally, the text mentions the ease of sharing GIFs across various platforms due to their image format.

Nick Chadwick from Mux discussed per-title encoding at Facebook's Video @Scale event. He explained that each piece of content requires a different ladder to maximize its quality, and presented a convex hull of quality bit rate curves for high complexity and low complexity scenes. Per-title encoding can help deliver encodes that maximize quality in HLS or Dash platforms. Mux has developed a machine learning approach for per-title encoding, which is currently live on their platform. The company plans to improve performance, particularly for shorter clips, and may open source the model to gather feedback from users.