A method for controlling a treadmill using a workout video comprising receiving a workout video at a treadmill comprising a display, an endless belt, one or more actuators configured to control a speed and incline of the endless belt, and one or more processors. The method may also include displaying, by the one or more processors, the workout video on the display, wherein the workout video includes audio of a song, text of lyrics of the song, and an indication of progress through the lyrics. The method may also include controlling, by the one or more processors, actuators of the treadmill according to control signals associated with the workout video.

A client device is disclosed that receives, from a server, a live video stream and a production quality live ambisonic audio stream generated during performance of a live event at a venue. The live ambisonic audio stream is generated from audio channels captured by audio capture devices disposed at the venue. The audio channels captured at the event, and modified by a producer, can be compared to audio captured by an ambisonic microphone positioned within the event space to determine the phase and relative amplitude of those channels as received by a particular ambisonic microphone channel. In this manner, raw and/or produced audio channels captured at the event can be shifted and mixed together to generate a production quality ambisonic stream.

Alert directives and focused alert directives allow a user to provide feedback to a behavioral recognition system to always or never publish an alert for certain events. Such an approach bypasses the normal publication methods of the behavioral recognition system yet does not obstruct the system's learning procedures.

A client device is disclosed that receives, from a server, a live video stream and a production quality live ambisonic audio stream generated during performance of a live event at a venue. The live ambisonic audio stream is generated from audio channels captured by audio capture devices disposed at the venue. The audio channels captured at the event, and modified by a producer, can be compared to audio captured by an ambisonic microphone positioned within the event space to determine the phase and relative amplitude of those channels as received by a particular ambisonic microphone channel. In this manner, raw and/or produced audio channels captured at the event can be shifted and mixed together to generate a production quality ambisonic stream.

Implementations generally relate to the television viewing experience. In some implementations, a method includes detecting a program being presented on a television. The method further includes determining content of the program, where the content is displayed on a screen of the television. The method further includes displaying light on a wall behind the television based on the content being presented.