Client-Side Live Stream Splicing for Arbitrary Ad Insertions

Instead of the server controlling segment cadence and requiring manifest polling to accommodate variable splice points, the patent inverts the approach by having the client application control the playback cadence. The client downloads segments at a fixed cadence but plays back frames at variable rates to achieve arbitrary splice points, eliminating the need for server-side manifest polling while maintaining both fixed segment structure and flexible insertion timing.

The patent applies preliminary action by having the client application pre-download multiple media segments at fixed cadence intervals before the actual playback occurs. The client stores these segments locally and then plays back frames at variable rates to achieve arbitrary splice points. This eliminates the need for manifest polling during playback, as all necessary segment information is already cached, thereby preventing network bandwidth waste while maintaining splice point flexibility.

The patent applies dynamics by decoupling the segment download cadence from the playback cadence. Media segments are downloaded at a fixed cadence for efficient mapping, but the client application dynamically adjusts the frame playback rate to accommodate arbitrary splice points. This allows the system to maintain efficient fixed-cadence segment acquisition while implementing variable cadence playback to insert supplemental content at any desired timing, resolving the contradiction between mapping efficiency and insertion flexibility.

The patent applies self-service by shifting the content insertion operation from the server to the client application. The client application independently manages its own playback, downloads segments at fixed cadence, stores them locally, and dynamically adjusts playback rates to insert supplemental content at arbitrary splice points. This eliminates the need for server-side manifest customization for each client, enabling scalable deployment where multiple clients can receive the same fixed-cadence segments and independently perform their own content insertion operations.