The Over the Top (OTT) environment facilitates a truly personalized approach to advertising. The OTT advertisement insertion process (while still based on SCTE-104 and SCTE-35 standards) is very unique in its ability to overcome obstacles such as geographical and timing constraints, opportunities for corruption of data during transit, improper interpretation of what the programmer is actually signaling, and outdated legacy equipment. Using adaptive streaming protocols like HTTP Live Streaming (HLS) and Dynamic Adaptive Streaming over HTTP (DASH) for example, video signals are broken into small chunks that are transmitted and put back together downstream to create the video. The only problem with using these protocols unaided is being able to precisely determine the location of advertisement boundaries. Advertisements vary in length and do not all conform to static segment delineation (i.e. 6 second segment, 6 second segment, 6 second segment). It is easy to see why a video segment containing a partial ad, or the start of one advertisement and the end of another could prove very problematic. For OTT to be a profitable, seamless consumer experience, the transition from programming to advertising needs to be accurate and under the control of the provider.
Using Cable Labs’ Event Signaling and Management API (ESAM) (a set of interfaces between key components in the stream), operators are able to condition the content in a way that allows for proper identification of advertising start/stop times and location in the broadcast stream. And Crystal suggests that using ESAM in the Crystal Metadata Cloud™ (our unique combination of advanced metadata transmission and patent-pending Temporal Fingerprint technology) provides all the information an operator needs to condition content for ad insertion in the OTT environment. Crystal identifies the ad/programming boundaries and the fingerprint of the video and audio that surrounds it. Receive locations perform the same fingerprinting calculation and, by using the cloud service, can receive exact timing information about when the breaks will occur in the stream they are receiving. The timing information has a high degree of accuracy – easily within a video frame – ultimately leading to precise segmentation and a high-quality experience for the end user.