Manipulating the manifest of streamed video allows localisation of adverts with the option of per-client customisation. This results in better monetisation but also a better way to deal with blackouts and other regulatory or legal restrictions.
Using the fact that most streamed video is delivered by using a playlist which is simply a text file which lists the locations of the many files which contain the video, we see that you could deliver different playlists to clients in different locations – detected via geolocating the IP address. Similarly different ads can be delivered depending on the type of client requesting – phone, tablet, computer etc.
Here, Imagine’s Yuval Fisher starts by reminding us how online streaming typically works using HLS as an example. He then leads us through the possibilities of manifest manipulation. One interesting idea is using this to remove hardware delivering cost savings using the same infrastructure to deliver to both the internet and broadcast. Yuval finshes up with a list of “Dos and Don’ts” to explain the best way to achieve the playlist manipulation.
Sarah Foss rounds off the presentation explaining how manifest manipulation sits at the centre of the rest of the ad-delivery system.
With the rapid increase in video-on-demand (VoD) viewing, over-the-top services such as Netflix and mobile TV, working out who watches TV – and when and where – has become a complicated business. Can the TV industry keep up with the changes and the ever-growing need to measure TV viewing habits across devices, platforms and other new ways to watch?
This panel from the RTS, includes Justin Sampson from BARB talking about how they’re capturing iPlayer views which are increasingly important, particularly has sometimes a TV programmes are available on OTT before linear transmission. But there is still work to do capturing views on Netflix, Youtube, Amazon and other services.
In the same vein, Rich Astley, Finecast CPO, pointed out there is a lot of advertisement viewing which is also uncaptured.
Sky Media’s MD John Lister gives his opinions including discussing the strong ability of linear TV to build brands which, clearly, is more important than individual advert spots.
Sarah Rose from Channel 4 discussed the continued importance of overnight ratings which are still highly correlated to having a hit on your hands.
A great look at SCTE35 and how it’s used from Roger Franklin and Alan Young, CEO and COO at Crystal given at the SMPTE Conference 2017.Watch Now.
SCTE 35 – “Digital Program Insertion Cueing Message for Cable” – is routinely used to identify the location and composition of programming content and advertising breaks in linear television for OTT providers and has been for a long time.
SCTE 35 specifies metadata that can be inserted into the MPEG-2 Transport Stream carrying the compressed content. SCTE 35 contains the precise frame of the beginning and end of video segments, content identifiers and rights-related information. However, the real-world implementation of SCTE 35 by content providers is inconsistent despite SCTE 67 – “Recommended Practice for SCTE 35 Digital Program Insertion Cueing Message for Cable”. Worse, the ever-increasing complexity of distribution and transcoding for delivery to multiple devices has taken its toll on SCTE 35. It rarely survives delivery to the OTT provider without being corrupted. This is obviously a problem for both the OTT providers and the content providers not only because it limits their ability to monetize the content but also because it makes it much harder to effectively automate the implementation of the complex rights associated with online content in an auditable manner.
This webinar describes a method of delivering SCTE 35 out of band using temporal fingerprints to re-synchronize the SCTE metadata with the video at each receiving point. This not only solves the core problem but provides many side benefits including automatic lip sync error correction, enabling broadcast adverts to become ‘clickable’ and enabling graphics to become customizable and user selectable.