Discussion of the SCTE 224 interface is relevant and timely to the vast array of NAB audience members. In the last couple of years, SCTE 104/35 and other frame accurate triggering and marking technologies have begun to take a major foothold in the broadcast industry. The glue that makes this all work is the out-of-band metadata carrier, which is at the heart of SCTE 224.
The SCTE DPI Working Group 5 has seen tremendous uptick in discussion and use of SCTE 224 not just in the United States, but worldwide. It is now being considered as a solution well beyond its original use cases of alternate content and sports blackout. It is now being used for geographic boundary enforcement in the U.S. as well as country boundary enforcement under the new EU laws. It is also becoming the latest vehicle to carry business data as well as advanced advertising data, especially for linear broadcast. As such, understanding this dynamic technology is a must for all broadcasters.
Produced in partnership with the Society of Cable Television Engineers (SCTE).
Prefer to read? Here's the full discussion:
Hello, I'm Stuart Kurkowski, a distinguished engineer from Comcast Technology Solutions. I'm here today to talk about SCTE 224, or ESNI. Hopefully, at the end of this, you'll understand what it is, many of its use cases, and why everybody's talking about it and why it's gaining popularity.
What is ESNI?
So what is it? ESNI stands for Event Scheduling and Notification Interface. You've probably heard that name from previous versions of the standard. Its primary mission is to describe the schedule for a linear playout channel, along with the timing, information, and metadata that go with it.
Traditionally, it's been thought of as the blackout or alternate content solution. But you'll see from today's use cases that it's growing in popularity and its use is expanding greatly in the community. It now includes VOD, start over and look back rights and restrictions, and all kinds of other actions that can be taken on content.
The ultimate goal of the protocol is to allow a content provider to convey to a distribution partner, machine to machine, the playout restrictions and actions that need to be taken in association with its content.
The five objects of SCTE 224
There are five main components, or objects, in the SCTE 224 protocol. SCTE Working Group 5 is the maintainer of the protocol, and they've described these five objects in order to execute this model.
Media and media points
The first one is a media. A media is really a channel. Think of it as a container around a linear channel or linear playout. It contains media points, and those media points are individual points in time on that linear playout where you want to do something. You can identify the start of a program, the end of a program, the start of an ad break, the end of an ad break, and lots of other things. All of those go in what we call media points inside that container.
A media point is then the opportunity, at that point in time in the video, to activate a policy, and the policy brings with it viewing policies and audiences. We'll look at that in more detail so you understand the five objects.
So again, a media is a container, and it contains media points. If a media contains, say, 10 to 14 days of a linear schedule, it could contain 900 to a thousand media points to describe all the events coming over the next 10 to 14 days.
A media point is a point in time where I want to do something. I can describe it with an ID and with a description. If it's the start of a show, I can provide the show title or show name, which is the traditional use. I can provide the source of the show. I can also provide alternate IDs, such as TMS IDs, a content provider's house ID, an EIDR, or something like that to describe the content. There are also effective and expires windows, so I can reuse media points over time. And again, the goal of a media point is to apply or remove policies, which we'll talk about more.
Triggering a media point
There are two main ways to trigger a media point. One is by match time, which is just wall clock time. I have something happening today at two o'clock on this particular feed, so at two o'clock we're going to apply this policy.
I can also trigger it with in-band signaling. SCTE 104 and SCTE 35 are the key in-band triggers that pair with SCTE 224. If I have an SCTE 35 marker in-band in my feed and it triggers a moment in time, I can do something more frame-accurate than wall clock time.
A media point can also carry a lot of metadata. I can include episode, season, and series information. I can include start over and look back information. I can also include a full ADI, with what you'd traditionally expect around actors, genre, ratings, categories, and so on. Remember, SCTE 224 is an out-of-band capability that mirrors the in-band SCTE 35 markers. This out-of-band container is fairly large and can grow in size; it's essentially unlimited compared to what you can put in an in-band SCTE 35 marker.
Audiences
On the other end, if I have this point in time where I want to do something, I need to be able to describe who I want to be affected. I can characterize my viewer, or my audience as it's called within the SCTE 224 standard, in many ways.
There are country restrictions, so I can use country names. I can do states, or postal codes internationally. I can also do DMAs and zip codes in the US context. I can have home zip codes as well as geographical zip codes, for example while I'm traveling, and I can differentiate those. I can do traditional cable-style virtual IRDs. I can do many other audience definitions, such as latitude/longitude. It might even be the situation a particular user is in: if they're having technical difficulties, I can do things differently.
There are also many device restrictions. I can restrict content on a mobile device like a tablet or a phone, or by other characteristics of a device. So I have a broad way of describing my audiences.
I can also put audiences into a group. Say you have a particular zip code. I can use the matching attribute "match any," so if your zip code is any one of the zip codes in this audience, you're in that group. I can also do exclusion: if your zip code is not in the list, you're outside of this group. Obviously the interpretation is up to the operator, especially with device restrictions; that's left up to the distribution partner.
Viewing policies and actions
So the paradigm is that I have media points identifying a point in the video where I want to do something, and I have the audience, the people I want that action applied to. Those two come together in the form of a viewing policy. The viewing policy determines when I want to do it and who it applies to, and then I can apply the action.
There's a long list of actions. The first one is content. This is the traditional one you'd think of for sports blackouts: I want to do something to your content, like take you to a slate because you're blacked out, or take you to an alternate content feed. I can do that through the content action.
There are many other actions, such as fast forward and rewind controls if we're talking about live DVR. I can restrict DAI: if you have DAI rights for a particular audience for pre-roll or mid-roll. I can control capture: whether you're allowed to record it, when, and how much.
We can also get into very detailed actions, such as signal point insertion and signal point deletion. One use case: a content provider needs to send out to two distribution partners, but with two different sets of signals. I can put both sets of SCTE 35 markers in the video feed and then, through ESNI, strip off or delete one set of signal points for one audience and not the other. When the feed reaches each partner, it has only one set of signals, done through signal automation. I can also do signal point insertion: if I'm time-based and don't have automation, I can use the SCTE 224 protocol and the viewing policy action to insert signals. So there's a lot I can do in that action space.
Putting it together: a sports blackout
Bringing it all back together: I have a media point triggered either by time or by an in-band signal, and I have an audience I want it applied to. I bring those together to execute the action.
In a sports blackout use case, I have a media point that represents the start of the sporting event. At that point I apply a policy that says I'm going to send you to a slate. At the end of the game, likewise, I have a media point that represents the end of the event, again an in-band trigger or a time-based one. Then I remove that policy, or lift the blackout. If you went to a slate during the game, I bring you back to the main channel. I can do that show after show, taking different actions for individual audiences or multiple audiences.
Use cases
So with that paradigm, I have a point in time, an audience I want to apply it to, and an action I want to take. Sports blackout was the original use case, but the number of use cases is exploding as people really get into SCTE 35 and SCTE 224.
Blackouts
For controlling playout, sports blackout is still the traditional use case: restricting by geographical zip code or home zip code, restricting on a device, or applying it to a virtual IRD in the traditional sense.
One thing that surprises some people is that it's not always sports that are blacked out. A lot of content today is old black-and-white content acquired by a content provider that didn't get mobile rights when it acquired it. So there are many playout restrictions around black-and-white content on devices that have nothing to do with sports.
Resident media points
Beyond the traditional blackout, which is where many people think SCTE 224 ends, there's the use of resident media points. You can define media points in a media, the bigger container, that aren't specific to a show. They run underneath the whole feed. They might define default behavior if something goes wrong or something is missing, for example a default ad that plays if there are no ads in the video. You can do a lot with resident media points that is independent of an individual event, so there's lots of work in that space.
Alternate content
Related to sports blackout is alternate content, and people are getting creative with it. I can switch you to alternate content because you can't get the main content. In the future, there could be an optional choice where you opt out of something and are diverted to alternate content. I might switch you to an alternate feed for regionalization: you're in one region, and the region next door has content available that you could be switched to. That way I keep eyeballs on my network. I can also put in a VOD asset: if there's something you're not allowed to have or are opting out of, I play the VOD asset and keep you on my channel.
There's lots of work in the alternate content space that you couldn't do in the traditional satellite world with downlinks, transponders, and muxes. It's much easier now with SCTE 35 and SCTE 224 working together.
Live DVR and per-partner rights
Likewise, there's a lot of work in live DVR. If I'm conveying information on a live feed, I can provide ADI information that covers the asset, and I can pass recording rights and start over and look back rights through this out-of-band channel, all defined for your specific audience.
Another benefit of this paradigm is that I can make all of this unique to each distribution partner. If I'm a content provider sending video to different partners, I might have different carriage agreements with different controls: one can record it, one can't. I can control all of that based on the audience, the media points, and the resulting actions.
VOD to live
I can also flip that around and do VOD to live. Maybe I have a VOD asset on a CDN. During the live playout, if I have a particular audience I want to substitute that content for at that moment, I can provide the location of the asset and the stitching instructions. I have the moment in time from the media point and the audience identified; the action this time is to stitch in the VOD content and play it out.
Playout controls
There's a rich set of playout controls. With live content that viewers have recorded, for example with 72 hours of start over and look back, I can apply fast forward restrictions and restrict ad skipping. I can do all of that in the out-of-band SCTE 224 ESNI information, which I couldn't do with in-band signaling.
Geographic enforcement
SCTE 224 is also being used for geographic enforcement. In the United States, this includes travel rights: I have a home area, but I want to get that content on the other coast. There was no way to do that in the traditional satellite world, but now I can with terrestrial delivery.
In the European Union, there's been a lot of legislation in the last couple of years around enforcing country boundaries, and satellite downlinks don't match country boundaries well. You can do that with SCTE 224. There's also new EU legislation saying that if I have a home country and travel to another EU country, I must have access to my home country's content. You can do that by characterizing the audience and setting up those restrictions, and content providers can easily do that.
Electronic program guide
Another exciting area for SCTE 224 that people often don't think about is the electronic program guide and how information gets into it. That might be the fact that a game is running long while the guide still shows it as over.
Another important one for a great user experience: I look ahead five or six days to set up a recording. A couple of days later I go to watch it, and it just says the show isn't available in my area. That's disappointing. For a better user experience, you'd like to apply the same rules ESNI applies in real time to the program guide. Then I don't even show you that show in the future to record if you're not going to get it. Or, if you're going to get alternate content in its place, I show you that alternate content, so the EPG is accurate. We see a lot of that happening on the EPG side with ESNI.
Advertising
The last category: SCTE 224 just went through a ballot process for the 2020 version, and a lot of that centered on advertising. Content providers are putting more and more information into their ESNI feeds and getting it out 10 to 14 days ahead. They've been good about marking program starts and ends, but now we see content providers starting to mark their ad breaks.
Now I can provide instructions on whether an ad is local or national, whether it's replaceable, and whether it can be targeted. I can also include ad inclusion or ad exclusion information. If a national car ad is playing, I can say, "In your local ad slot, plug in a local dealer's ad, because it'll align." Or I can say, "This was a car ad, so don't show a competing one." For live DVR, I can also include C3 and C7 restrictions. I can put all of that in this out-of-band model.
Closing
As we think about these use cases, it's an exciting time for SCTE 35 and its ability to trigger in-band and get the attention of the playout system. When I describe my audience, I can now take this rich set of actions, put it into play, and do many things with content, well beyond the traditional blackout and alternate content uses we usually associate with SCTE 224.
Thank you for your time. I'm Stuart Kurkowski with Comcast Technology Solutions, and I appreciate it. Hopefully you learned about SCTE 224. Thank you, bye.