We’ve recently hit a problem based around a lack of clarity around the “equivalence” of assets. For people in the broadcast industry, this is an old chestnut, but I think it’s worth exploring here.
Lets say you have a program, for the sake of argument: “Vertical City” (1st Episode). Let us also assume it has a total of 6 episodes. (For those with Deja-Vu, it is a real program on Channel 4, but I’m just using this as an example).
Let us also suppose (this is not the case as far as I know) ZDF buys this series, as does NED1, and then RTE.
Let’s introduce a viewer, Bob, who has decided that he wants to watch the entire series. He gets all these broadcaster/channels: RTE, NED1, ZDF, and C4 but he’s come late to the party and C4 is at episode 4 (I’m ignoring the iPlayer/4oD part here :)). Now if you had a Tivo, and told it to get the whole series, it would make this assumption:
A program is the same regardless of channel
So it could reconstruct the series as follows:
- Ep1 – from RTE
- Ep2 – from RTE
- Ep3 – from ZDF
- Ep4 – from NED1
- Ep5 – from C4
- Ep6 – from C4
And that’s the same as the original series on C4 ? Isn’t it ?
If you did this, you’d probably find that NED1 burns subtitles in (it’s a Dutch channel), ZDF might dub the program so the sound track is now German, and RTE might burn a graphic into the program.
The point is that the original asset produced (in this case by Electric Sky) by the program maker is the only part that you can make any assertion of equivalence on. This is how a broadcaster like the BBC can say that an asset they buy to put out on BBC1 is the same as the instance they put out on BBC3 a week later/earlier. Once it’s is on the channel and consumed, it is a different program, even if the viewer might say its a equivalent. BBC3 has the channel logo for instance, the time of showing might be different.
This difference is even more marked in a multi-language environment because of the differing audio/video tracks.
From the point of view of a Broadcast Network owner, such as UPC or Sky, this is frustrating, because you cannot assume equivalence, so each program is different unless the channel operator tells you explicitly that they are equivalent. This affects your approach to PVRs, network PVR, and meta-data. Essentially only the people who fed the tape into the ingest at the transmission stage can tell you if the program is the same.
From an architectural stand-point, this is something to watch where a development view will tell you that there is multiple copies of the same data in the system, even though the “copies” are actually different based on the business rules.
Just because you can run a string ‘==’ on the data with a result of ‘true’, doesn’t make it the same.