Getting to Know VOD Metadata 2.0

By now, you've undoubtedly bumped into the “metadata” term a few times. It's not new. In a video sense, metadata is the information applied to a title at its creation to describe who's in it, what it's about, how long it is, when it needs to be removed from a video-on-demand (VOD) server — details that describe what's being stored.

At this moment — just after Thanksgiving, 2006 — the metadata used by cable operators in their VOD systems is pretty stable. It grew out of a specification issued by CableLabs in 2002, which means it's had four years to bake into the gear used by content owners, aggregators, and headends.

Yet, as any tech-side person will tell you, the blessing and the curse of technical specifications is their tendency to engender additions. Metadata isn't immune. As the on-demand sector continues to evolve, so does the need to improve its component parts.

That brings us to “VOD metadata 2.0,” a specification released by CableLabs earlier this year. This week's translation is about what it does, and three reasons why it matters.

For starters, know that most of the goodness in the VOD metadata 2.0 spec happens way in the background — far from consumers, their couches, their TVs and their remotes. This is back-end, behind-the-scenes, make-your-life-easier stuff.

If you ask a technical person to state the No. 1 most important thing about metadata 2.0, they'll probably explain how it eliminates the need to “double-pitch” a video asset (meaning a title and its accoutrements) if something changes.

In today's VOD world, when you want to change anything about a title that's already been populated into a server, your only option is to ask for a resend of the whole thing. That wastes bandwidth. The new spec fixes that. More title volume, less bandwidth used. So that's one thing.

If you pose the same question to a VOD operations person, they'll waste no time in telling you what a pain it is, pre 2.0, to correct metadata that comes in botched.

You've seen the botches: You're browsing through the on-demand library. You come upon a listing where the title is repeated as the description, or the description is so long, it falls off the screen.

To be able to correct just the botched metadata, without having to resend the entire show, is quite the hallelujah for systems people. (And yes, content owners, that does mean you'll be asked to resend your metadata if it comes in goofy.)

The third item of note about metadata 2.0 is the grace it imparts to packaging. Right now, for instance, if you've received and stored a title, it probably has specific rules attached to it about how long it can stay active and how to price it.

With 2.0, you could take that same title and bundle it into a double-feature, or a weekend special. You could switch out promos or advertisements, offer a 99-cent special, or do whatever you want (assuming you have permission) — again, without having to go through the process of requesting a re-pitch.

Plus, 2.0 improves search features, adds substantial multilingual support, and lets you use multiple display formats.

That comes in handy if, for instance, a title can be played out in HD, SD, and a super-compressed mode, as may be needed by a handheld player.

Those are the highlights of what you'll be able to do with the metadata 2.0 spec. The next step is for the supplier community to build it into actual applications, and for the content community to start marking its wares for 2.0.

It'll happen in the background, but it'll happen.

Stumped by gibberish? Visit Leslie Ellis at