Concurrent CTO speaks out on 'Start Over'

Last month during the Consumer Electronics Show in Las Vegas, Concurrent Computer was recognized with an Advanced Media Technology and Engineering Emmy for its work on the development of "Start Over," a service that allows Time Warner digital subscribers to jump to the beginning of any show they tune to in progress.

Concurrent, which was honored along with Time Warner, BigBand Networks, Harmonic and Scientific-Atlanta, for "Outstanding Innovation and Achievement in Advanced Media Technology for the Best Use of 'On Demand' Technology Over Private (Closed) Networks," spearheaded much of the development.

While the Start Over application is being used in the cable industry by Time Warner, it is equally well-suited for IPTV service providers. IPTV Update spoke with Bob Chism, Concurrent Computer chief technology officer, about the project and where viewer interactivity is headed.

IPTV Update: Could you please describe the basic technology requirements for the Start Over feature Time Warner required?

Bob Chism: The basic requirements from a VOD server perspective is you have to be able to follow a live broadcast within a couple of seconds. When I say "follow," technically what you are doing is being able to write the storage device and then play out of it within a couple of seconds so that your network DVR capability mirrors the same type of user experience you would get with something that everyone seems to be familiar with today, the TiVo device.

IPTVU: Could you describe the primary challenges you had to overcome in designing the Start Over function?

BC: One of the primary hurdles you have to run against is the number of channels to ingest. Although Start Over right now is incremental in terms of their strategy relative to the number of channels they are trying to ingest, they could easily be in the range of 30 to 50 ingest channels that are commercially deployed in an operational environment like we have with the Time Warner, Columbia, SC, site.

But the original model called for as many as 150 channels and above and being able to hold that content on the system for a week or two. So if you do the math on that, you are easily above 40 terabytes or so — quite large.

When you talk about that type of model, you really can't do it with a RAM-based solution. You have to have a solution that is disk-based when you talk about that magnitude of storage.

Some of the challenges associated with that include the need for a very scalable solution, and at Concurrent, we were the first to separate streams and storage and ingest to their own components so that you could independently scale those. You need a very scalable solution to be able to handle that type of volume of channels and the amount of storage that is going to be required.

Another big hurdle is being able to have the software necessary in terms of not only an operating system and file system, but also the application that sits above that. It must not only be scalable but also give you the performance you would need so that you could follow a live broadcast within a couple of seconds.

A lot of the architectures that are out there will not scale appropriately and don't have the level of response necessary to follow that close. Some of the scenarios described to us by some of the MSOs is they didn't want to have situations where the follow of your live broadcast was a large increment of time. They didn't want that.

IPTVU: Could you envision complementary technology to Start Over in the realm of advertising and commercials?

BC: Sure. In my opinion that's a next step of that. Clearly, right now what they are doing is the entire capture, but eventually ads would be dropped out, and you would end up with what would be utilized as a technique called playlist, so you could construct playlists of the content that would be captured and you would intermix that in eventually with addressable advertising so that you could get unique advertising to that particular profile that you're trying to match.

IPTVU: Do the MSOs have a legal obligation to run the commercials in the broadcaster's program?

BC: Clearly they do in the live broadcast, but on replay, I don't know. I think that will depend on whatever the nature of the negotiated contract would be. Clearly, I don't get into that.

Right now, a lot of them don't keep the content on the system that long, so maybe it's not that big of an issue whether their ads are rerun or not. Eventually, as they keep the content on the systems longer, they will want to be able to insert ads that are fresh and even beyond that and ads that have a particular profile. Maybe the profile was something as rudimentary as a zip code or something like that, but eventually they will be addressable.

IPTVU: Couldn't the same technology drive a service to allow viewers to request extended advertising on something that interests them?

BC: There has been a lot of discussion and a lot of work behind the scenes to accomplish that, to be able to telescope ads into longer advertising segments. Maybe they get into long-format ads. Certainly, there has been a lot of discussion on how to achieve that.

IPTVU: Do you see an aligned application for this Start Over technology in the IPTV market?

BC: Oh, yes. All of this technology that I am talking about is really independent of whether it is an HFC cable application or whether it's an IPTV application. At the end of the day, the device that the media is eventually received on may be different, but what we are really talking about here is all of the infrastructure that feeds those types of devices.

So, a lot of all the hard stuff is being addressed as we speak. It will be directly portable to over applications like IPTV.

IPTVU: Is there anything else you would like to add?

BC: Clearly, rights issues will continue to be negotiated. More and more content will be available, and more and more content will be made available for a longer period of time on the system. So, there will be massive amounts of storage requires. You are correct: This will lead us down the path of being in the addressable advertising space, and I really believe that at the end of the day, all of this will get combined and all broadcast and on-demand content services will run through video server complexes. So maybe at the end of the day, a number of years from now, we may end up with the major networks — ABC, CBS, NBC and FOX — as broadcast channels and everything else being captured on the system and being able to playout on demand at the subscribers' discretion. The four broadcast networks I mentioned would be captured as well.

But I wouldn't at all be surprised if that's the model we end up with in the long run, where broadcast and on-demand is combined. We've started to see that in the QAM sharing with switched digital, so I just think the plumbing changes a little bit farther upstream from all the server complexes.

Tell Us What You Think!
IPTVU invites response from our readers. Please submit your comments to We'll follow up with your comments in an upcoming issue.