Title: Chief technology officer, TVN Entertainment
Broadband background: Stasi leads an engineering and operations team at TVN that is responsible for managing over 3,000 hours of content each month. He is also president of the Society of Satellite Professionals International, (SSPI) a nonprofit member-benefit society that serves satellite professionals. He has served on the NCTA Engineering Committee for 25 years, and is a member of the Digital Cinema Technology Committee of SMPTE.
We have an article on digital rights management (DRM) in the current issue of Communications Technology. Would you place that topic at the crossroads of content and technology?
Yes, without question. There’s a co-dependent relationship between content and technology that only grows with time. It goes back to the earliest days of my career. I was at HBO and we were going up on the satellite. Suddenly, by being able to put high-value content across the entire country from a single point, we changed the television paradigm. It also exposed high-value content to interception, to random and anonymous interception from anyone. That was the beginning, when rights management and signal denial and encryption saw their first manifestation in the popular space.
As we get into better, more efficient compression systems and we wind up with more high value and very high quality digital content, that content starts to lend itself to file sharing. As VOD becomes more and more popular, we’re starting to see a greater interest in moving the pay-per-view and VOD windows up closer to day-and-date with DVD release. That is profound.
But it’s going to require new levels of protection and control, new levels of traceability and forensics. We’re doing file-based AES at TVN already. We’re doing both link and file encryption. And now in addition to copy protection and certain aspects of rights management, we’ve just begun watermarking in our high-value and HD content. These are significant if complex steps. Our ADONISS systems make managing the complexity possible.
What are the dynamics of this question with cable operators?
Right now, we’re operating in a very comfortable area, where the pay-per-view and VOD window really only represent some 5 percent of the piracy on feature films. If we move closer and closer to day-and-date, if we get to concurrence with theatrical, with very high quality and ubiquitously distributed content, we’ll have moved into a period that represents some 72 percent of piracy. We have to be prepared to increase our levels of security at least to these minimums – if not to prevent the piracy, then at least to defend ourselves against claims of piracy.
With watermarking, you can determine if a claim was correct or not, even the set-top box it was detected from. In those regards, we are starting to reach out and work with cable operators.
Does the rise of switched digital video (SDV) play into this?
(SDV) increases the number of assets moving simultaneously, and it gives you a little more traceability as to where those assets are propagated, but some forensics would be needed to reverse engineer to determine where content might have been intercepted.
I don’t want to overstress watermarking, but I think it’s an area where programmers, content owners and cable operators are going to ultimately sit down and have some serious discussions about what the content and integration mechanisms are going to be for this level of forensics because they’re backed up by a pretty substantial economic opportunity. By that I mean early release content.
Who’s driving this discussion?
Content owners looking at a changing distribution market. In a green-conscious world, there’s an added benefit to moving content without any physical assets like tapes and release prints. An electronic file that requires reusable storage is ecologically quite sustainable.
With files, you give yourself an opportunity to maintain and control quality and security. But you also expose content to the enormous potential for file sharing, particularly as we compress what might have started out as a 40 or 50 GB AVI file down to 8 GB, and it’s a magnificent piece of high-definition, early-release content. We might get our hands on it before it’s even out of the theaters and have to move it across the network. That has to be very well-protected against interception and redistribution. That starts with the owners, ends with us.
So where are we with MPEG-4?
Here’s some context. Way back when we first did MPEG-2 VOD in 1998, we invited our studio colleagues over to TVN’s NOC to look at different encoding rates. We were doing very high-quality scene-by-scene encoding and trying to reconcile what cable operators needed to move eight channels down a QAM against what the studios would be willing to accept. It was a very interesting and informative experience.
For the past year or so, we’ve been undergoing that same, friendly discourse with our counterparts at the studios and the content providers, looking at MPEG-4 and high-definition MPEG-2, and a whole new set of criteria.
The goal is trying to reconcile the very divergent needs of a cable operator who wants as low as possible data rates with some minimum measure of quality against a studio whose orientation might be more toward quality first, data rate be damned. That’s tough to reconcile, but we always manage to do it.
In several ways, it sounds like you’re moving into a new phase of preemptive and constructive conversations.
Yes, on the forensics side, we’re introducing that to cable. And on the MPEG-4 and high-definition side, we have studios coming in and saying: "The broadcasters are running their high-definition channels at 19.4 Mbps. What are you doing?" And we say: "Well, we’re trying to be cable compatible, and we’re looking at scene-by-scene encoding at 15 Mbps as opposed to casual encoding at 19.4. Do you see a difference?" And they say, "No, I really don’t."
On the high-def side, and this is my particular view, for the first time we have receive and display mechanisms that are superior to most of the transmission parameters. In other words, in standard definition, a video camera out of a studio is probably capable of 720 pixels per line. A standard-def TV set can only discern about 300 to 400 horizontal resolution lines, so for the most part, we’ve always been able to drop the resolution down to half D-1 resolution, which is about 352 pixels per line. Now, however, in high-definition, if a TV set’s spec says 1,920 by 1,080 resolution, that’s like a medical machine. If you try to cheat on the distribution end and lower the resolution a little bit, it’s going to be noticeable. So you’ve got to deliver spec. And that’s been an interesting dichotomy: trying to save bandwidth and still deliver high quality, or avoid what I call low-resolution, high definition television.
Any update on the HD-VOD club? Was it the second transponder that really got this going?
We filled up the first one with about 3,000 hours of standard definition MPEG-2 content. We lit up a second transponder earlier this year. We’re working with Intelsat and SES space segment, taking portions of each transponder as our throughput grows. When we started distributing high definition, we started distributing at 19.4 Mbps. Over time, we brought that down to 15 Mbps, which is the CableLabs safe harbor, and which seems to reconcile pretty closely with the cable QAMs. So the ability to illuminate that second and third transponder without compromising the first with these big files was how we moved into high definition.
With the HD-VOD club, what we anticipated was we’d probably get a lot of questions from cable operators as this thing started to flesh out. And clearly that has been the case. "What’s your data rate? What kind of DRM? How about the digital outputs on my set-top boxes?" All those kinds of things that really well-informed and curious operators had questions about before they made any decisions.
What about your VOD Complete solution?
It’s gotten a little more activity than we had anticipated. It’s the kind of thing we did back in my days at HBO when we first went on the satellite. What I learned then was that this kind of service really made a lot of sense both to our affiliates and the network. It also made a lot of friends. The cable operators have their job, and you have yours, and if your skills can be complementary – aid in making their life a little easier – great. We brought that same ensemble-solution idea to this VOD Complete program for smaller operators, and it seems to be paying dividends.
What’s the size range for operators?
We’ve never set a top end. We’re actually talking to systems as small as 300 subscribers. However, if you have a small system, as few as 1,000 subs, and the right architecture, we can integrate VOD Complete.
You’re serving as president of the Society of Satellite Professionals International. Is there any crossover with that group and cable’s technical teams?
For the most part, I don’t see a great deal of discrete cable membership, but there should be. We have common concerns. Right now, broad membership is on the content side, from virtually all of the major content providers, like HBO, Starz Encore, CNN, Turner, TVN, so on. There’s heavy participation by content providers, networks and studios; heavy participation by academics and by the satellite operators themselves, SES and Intelsat. If a cable operator like Comcast and Time Warner operate a satellite facility, they have members. But even a receive-only operator should belong.
What kinds of things does the SSPI do?
SSPI is a nonprofit, member-benefit organization. We concentrate on education and issues of common interest to the science and commerce of satellite communications. But we do work behind the scenes as well, informing our members of pending regulations and advances that are important to their careers or their companies. We stress education in the satellite space and encourage bright young people to go into the field through young leadership recognition, scholarship programs, stuff like that. Good work, good info, good fellowship.
Any particular issues you’re tracking that might be relevant to cable?
Absolutely. For example, most cable operators get their content from C-band satellite. Right now, there are issues where land mobile services are trying to recover some C-band spectrum because its propagation characteristics are extraordinarily good and solid.
In fact, a lot of that traffic might be backhauled by C-band. That’s something that could affect a cable operator. So could interference. We can’t predict what’s going to interfere if there are C-band unlicensed devices. Is that going to affect your cable system and cost money to fix? Understanding and illuminating these issues to our membership is one thing that SSPI is involved in. Cable operators should be SSPI members, yes.