IS KQED THE STATION OF THE FUTURE?

While some are thinking about building the TV station of the future, Ethan Bush has been doing it.

As a senior project director for National TeleConsultants, a leading TV systems integrator, Bush has been deeply involved in transforming noncommercial KQED San Francisco from, as he puts it, “a very labor-intensive, manual, single-channel analog operation to a very highly automated multichannel digital HD operation.”

Northern California Public Broadcasting’s flagship station is now capable of centralcasting not one, but six, channels of digital programming to its sister stations in San Jose (KTEH) and Watsonville (KCAH, serving Monterey and Santa Cruz) and of HD production with a tapeless, file-based work flow.

Why six digital channels? In addition to the HD channel, KQED also produces and sends downstream to the other stations a digital version of the main analog signal; Life, a how-to and lifestyle channel; World, an international channel; Encore, which repeats the previous night’s primetime programming; and Kids, a channel for school-age children.

The stations have a deal with the local cable operators for carriage of all the channels and it can also broadcast them when it is not broadcasting HD.

The station is not only multichannel, it’s multiplatform. The workflow for the production of Quest, the popular science program, has been designed to deliver not only a conventional TV show, but also content for the radio and for the Web.

In this interview with TVNEWSDAY, Bush talks about the KQED project and what lessons it might hold for commercial stations.

Do you think centralcasting makes sense for commercial stations?

Well, certainly for station groups, particularly if they’re sharing programming, there are opportunities to reduce costs by centralizing their operations.

Most of the time when we’ve analyze centralcasting for groups, we have found the interconnect costs often were very, very high. Sometimes, we would convince them not to go to a centralcasting model. But, over the years, we’ve seen the cost of interconnect dropping, so it’s becoming more and more affordable.

There are a number of models developing that would allow one facility to perform all the common ingest and then push content either out to remote servers or actually originate it centrally.

At many commercial stations, local news is the premiere broadcasts. The rest of the day is pretty much routine play to air. So I think as long as the local stations can maintain their community identity through their local newscasts, upper management is willing to look at reducing some of the labor necessary for the day-to-day playback of the daily programming.

So you think one way to centralize would be on the ingest side?

That’s certainly one way. Then it’s a question of whether the group wants to have play to air at individual stations or centralized from one location.

From talking to some of the broadcast engineers, the idea of centralcasting seems to have fallen out of favor for the exact reason you just cited—the cost of the interconnection.

The Corporation for Public Broadcasting and PBS have been telling stations to really try to improve the overall efficiency of public broadcasting nationwide, to reduce the costs of it. There are many areas, in San Francisco and Los Angeles and New York certainly, where there are multiple PBS stations with overlapping coverage and they wanted to encourage the stations to combine operations wherever possible.

So in that case, it was kind of a push from high above down to the stations. Many stations have resisted it because they don’t want to lose their individual identity.

I also noticed that you went with the Sony XDCAM for field production. Was that your call?    

That was not our call. That was the station’s call and a lot of it had to do with the incentives available to them for packaging of a variety of products that they needed for HD upgrades.

How does XDCAM fit with what the station is doing?

KQED is using XDCAM in the field, particularly for one of its productions called Quest. One of the advantages of that is it’s an optical disc recording medium, which is relatively inexpensive compared to some of the alternatives and also offers some compatibility with their post production process through Avid.

They are also people who like to have something on the shelf. Even though the discs are re-recordable, they keep them in their libraries just as a production backup. You can do that economically with discs. You can’t do that economically with some of the other solid-state media.

Some of the commercial broadcasters have been experimenting with Apple Final Cut Pro and other off-the-shelf software and gear. Is there any particular reason why KQED stuck with Avid here?

I guess they remained with Avid because they had them in there for so many years. They also had storage systems that they would have had to abandon if they wanted to switch to Final Cut Pro.

KQED’s Quest program is touted as a multimedia production—TV, radio, the Web. As a practical matter, what does that mean?

Quest was fairly innovative in that they’re really trying to tie together all the various media channels available to KQED during the course of production.

While they’re in production, they make special arrangements to accommodate the different distribution models. For example, depending upon the particular shoot, the camera crew may pick up material specifically for use by the Web site. The audio people may record specific sound bites or ask specific questions for the radio and they may also even record separate content for their educational Web site.

So they try to anticipate how each production can make best use of the various distribution media. They make sure that while they’re out on these shoots they cover the shoot appropriately for each of the different media and for the educational platform.

What is the educational platform?

They’re offering service schools access to some of their content in a raw form. So, basically, with a low-cost editing system, a teacher and her students can build their own little video segment from KQED’s raw content.

And it’s distributed over the Web?

Yes, the schools get it as a low-res digital distribution over the Internet.

One of the other things that the commercial guys are doing more and more is going to automated production. Is that part of the KQED setup?

No. KQED doesn’t do a daily newscast per se. They have a weekly newscast called This Week in Northern California. Most of their other productions though aren’t as repeatable as a newscast. They don’t use robotic cameras. They don’t use a newsroom computer system.

So it’s just not economical to go there?

It’s not that. It’s just that such productions don’t really lend themselves to preset shots that are used by robotic cameras. That and I think also it’s also a union station so there’s probably some resistance on the part of the labor force.

NTC has been developing a means of interconnecting the various systems within a station called service-oriented architecture. Could you explain that?

It’s designed to provide better integration of the entire television station operational work flow from traffic, to automation, to billing, to the newsroom system. You buy a traffic system, you buy a automation system, you buy a newsroom computer system, you buy a bunch of productions systems. They just don’t interface well. They have no way of getting information from one to another, either to assist in the work flow or to monitor the work flow.

So what NTC is doing is working with a variety of vendors of the individual systems to see whether they are compliant with this newer service-oriented architecture model of software development.

Well, isn’t this one of the selling points of Harris and some of the other big vendors. You know, “Come to us. We’ve got all the pieces, we’ll put it together and we’ll make sure it all works.”

That’s the basic idea.

I guess the idea behind the SOA approach is that you could use multiple vendors.

It’s not often that one vendor’s solution really fits everybody’s application. So at NTC we don’t treat every job as a cookie cutter job. You know, “Here’s the Harris solution that we offer and we support. We’ll put this in for you.” Instead, we work very closely with our clients to really understand what their needs are, what their unique requirements are, and, more often than not, they don’t fit into a single-vendor solution.

But then you’ve got to make sure it all hangs together, right?

Yes, we have to make it all hang together, but, more important, we have to make sure that what we’re putting together really serves the clients needs. I don’t know how many times stations have been sold a single-vendor solution that doesn’t integrate even within itself. I have heard horror stories of one particular vendor whose products just don’t work well among themselves.

How do you approach new jobs?

We try to work kind of from the top down. Rather than sitting with engineers, we tend to try to hit it from a higher level and sit down with their upper management and find out what the stations goals are and also what issues they have. Then, we try to work down towards what’s the appropriate technological solution. Often, engineering’s perspective and management’s perspective aren’t in alignment with each other.

What’s your advice for stations looking to overhaul operations as KQED has done?

Let’s say a station is going to incorporate a newsroom computer system, make that leap. It takes a lot of forethought and planning. They really have to understand their work flow, understand how not to repeat the mistakes of their current work flow, not to force their current work flow onto a system that has alternative work flows that may be an improvement.

In other words, don’t carry forward your mistakes, your current work-arounds, into the new technology. A lot of people have the tendency to do that. Look at the newsroom computer systems that are out there. See how they can adapt to your requirements and also see how you can take advantage of their flexibilities and their new methods of working. But it does take a significant effort to really understand the complexities of these systems and how they can be used.