Pinnacle Systems

As the Great One says, skate to where the puck will be

Wide-area-network connectivity is one of the major challenges for station groups interested in centralcasting, but Al Kovalick, CTO of Pinnacle Systems' broadcast solutions division, says the same attitude that made Wayne Gretzky the greatest hockey player ever can help broadcasters stay off the thin ice.

"Like Wayne Gretzky says, don't worry about where the puck is, worry about where it's going to be," he advises. "And that's the way you have to look at WAN technology and its pricing." Those planning to adopt centralcasting should find out when WAN technology will be attractively priced for them and then look to have any facilities completed in the same time frame.

No doubt anyone interested in investigating WAN technologies is also sorting through operational models. Kovalick says he has developed three basic approaches to centralcasting. The first is a group that wants to do 100% centralcasting, as in the model the Ackerley Group chose. He says whether that is the best approach depends on how much of the operation a group can get under one roof as well as what kind of deals can be reached with transmission providers on bandwidth, quality and guarantees of service.

"Some of these stations have two DS3 lines from the hub to the remote so, in case one line dies, the other one is available," he explains. "It also has to do with geography. If you have one station in Maine, one in Florida and one in California, you probably won't want to centralcast."

The other model involves content's being prepped in one place and then sent to remote locations. But there is no streaming occurring from the central location. "It's all playing out from the station, but segmentation and generation of promos is done in a central location."

The third model is a hybrid, incorporating a bit of both of the other models.

Kovalick says that centralcasting is not a solution for every station. Independent stations in small markets dominated by UHF stations probably won't ever find a great need for centralcasting. And stations in markets that can hire employees for less than $10 an hour might be hard pressed to see the advantages as well.

But, for those stations that do find themselves pulled towards centralcasting, he offers this advice: Engineering should work closely with the IT department given the use of WAN technologies that play a role in centralcasting.

"If you're in the IT department and you're going to purchase WAN technology, you're going to sit down, figure out what you want, write out an RFP and talk to at least three or four carriers," he says. "Broadcasters don't know WAN technology, so they can learn a lot from the IT department. And until people do that, they're living in a fantasyland thinking it's just going to work."

With WAN technologies helping interconnect far-afield facilities, it will be the storage-area networks (SAN) that will help get centrally stored material out of the hub.

"We now have a SAN-based architecture for our Media Stream server," says Kovalick. "And then you can have 50 to 100 channels with no single point of failure in the system. We're also working to get an archive system included so it's a simpler buy for people."

Kovalick is involved with developing the Media Exchange Format (MXF) standard, something he says will help make it easier for broadcasters to use different servers within the same WAN.

"MXF has at least six months to go before it becomes a SMPTE standard," he says. "It will help in the file-transfer world so that groups can more easily prep material at one site, on one type of server, like a SeaChange, and send it to another site and another type of server, like a Pinnacle or Grass Valley server."