Paul Shen, founder and CEO of TVU Networks.
This summer I had a chance to sit down with Jimmy Goodmon Jr. in his office at Capitol Broadcasting in Raleigh, N.C., for a face-to-face interview on ATSC 3.0.
During the Q&A, Goodmon revealed WRAL planned to begin offering HighSchoolOT Live — an NFL Red Zone-like over-the-top service to viewers of Friday night high school football games at the beginning of this season.
(My colleague Paul Greely reported on the rollout in “WRAL Revolutionizes High School Football Coverage.”)
As broadcasters like WRAL look for ways to exploit the opportunities OTT — and eventually expanded OTA channel count in an ATSC 3.0 world — enable, keeping production costs in line will be a priority.
I remembered a brief conversation I had at the 2016 NAB Show in Las Vegas, with Paul Shen, founder and CEO of TVU Networks, about a new product his company was introducing that allows TV networks, stations, schools, production companies and others to sidestep the expense of sending production trucks and full production crews into the field.
The product, TVU RPS, which stands for remote production system, allows a broadcaster to send just cameras and camera operators into the field and contribute synchronized camera feeds via IP over the conventional internet to a centralized control room — presumably back at the station — where switching is done and graphics, effects, text and other production elements are added.
As Greely reported in his article, WRAL is not using TVU RPS for its production of HighSchoolOT Live.
But it seemed to me like this is sort of application the product is designed for. So, when I had a chance to interview Shen right before I left for Hollywood, Calif., to cover the 2016 SMPTE Technical Conference & Exposition, I took it.
I wanted to find out if broadcasters are in fact beginning to use RPS to increase the number of games and other live events they produce, how the product has been performing and where it is headed.
An edited transcript:
TVU Networks unveiled TVU RPS at the 2016 NAB Show in Las Vegas. How have broadcasters responded?
Consider one of our major customers. I can’t mention their name, but they say every year they have the rights to over 5,000 games. Maybe they can use 200 to 300 games at the most. Most of the content is not produced because the cost is too high.
Another application is for broadcasters who want to produce over-the-top channels to cover events like high school football.
Consider lacrosse. It is one of the most popular games, but you cannot find it on TV. You simply can’t.
The whole RPS approach to remote production opens up new opportunities that were never available in the past because the cost of production.
Could you compare the cost difference between the conventional approach and RPS?
Today, for any TV station that wants to do a production in the field, there is the truck cost, the cost of all the equipment and the crew.
The time traveling to go to the production and back is generally two to three days. Sometimes it is way more.
But imagine now the production crew does not have to go out into the field. All you need to do is send the camera operators. Now, the production crew you once sent out for two to three days can produce more than one event in a single day.
Today, this is what the industry needs. It needs more content.That content will find audiences, but broadcasters must produce it at much lower costs. RPS fundamentally solves that problem.
Have you also seen RPS used for news and entertainment programming?
Of course. One TV station has used RPS for a vice presidential debate.
How tight is the sync between RPS contributed cameras, and is that difficult to maintain over the public internet?
This is one of the most important parts. To make remote production happen, we need to make sure all of the cameras are in perfect sync — within one frame.
Once they are in perfect sync and they can be transmitted in a bandwidth-efficient way, we have a solution. That’s what RPS does.
The technology behind what we have done is network management, error correction, and how efficiently we use the network. All of that plus a very important component: the synchronization of video.
They are in perfect sync and will never offset. So, when we switch from one camera to another, you will never go back in time or into the future for that matter.
So, one RPS can stream back four synced cameras?
Yes, but we can go from four to six or even eight as an option. At this point, the first product we’ve gotten into the market does four cameras in perfect sync. We believe with six or eight it is not an issue.
For most small productions, four cameras are sufficient and for the big games, eight are enough.
What data rate is supported for HD?
It is over 10 Mb/s. Each of those [camera feeds] is being sent as H.264 at 10 Mb/s. That is way beyond most broadcast quality. It’s better than satellite transmission.
And this is being done over the public internet, not some special leased bandwidth or fiber connection at the venue?
The conventional internet. That’s our core technology. That’s what is most important to reduce operational cost.
Is 4K possible?
We are going to put a 4K mode in there.
Are you using your own technology to maintain sync among the cameras, or are you relying on something like PTP — Precision Time Protocol?
Currently in our system we are using our own protocol. The reason we use our own is because it’s supposed to end up in our equipment. So, it is crucial to use our own protocol because it gives us more precise synchronization.
But once it arrives at our receiver, we can push it back out into other IP-based switching systems. We can embed another timing protocol on top of that and do perfectly synced switching.
Another very important part for us in the transmission is error correction. Our signal is transmitted over an unconditioned public internet. We have to make sure it is properly handled given the uncertainty of the network — network fluctuation and bandwidth packet loss.
So we need our own protocol to handle those network conditions.
What do you think about all of these IP protocols being developed for media?
The beauty of the IP protocol is that from one protocol to the next the cost is beginning to be very low. It is no longer in the physical layer which made converting from one to another very expensive.
Today, to convert from one protocol to another has almost no cost. Software can handle all of those conversions.However, everyone’s format needs to be open. Once the format is open, it is easy for others to integrate.
For example, in the past in the software industry it was very expensive for one piece of software to integrate with another. But once people adopted an XML interface it became very easy because it is a generic syntax. You can pass any data you want.
The same is the case with IP video. It’s not essential that protocols have to be identical, but they have to be open. If they are open, it’s easy.