Interview: Dr. John Abel, President of DataCast, on Datacasting and DTV's Business Future — Part 1

by Dale Cripps
Tuesday, January 19, 1999

Interview with Dr. John Abel, President
DataCast Corporation
Reston, VA

Interviewed by Dale E. Cripps
HDTV News Online

I recently talked to Dr. John Abel, now immortalized in Joel Brinkley's book Defining Vision as the father of HDTV. John held the post of vice-president of the National Association of Broadcasters from 1986 through 1996. He distinguished himself at the NAB as the major force in the metamorphosis of broadcasting from its traditional roots to the idea of "datacasting."

John joined Chris Craft as president of its DataCast division in late 1996. He has been busy since developing data broadcast applications. There is likely no one more advanced on the theory and implementation of this new craft as is this notable and decorated industry executive. To suggest his view, commitment, and direction he carries with him at all times numerous portable computers. He is thoroughly convinced that there is a profitable datacasting future for anyone capable of radiating bits.

I started our conversation with a focus on money—was there any to be made in data transmission, and were the venture capital people showing any interest?

Dr. Abel:
The approaching digital broadcast era has raised the interest in venture capital people, though perhaps the IPO of Broadcast.Com also stimulated the notion that there are many companies that want to stream video through the Internet. But the Internet is not a very good medium for doing that. There are between 300 and 400 thousand web sites that have some level of video on them. There are many content providers that want to stream video to end-users, including large companies like Microsoft, who want to use it for training of people on various pieces of software they sell into the business environment. Companies want to use Broadcast.Com for stockholder meetings and other things. I would presume there are many content providers that would prefer to have video, but they don't do video on the web because it is an awful medium for it. The web is not so bad for streaming audio, but for video, the style is cramped.

So these things have caused people to think that oh my goodness, there is an opportunity here take a lot of the content that is in the web and put it over DTV channels, because in DTV channels there is so much unused bandwidth, especially if broadcasters do multicasting of 3 or 4 SDTV channels. Even if they do an HDTV channel, they will have many unused packets coming off the transmitter.

HDTV News: Is the 11th broadcast commandment "thou shalt not send an empty packet."

That is totally right. It takes the same amount of power and infrastructure to support a TV station that is broadcasting one SDTV channels or one HDTV channel and they have many null packets coming off that transmitter. Those null packets can now be stuffed with some sort of data. It is also true that cable is not going to carry DTV channels in the early stages. Broadcasters are going to rely upon wireless antennas to receive DTV channels. The ancillary data component is probably not going to be carried by cable anyway even if the "must carry" rules go through. Ancillary data is not required to be carried by cable systems now.

So, it is going to be antennas on PCs. Philips announced a PC TV receiver card. Intel also has a card. LG Electronics just sent me an announcement that they have such a card. Panasonic made their announcements of an ATSC compliant card set. These cards are going to be less costly than a DTV receiver by a significant amount. Some of these are looking at a device that has a lot more signal processing on the card in a DSP variety as opposed to doing this is software. Intel is more reliant on software. These other cards with greater DSP on board could be retrofitted into older computers (233MHz or greater) with a lot of the processing done on the card—not drawing upon the resources of the machine. The PC 99 spec has the expectation of device bay that can be hot-pluggable into the PC, so the tuner cards plug into the device bay rather than having to open up the computer. That goes into the 1394 firewire connector, or even to 10Base T or 100Base T. That makes it all much easier to fit these pieces together. Whether this is going to be multi-millions of these devices, I don't know. It is certainly going to make the PC a more affordable place to send the DTV signal.

HDTV News: That would suggest that the business environment is one of the first real markets for the DTV signals.

Absolutely. That is another important point about the venture capital community. VC want to get to the consumer markets, but like many things, they start in the business-to-business markets. Something on the order of 45% of Broadcast.Com's revenue comes from business-to-business applications. The fact that you can now do other kinds of data broadcasting —catalogs with multimedia components in them—make data broadcasting...well, I am somewhat passionate about this—but to me its the only way that a broadcaster sees new revenue. I am not saying there is not a value in HDTV. I think there will be. But it is limited to certain kinds of content. It is not new revenue. It is replacement of old revenue and trying to hold on to the past.

HDTV News: What is the end game for data distribution? What am I going to do with so much data? Where is it going to go? How is it going to get to some place where I can use it?

There are a couple of different answers. In the streaming application—let's say you are using one Mb/s. We've been demonstrating that along with the 500 Kb/s video. That is infinitely better than what you can get on the web. It doesn't use any resources of the machine. It's passing right through it. Now, unfortunately, you're into real time again. You've got to watch this program when it's being streamed.

In terms of caching things to the hard drive—that's the other part of data broadcasting. You are, of course, going to occupy a lot of space on the hard drive. In the future there will be low cost memory and storage devices. I think the expectation is that we're going to have between 10 and 20 Gigs for around $2 or $3 hundred. From the projections I have seen by the year 2007 you will get hundreds of gigabytes for around $200.

HDTV News: What about the digital tape recorder that was standardized several years ago which is capable of recording and streaming back 19.4Mb/s?

Yes, that's another. I'm not a big fan of it. The whole storage area is becoming more efficient. The costs are dropping dramatically. You're going to have the opportunity to store entire television programs with just IDE or SCCSI hard drives.

HDTV News: What is not incorporated in the ATSC standard that in your opinion needs incorporation to accommodate those things clearly in your vision?

HDTV News: There is no data broadcasting standard, per se. There is a group working on it. They have never published the standard. I am of the opinion you don't need much of a standard. All you need is . . . well, let's take the streaming video for an example. If the (television) station has an encoder card that will do 2 standard channels—one they put on their up-converted NTSC—the other one can be devoted to 4 or 5 or 6 channels of streaming video and data.

All we care about then is; can the data be encapsulated in the MPEG2 packets at the transmitter site? That's easy to do. We can extract that data at the client side, bringing it to the PC bus. We just take the data out of the packets—de-encapsulating them, if you will. We don't care about whether there's a video picture on this PC or not. All we're using is the MPEG transport to get other data in there.

—Continued part 2