RF Column 31 - NAB Issue 1994 Copyright (c) 1994,1995 H. Douglas Lung ALL RIGHTS RESERVED TOPICS: What happened to HDTV this year? Technical sessions reveal Grand Alliance system details Compromises required in digital transmission systems "All or nothing" coverage concerns broadcasters Dr. Oded Bendov: HDTV coverage will vary by channel Transmitter manufacturers planning for HDTV Digital video compression systems for satellite transmission ------------------------------------------------ Most of the excitement at NAB this year was generated by the widespread introduction of digital video recording on inexpensive hard disks. The trend to digital hinted at last year was in full swing this year. I didn't see anything as exciting in RF technology, although the widespread use of digital video will have an impact on video transmission. Wandering the floors, you might have wondered what happened to HDTV? A hot topic in previous years, it seemed pas'se this year. From a broadcaster's viewpoint - the big news was out: The Grand Alliance was reality and we finally had a transmission standard. Chips from C-Cube, Thomson and others opened the market for satellite video compression. While General Instrument's Digicipher 1 system remained the leader, Scientific Atlanta, NTL and C.L.I. all had more advanced "almost" MPEG-2 systems on display. As in past years, there were some useful nuggets of information (other than sales pitches) in the technical sessions. I suspect one of the reasons HDTV wasn't as hot a topic on the floor this year was because there wasn't anyone pushing it. I'd gotten used to seeing Sarnoff, Zenith and General Instrument, along with a few others, pushing their flavor of HDTV. The Grand Alliance eliminated the need for each company to push its own standard. The "HDTV - Under Construction" logo adopted by the Grand Alliance set the tone. While HDTV wasn't a topic on the convention floor, it certainly was THE topic in the technical sessions. The most popular session Sunday morning was titled "The Grand Alliance HDTV System". The session started out standing room only and tempers flared when convention officials tried to move attendees (including some powerful people in the HDTV industry!) standing in the walkways. I'll touch on some of the RF aspects of the "Grand Alliance" (G.A.) system. For more in-depth coverage of the details, see Merrill Weiss' column or the copy of the system specifications in the back of the NAB 1994 Engineering Conference Proceedings. By now you've probably heard that the modulation method selected for US HDTV was 8VSB (8 level Vestigial Sideband with a pilot carrier). The original Zenith proposal specified 4VSB. I was concerned that increasing the complexity of the signal might also reduce the reliable coverage area. Wayne Luplow, who tackled the transmission side of the G.A. system, said no. By using 8VSB instead of 4VSB and using Trellis coding, the G.A. was able to achieve a 1 to 1.5 dB improvement in signal to noise performance. On the transmission side, the system will use a pre-equalizer filter to compensate for ripple in the passband and frequency roll off at the edges (from either the transmitter or antenna). On the receive side, an NTSC rejection filter sets the response of the system so that the receiver has nulls at NTSC carrier frequencies (video carrier plus chroma and aural subcarriers). The net result -- a payload data rate of 19.3 MBs (megabits per second) with a carrier to noise threshold of 14.9 dB. During his presentation Mr. Luplow flashed a slide on the screen outlining some of the latest RF system test results. I tried to copy down as much as possible as fast as possible, so bear with me if you find a typo or two. The desired to undesired ratios of the 8VSB system looked good - for NTSC to HDTV interference the ratio was 2.1 dB. For HDTV to NTSC a higher ratio of 33.8 dB was required and for HDTV to HDTV the ratio was stated as 15.9 dB. Single static echo (ghost) rejection was good - ranging from -1 dB for a one microsecond delay to -2.9 dB for a 15.0 microsecond delay. The system worked with impulse noise bursts up to 190 microseconds in length. During this test the payload data rate was 18.8 MBs. While he didn't mention it in his talk, I appreciate Mr. Luplow displaying the peak to average power ratio. Such information was difficult to nail down last year. The ratio for the 8VSB system was 6.2 dB, in line with my expectations. When can we expect to see some field tests? The VSB subsystem is supposed to be integrated by this summer. By first quarter 1995 the full system should be ready for field testing. During the question and answer portion of the session engineer after engineer expressed concerns about HDTV coverage. If you've read this column before, you know that digital HDTV coverage is an all or nothing proposition. The concept of City Grade and B grade mean nothing. Worse, there is a fear that consumers will do the minimum necessary to get a picture. After all, any improvements in the antenna system after threshold is reached would yield any noticeable improvement in the picture. No improvement, that is, until a heavy rainstorm or transmitter problem reduces marginal signals below threshold. I questioned the panel about this problem and the answer I got was that was something broadcasters and the set manufacturers would have to deal with. Manufacturers will have to provide some sort of signal level or bit error rate bar graph display on the TV. Broadcasters will have to teach viewers what levels are acceptable and what steps to take when the picture disappears. Last year two of the HDTV system proponents were pushing "soft fail" mechanisms in their systems. Zenith's system had high priority bits encoded on the higher modulation levels. The Advanced TV Research Consortium had a separate, lower data rate, higher power "priority channel" to carry essential data. The robustness of the ATV-RC's priority carrier was demonstrated in the tests in Washington DC when the HDTV signal was receivable at locations where a comparable power NTSC signal wasn't. I asked the panel why the G.A. system didn't have soft fail. I detected a bit of frustration in the response - this after a line of questions from other attendees questioning system robustness in real world conditions. The answer was that in surveys broadcasters said "soft fail" was a low priority and if it was a priority, broadcasters had their opportunity to push it and did not. Delving a little deeper into the subject, the real answer is "there is no such thing as a free lunch". Devoting system resources to providing two levels of service would have reduced the payload data rate, requiring more data compression or less error correction or both. It would also increase the cost of the receivers. While I liked the idea of a soft fail mechanism, the problems of viewers with minimal antennas or at marginal locations would remain. Because the soft fail systems would have extended coverage, the number of viewers experiencing loss of picture due to small variations in signal strength would be larger! Perhaps a single mode system is best. When digital HDTV was first proposed a few years ago, Oded Bendov from Dielectric Communications was one of the first to question digital HDTV proponents' claims that smaller transmitters would be needed for digital HDTV than analog NTSC. Now it is generally accepted that peak radiated powers for full coverage digital HDTV will be the same as those for NTSC. This year Dr. Bendov raised a new concern in a technical session titled "The Effect of Channel Assignment on Transmitter and Receiver Requirements for Equivalent HDTV/NTSC Coverage". He concluded that outside antennas with low noise amplifiers would be required for reliable HDTV reception at distances comparable to NTSC. Furthermore, channel choice would have a direct impact on coverage area. Stations operating on channel 69 are operating at a 5 dB disadvantage compared with stations on channel 14. This disadvantage is due to higher losses in free space transmission and higher losses in receiving antenna lead in cable. This 5 dB difference isn't significant now - viewers might notice a bit more snow in the picture. With a digital transmission system there is no graceful degradation. Drop below the carrier to noise threshold and the picture disappears. If the FCC takes the same approach with HDTV as it has with UHF NTSC TV, stations operating on higher channels will have a distinct disadvantage. Outdoor antennas and low noise amplifiers can reduce this disadvantage by offsetting cable losses. Write or call Dielectric Communications for a copy of Dr. Bendov's paper. You may disagree with some of the conclusions, but its tough to argue with his numbers. Let me leave the technical sessions and go back to the convention floor for a few closing comments on HDTV at NAB 1994. Transmitter manufacturers are planning for HDTV. Comark continues to refine its IOX line of HDTV ready transmitters. Harris, after a late start with IOT (Inductive Output Tube) transmitters, is catching up. Many VHF transmitter engineers kept hoping for a solid state UHF HDTV transmitter that would be as reliable and maintenance free as the current line of VHF solid state transmitters from Harris and Larcan. Those hopes seem to be dashed, except for the possible development of a silicon carbide high power transistor from Westinghouse. Of course, those transistors would operate at temperatures high enough to melt solder, would require pre-heating and probably won't be available in the 120 KW peak power range anytime soon. I take Larcan's merger with TTC and their display of a Larcan high power IOT transmitter as an indication tubes are going to be with UHF engineers for some time to come. PESA's solid state 1 KW 8VSB transmitter looked great, however, that 1 kilowatt was PEAK power - you'd need at least 30 of these for decent HDTV coverage in most markets - more if you want to reach the suburbs or the sets with indoor antennas. While my column is on RF, my job requires me to keep up to date with studio technology as well. The current trend in studio video recording is going the opposite direction of HDTV! The hot items on the floor this year were digital disk recorders, most using some form of video compression. Data rates were well below those necessary for HDTV. While heavy video compression would permit some of the systems to work as HDTV recorders, this could pose problems for the broadcaster, since about all that's known about repeated compression and decompression of video (concatenation) is that it is bad! Considering the rapid advancement of digital data storage and manipulation, this limitation may be gone by the time we see terrestrial HDTV transmission. On the other hand, I'm hearing more talk of using the digital data stream broadcasters will receive for HDTV for "multicast" NTSC instead. This was broached at the Grand Alliance session. The main hurdle seems to be regulatory. Once the transmission system is finalized and the channels assigned, that 19 Mbs. of data could be an HDTV picture, multiple NTSC pictures, interactive "Beatles' CDs on demand" music or lists of bad credit cards. In the digital world, it doesn't matter. Scientific Atlanta, N.T.L. and C.L.I. had satellite video compression / decompression equipment at the show. All claimed they used MPEG-2 or near MPEG-2 compression standards. Actually, the MPEG-2 standard was not fully defined at NAB - there was an MPEG-2 committee meeting going on in Europe the same week as NAB. Most of the specifications are set and chip manufacturers like C-Cube and Thomson have jumped ahead of the MPEG-2 committee by shipping chip sets matching the current state of MPEG-2 standards. Since these are computer controlled devices, minor changes can be handled with software. General Instruments was not pushing (and perhaps did not have to push) Digicipher at the show -- their Digicipher 2 product due out later this year includes the same "MPEG-2" functionality the other vendors showed. I found Digicipher 1 receivers hiding at the PanAmSat booth. Toshiba had one of the niftiest fly away packages I've seen. It used G.I.'s Digicipher SCPC (Single Channel per Carrier) video compression equipment and a Toshiba exciter and 35 watt final amplifier. In Japan, the system works with domestic satellites using dishes under a meter in diameter. For Intelsat approval a larger dish - up to 1.8 meter - will be required. I'll get back to more current day / current budget topics for the next column. I'm hoping to have factory test data on one of the nicest looking, reasonable cost LPTV transmitters I've seen recently and maybe even some real world satellite video compression test results from Telemundo's Scientific Atlanta installation (now scheduled for April). If your network or program source is switching to digital video compression, you're probably wondering if you need to buy one of those phase locked LNBs. The answer is maybe. More details next month, along with some unrelated tips skimmed from the NAB, phone calls, faxes and EMAIL. If you want to phone, fax or EMAIL, the appropriate numbers are 305-884-9664 (voice - after 6:30 PM Eastern), 305-884-9661 (Fax - anytime), CompuServe, ID Number 70255,460 or, via the Internet, 70255.460@compuserve.com. Mail should go to me at 2265 Westwood Blvd., Suite 553, Los Angles, CA 90064. Copyright (c) 1994,1995 H. Douglas Lung ALL RIGHTS RESERVED