CHAPTER

An Introduction to HD Radio Technology

1

1.1  What is HD Radio?

HD Radio is the trademark for iBiquity Digital Corporation’s in-band on-channel (IBOC) digital radio system. While there are differences between amplitude modulation (AM) and frequency modulation (FM) band HD Radio systems, an HD Radio signal can be generally described as a digitally modulated RF signal that is transmitted around, under and along side the present-day analog AM and FM signals. It should be noted that, strictly speaking, a hybrid HD Radio signal actually has two components – an analog modulated component (the legacy AM or FM signal) and the digitally modulated component referred to here. These digital signals are composed of multiple orthogonal frequency division multiplexed (OFDM) subcarriers, which are transmitted at a level to meet the specifications of the RF masks (AM and FM) as mandated in the United States by the Federal Communications Commission (FCC), and as specified in the digital radio broadcasting standard (NRSC-5-A) of the National Radio Systems Committee (NRSC). Since the OFDM subcarriers of the HD Radio signals are contained within these masks, and are therefore considered to be contained within the allotted channel for a given station without allocating any additional spectrum, it is considered to be an “in-band on-channel” system.

Eventually, when the penetration of HD Radio receivers in the marketplace is sufficiently high (over 85%, for example), the broadcast industry and the FCC will likely support the rollout of “all digital” IBOC which will allow the digital subcarriers to be moved into the area presently occupied by the analog signal, and the power level of the digital subcarriers could then be increased in an all-digital environment, improving coverage and robustness. Since the FCC has not yet authorized all-digital operation, this book will be concentrating on the hybrid mode of IBOC operation which is presently authorized and being transmitted by thousands of broadcasters, with more going on the air every day.

The audio in the HD Radio system is input in Audio Electronic Society-3 (AES-3) digital format. The digitized audio is put through an iBiquity Digital Corporation proprietary perceptual audio encoder called “HDC” which reduces the amount of bandwidth necessary to transmit the digitized signal. The output of the encoder is used as the information which is modulated onto the OFDM subcarriers.

To listen to the HD Radio signal, a new radio which can decode the HD Radio signal is required. When using a new HD Radio receiver, if the listener is tuned to an analog-only station, he or she will hear the normal analog signal. If the listener tunes to a station transmitting an HD Radio signal, the radio will first demodulate the analog signal. Once it has acquired the digital signal, the radio will then “blend” over to the digital audio, and the listener will be hearing the digital audio. If the radio should encounter a problem with the digital signal, it will immediately blend back to the analog signal. In this way, the listener will not lose the station and will hear audio (either analog or digital) continuously.

The FM HD Radio signal has more spectrum space available than does the AM HD Radio signal, as the FM channel has been allocated greater bandwidth. Therefore, the FM HD Radio signal can operate at a higher data rate than the AM HD Radio signal. This greater data rate can be subdivided to allow additional audio channels (or advanced data services – see below) to be transmitted on the same frequency. As an example, a station would transmit its main channel audio programming (which is broadcast in the analog signal) on their HD-1 primary channel. They could add a second, digitalonly audio channel (call it “HD-2”) and transmit a completely different program format on that channel. This is called “multicasting” and can only be done on the FM HD Radio system at present, though there is talk that it may be possible with the AM system, giving one high quality monophonic channel to the primary HD Radio channel and a lower quality monophonic channel to a secondary channel. The FM multicast channels can be stereo if this is what is desired by the station, and if enough bits-per-second are available to support stereo operation on these channels.

The FM HD Radio signal can also be split to add advanced data services, such as the transmission of stock market quotes to various data devices, or the ability of the station to transmit album cover art to a listener’s radio. This adds a new dimension to the listener’s experience, and/or increases station revenue if this data capacity is leased out.

There are companies presently working on a way for an HD Radio receiver to tie into a listener’s cellular phone so that, if the listener liked the song he or she was listening to, he or she could press a virtual “buy now” button on his or her cell phone and order the song or album. Yet another innovation that is in its infancy is “iTunes tagging.” With iTunes tagging,a listener has the ability to “tag” a song he or she has heard on his or her HD Radio. The HD Radio would have the ability to download this tagging information to the listener’s iPod or computer, and the listener then would have the ability to purchase the song or download online at a later time.

Additionally, title and artist data, called program associated data (PAD) or alternatively, program service data (PSD), is transmitted in both the AM and FM HD Radio systems. Analog FM already has the ability to transmit this information via the radio data systems (RDS), which operates on a 57-kilohertz (kHz) subcarrier in the FM baseband. PAD will be a new experience for AM HD Radio listeners, as their radios have the ability to display title and artist information for the first time.

The listener’s experience for AM HD Radio can be quite startling. The first thing the listener will notice is that the “muffled” sound of traditional analog AM radio will no longer be present. When an AM HD Radio receiver blends from the analog audio to the digital audio, the frequency response opens up from the 3.5 to 4.5 kHz audio passband of a typical analog AM radio to a frequency response of 50 Hz to 15 kHz. The digital audio, in addition to being easier to listen to because of the greatly expanded frequency response, will also be in stereo, with a separation exceeding 70 dB. The audio will be in stereo within the primary digital coverage area of the AM station; it will blend to digital mono and then to analog mono as the listener moves farther away from the transmitter.

The next thing an AM HD Radio listener will notice is that the buzzing, whistling, humming, and audio fades of analog AM radio will no longer be there. The AM digital audio signal is very quiet. When a station has a silent spot in its audio, it is truly silent. If the listener is driving and passes under a bridge, the audio does not drop out to a burst of noise. The listener simply hears the digital audio. There are, however, limitations to this. If the bridge is too large it is possible that the AM signal will either become too weak or distorted, and the HD Radio will quickly blend to digital mono, then to analog mono, and the listener will hear some noise. It is also possible that the digital signal will be fully listenable while the vehicle is passing under the bridge and play out its audio buffer, then quickly blend to analog mono after the vehicle emerges from under the bridge. In this case, the listener will hear an analog signal briefly until the HD Radio blends back to digital audio. Overall, the listener will find AM HD Radio a much better experience.

The FM HD Radio listener will not notice as dramatic a difference going from analog to digital as will the AM HD Radio listener. The frequency response of the analog FM signal is 50 Hz to 15 kHz. The frequency response of the FM digital signal is 50 Hz to 20 kHz. In most instances, audio subtleties will take up the additional audio spectrum space and many listeners will not notice; some, especially us older folks, simply cannot hear up to 20 kHz.

The FM HD Radio digital audio does not use audio pre-emphasis as does the analog FM signal. Pre-emphasis is utilized to boost the high frequencies at the FM transmitter by 17 dB at 15 kHz using a 75-microsecond (µs) equalization curve. This is done to suppress the inherent noise present in the FM modulation process. This high frequency boost generally requires that clipping be used on the high frequencies on the transmit side to prevent overmodulation of the transmitter.

The receiver utilizes de-emphasis using the inverse of the 75μs equalization curve, decreasing the high frequencies as received. Depending on design factors and the quality of the parts used in the de-emphasis network, the de-emphasis may not be as accurate as it should be. This can result in audio phase shifting, in duller or brighter high frequency response than contained in the source material, and other distortions of the audio signal.

As a result of the FM digital audio not using pre-emphasis, the listener will notice that the high frequencies are smoother sounding and do not have an “edge” to them. Additionally, since the audio processing is not clipping the high frequencies to prevent overmodulation due to the preemphasis used on the analog channel, the audio spectral content tends to be closer to what is input to the audio processor. Some industry professionals have commented that the lack of pre-emphasized audio gives the FM digital audio a sound that does not “sound like radio.” This is not necessarily a bad thing.

What the FM HD Radio listener will notice in the mobile environment is a lack of multipath distortion or “picket fencing,” though most listeners do not know this distortion is due to multipath fading. Most listeners will probably refer to it as that “hiss-hiss-hiss” or brief fading that occurs on FM. The FM digital signal is fairly immune to this phenomenon, and it makes for a much better listening experience.

Keep in mind that to keep the digital signal within the allotted channel space of an AM or FM station, the audio is transmitted in a data-reduced format. While not perfect, the HDC perceptual coding technology used in the iBiquity HD Radio system is optimized for low (under 100 kilobits per second or kbps) data rates. You may have heard that the HD Radio digital audio sounds like “computer audio.” This is not the case, and in general, the listener will have a good experience with HD Radio.

The audio in the analog channel and the primary HD Radio digital channel must be both time and level synchronized so that there is not an abrupt change for the listener when the radio blends between analog and digital audio modes. It should also be known that, due to the digital conversion processing time which includes error correction, it is necessary to delay the analog audio on the AM HD Radio system by approximately 8.4 seconds, while in the FM HD Radio system the delay is approximately 8.7 seconds. I will discuss ideas to handle this delay for a station, which monitors an off-air signal, in Chapter 2.

It should also be known that, because the HDC codec provides an audio stream in a data reduced format, audio processing methods on the digital channels must be different than techniques used on the analog channel. You will need to add an audio processor specifically for the digital channel. This will be covered in Chapter 4.

1.2  The HD Radio signal for AM radio

The HD Radio signal, as generated by the iBiquity exciter, and as transmitted, consists of 81 OFDM subcarriers utilizing one of two modulation schemes. Most of the subcarriers reside in digital sidebands starting at ±4905.5Hz from the channel center frequency, with the last subcarrier being located ±14716.6 Hz from the channel center frequency.

The first set of subcarriers is located within the audio passband of the analog signal and is transmitted in quadrature with the analog RF carrier using quadrature phase-shift keying (QPSK) modulation. QPSK works by phase modulating the digital subcarrier, similar to FM, but the subcarrier frequency does not change except its phase relationship. The data therefore is transmitted by changing the phase of the QPSK subcarrier.

FIGURE 1-1

The AM HD Radio waveform.

The primary digital subcarriers carry the “core” audio. They are located ±10356.1 to 14716.6 Hz from the center of the channel, and are transmitted at a level of –28 dB as referenced to the unmodulated analog RF carrier. The secondary digital subcarriers are located ±5087.2 to 9447.7 Hz from channel center, and are transmitted at a level –43 dB as referenced to the unmodulated analog carrier. The secondary carriers carry enhanced audio data and are responsible for the stereo information contained in the digital audio signal. The tertiary digital subcarriers are transmitted from ±363.4 to 4723.8 Hz and are in quadrature with the analog RF carrier. The tertiary digital subcarriers are presently transmitted at a level of –45 dB referenced to the unmodulated RF carrier.

Finally, there is a “reference” subcarrier transmitted at ±181.7 Hz at a level –26 dB as referenced to the analog RF carrier, and two IDS subcarriers, transmitted at ±4905.5 and 9629.4 Hz, respectively, at a level –37 dB as referenced to the analog RF carrier. The IDS carriers contain information regarding a particular station’s HD Radio operation which is used by the radio in the decoding process.

The AM HD Radio main channel has an audio bandwidth of 20 Hz to 15 kHz, a minimum stereo separation of 70 dB, and a dynamic range of 72 dB. Additionally, when all-digital AM HD Radio is someday authorized, two audio streams will actually be transmitted – “main” and “backup.” The AM backup channel has an audio bandwidth of 20 Hz to 10 kHz, a dynamic range of 60 dB, and is mono. Audio quality of the AM main channel can be described as “FM like.” Audio quality of the AM backup channel can be described as “AM mono.”

There are two modes of operation for the hybrid AM HD Radio signal. The first mode, called 5 kHz mode, limits the analog audio bandwidth to approximately 5 kHz. The second, called 8 kHz mode, limits the analog audio bandwidth to approximately 9 kHz. No, that is not a misprint – the “8 kHz” mode name was originally assigned by iBiquity when the bandwidth was 8 kHz, and when the bandwidth was widened to 9 kHz the name was not changed.

In the 5 kHz mode, the radio constantly monitors the digital sidebands on either side of the analog AM signal and can choose to decode either one or both sideband groups, whichever will produce the best recovered data. I have participated in testing with the signal of WOR Radio in New York City which has shown that the digital sidebands of a hybrid AM HD Radio signal can be affected differently and independently by various causes. I have seen the lower digital sideband of the WOR signal literally disappear in the null of the antenna system on the northern portion of Route 17 in New Jersey. Even with the loss of the lower sideband, the digital audio was still being recovered properly, as the radio was getting its data from the upper sideband.

In the 8-kHz mode, the radio must constantly receive both of the digital sideband groups at all times to recover the HD Radio signal. In testing with the WOR signal, the coverage of the HD Radio digital signal was impacted significantly by operation in the 8-kHz mode, particularly in the null on the northern portion of Route 17 in New Jersey. Obviously, operating in the 5 kHz mode will make the AM HD Radio digital signal more robust.

Your station can also transmit PAD along with your audio. Depending on the software package and digital audio delivery system your station is running, you can put song title and artist on listeners’ radios, put up messages based on time, and put up information associated with commercials. This will be a new experience for AM listeners and is something new for AM stations to bring to sponsors.

Additionally, the AM HD Radio system has “extra” data capacity that can be made available for sale, much like FM subcarriers. This data capacity can also be used for numerous purposes by the radio stations. AM HD Radio brings many things to the table that stations can use to their advantage.

1.2.1  FCC approval, interference, and nighttime operation

The iBiquity HD Radio IBOC system was granted an “interim authorization” by the FCC on October 10, 2002. As part of that authorization, AM radio stations were allowed to operate using the hybrid IBOC mode of operation from local sunrise or 6 AM, whichever is earlier, until local sunset or 6 PM, whichever is later.

FIGURE 1-2

The AM HD Radio signal shown with the present-day NRSC mask limits.

Looking at the table in the previous section on HD Radio subcarrier placement in the transmitted spectrum, it is obvious that the energy of the HD Radio subcarriers is transmitted not only within the ±10 kHz used by the station’s analog signal, but on the frequencies which may be occupied by first and second adjacent channel signals, as well. The energy of the primary and secondary HD Radio subcarriers which exists in these adjacent channel regions is well below the specification of the NRSC (and FCC) mask, as the system was designed to operate within the mask so that it would be compliant with FCC rules.

Due to its spectral occupancy, the AM hybrid HD Radio signal may cause unintended interference to first and second adjacent neighbors of a station transmitting an HD Radio signal. iBiquity has performed a great deal of analysis of AM radio stations throughout the United States, both analyzing the station’s coverage as specified through antenna proofs of performance on file with the FCC, and through field strength measurements on both HD Radio and non-HD Radio stations.¹ This information has showed that only a handful of stations would likely experience interference within their normal nighttime interference-free coverage contours.

FIGURE 1-3

The NRSC-5-A mask. Note that the limits are much more stringent than the mask for analog AM specified in NRSC-2-A.

This is particularly important information, especially where it concerns nighttime operation of AM HD Radio. As the atmosphere cools starting at sunset, the ionosphere moves closer to the Earth’s surface. When this happens, AM radio signals are reflected off of the ionosphere and can travel great distances, much farther than they do as a result of normal, “ground-wave” coverage. The resultant “skywave” signals have the potential to cause interference to co-channel and adjacent channel stations, hundreds, and in some cases, thousands of miles away. Consequently, one aspect of the iBiquity nighttime testing was designed to determine the effect of skywave propagation on co-channel and adjacent channel signals.

I participated in hybrid AM IBOC nighttime testing utilizing signals from stations WOR and WLW. WOR is on 710 kHz, located in New York City, with the transmitter located in Rutherford, New Jersey. WLW is located in Cincinnati, Ohio, with the transmitter located in Mason, Ohio, and operates on 700 kHz. Both stations operate with 50,000 watts. These tests were conducted over three nights in December 2002.

Both WOR and WLW were operating with a special version of the iBiquity operating software created specifically for this test. This software was synchronized at both stations by iBiquity engineers, and would turn on the HD Radio subcarriers at WOR, and simultaneously turn them off at WLW. Several minutes later, the HD Radio subcarriers at WOR would turn off, and the HD Radio subcarriers at WLW would turn on.

The iBiquity test van used in the New York tests contained a spectrum analyzer, an iBiquity HD Radio test receiver, and a computer running specialized software that would record spectrum analyzer measurements, audio, and data on the HD Radio signal recovered from the test radio. The van also included a myriad of analog radios, including a home tuner, a General Electric SuperRadio III, and a few others whose names I forget. All were typical radios that could easily be found and available for use by consumers. The only modifications to these radios were to make the audio output signal directly available for recording purposes (if these radios did not already have that facility).

A listening location was chosen in Pennsylvania, in a parking lot alongside Interstate 78. The WOR signal, as measured with a calibrated Potomac Instruments FIM-41 field intensity meter was just about 0.5 millivolt per meter. The consumer radios were set up outside the vehicle, each connected to the recording computer.

When the WLW HD Radio subcarriers were on, a very slight hiss was heard under the WOR programming. When the WOR HD Radio subcarriers were on, WLW could still be heard, though there was a hiss under their signal. It should be noted that these evenings were very active skywave nights, and the level of noise under the WLW signal with the WOR HD Radio subcarriers off was the same as when they were on.

Another listening location was chosen in Allentown, Pennsylvania, again in a parking lot alongside Interstate 78. This was skywave territory, and we were outside the WOR 0.5 millivolt per meter contour. Again, the radios were placed outside the vehicle and connected to the recording computer.

Under skywave conditions, if the WOR HD Radio digital subcarriers were on and the WLW signal strength was equal to or exceeded the WOR signal strength, WLW was very easy to listen to. If the WOR signal strength increased beyond the level of the WLW signal, WLW would be unlistenable. It worked the same way for the WOR signal when the WLW HD Radio digital subcarriers were on. It must be noted, however, that when either station was severely interfered with by the other HD Radio digital subcarriers, if those subcarriers were turned off, the level of noise on the band was just as bad. My analysis of the situation was that there was very little interference being caused by the HD Radio digital subcarriers, and that the majority of the interference experienced during these tests was caused simply by skywave conditions.

Based on tests such as the aforementioned, the FCC approved use of HD Radio on AM stations during nighttime hours in their Second Report and Order issued March 22, 2007. This authorization took effect on September 14, 2007.

1.3  The HD Radio signal for FM radio

The FM hybrid HD Radio signal operating in the primary mode of MP1 consists of 190 OFDM subcarriers starting at ±129,361 Hz from the channel center frequency and ending at ±198,302 Hz from the channel center frequency. These OFDM subcarriers are the primary set of subcarriers for the FM hybrid HD Radio signal, and operate at a data rate of 98.4 kbps. The level of these subcarriers, compared to the unmodulated FM carrier, is –45.8 dBc. None of these subcarriers are contained within the analog audio passband as is the case with the AM HD Radio signal. It should be noted that these levels will allow the FM HD Radio waveform to fit within the FCC mask for FM stations.

FIGURE 1-4

The FM HD Radio waveform.

It is important that the FM HD Radio subcarrier levels be compared to the unmodulated FM carrier for purposes of establishing compliance with the mask. Don’t forget that when modulated, the FM carrier level will be constantly changing and in many cases the carrier will disappear, as the total energy of the sidebands cannot exceed total carrier power in FM. This is as opposed to AM modulation, where the sideband energy is added to the carrier and the (unmodulated) carrier level is constant.

An FM station has several choices available on how to utilize the available 98.4 kbps data capacity. One is to use the entire 98.4 kbps for the main channel audio. Because the FM HD Radio system has more bandwidth available and can operate at a faster data rate than the AM HD Radio system, it is possible to subdivide the available data capacity and add additional audio channels, called multicast channels. Contrary to what you may have heard, the digital coverage of a multicast channel will equal that of the main HD Radio digital audio channel – they are contained within the same data stream that is transmitted and do not operate at different power levels. However, the multicast channels will not blend to analog when the main channel audio does, but instead will mute. The station also has an available data capacity of 860 bits per second which can be utilized for many different purposes. Incorporation of PAD information and station information is standard as with the AM HD Radio system.

FIGURE 1-5

The FM Hybrid HD Radio waveform with the FCC mask.

A station could, for example, run the main digital audio channel at 64 kbps and add a second digital audio channel, which could be another program format, and run it at 32 kbps. At this data rate, the second digital channel would be stereo and have pretty much the same frequency response of the main digital audio channel, though the data reduction would be greater with a greater possibility of hearing data reduction artifacts.

The station could also subdivide the available 98.4 kbps into three different stereo audio channels to provide, say, two program formats, with the third digital channel being used for a radio reading service. There are many choices to make, which will be discussed in Chapters 4 and 6. It should be noted, however, that there is a point where reducing the data rate will affect the audio characteristics of the main digital audio channel. The FCC, in their Second Report and Order issued March 22, 2007, has stated that the audio quality of the main digital audio channel must be equivalent to the audio quality of the analog audio channel. This will vary by station, for example, a talk radio station will be able to achieve the required audio quality with a lower bit rate than will a classical music station.

If a station presently has 67 or 92 kHz analog subcarriers in operation, they are compatible with this mode of FM HD Radio operation, with one exception. 92 kHz analog subcarriers are not compatible with the extended hybrid mode of FM HD Radio operation (this is discussed more below).

The main HD Radio digital audio channel has a frequency response of 20 Hz to 20 kHz, with a minimum stereo separation of 70 dB, and a dynamic range of 96 dB. It can be described as “CD like.” In the all-digital FM HD Radio configuration (not presently authorized by the FCC), there is a digital backup channel transmitted. It has a frequency response of 20 Hz to 15 kHz, a dynamic range of 65 dB, and is mono. For purposes of blending, this backup digital channel takes the place of the analog, that is, in the all-digital mode the main channel digital audio blends to the backup digital audio channel as impairments are encountered or signal strength at the receiver is diminished.

Since there is more room available in the FM channel, additional OFDM subcarriers, called Extended subcarriers, can be added to the FM HD Radio signal, and the mode then becomes the Extended Hybrid mode of operation. The extended subcarriers can be added in groups, called partitions. There are four available partitions that can be added in the extended hybrid mode. The maximum extra data capacity added by operating with extended partitions is 49.6 kbps.

A station operating with all four extended partitions will find them starting at ±101,744 Hz from the channel center frequency, and ending at ±128,997 Hz, at an operating level of –45.8 dB below the unmodulated analog carrier. Operating with extended partitions will adversely affect an analog subcarrier operating at 92 kHz. An analog subcarrier at 67 kHz and an RDS subcarrier at 57 kHz may notice a slight increase in signal-to-noise level, but otherwise should operate without problems.

An FM station operating in extended hybrid mode will have the ability to utilize the data capacity created by adding the extended partitions in a myriad of ways. It can be sold and utilized for data transmission by an outside company, providing an additional revenue stream for the station much in the same way analog subcarriers do. This capacity can potentially be utilized to provide album cover art to listeners’ radios. There are uses for this data-casting ability that have not yet been dreamed up, but stations should know that it is available for their use. It should be noted that the extra data capacity gained from using extended partitions can also be used for the audio payload.

The actual transmitted power level of the FM HD Radio subcarriers, taken as a group, is 1 percent of the transmitted analog power. There has been much discussion regarding the possible need to raise this level. This is beyond the scope of this book, as testing is being done, but many hurdles would need to be overcome, including the re-configuration of existing HD Radio facilities to accommodate the additional digital output power.

There has also been much discussion regarding the possibility of interference being caused to first adjacent stations. As an example, a report was issued in 2003 regarding FM HD Radio that listed KWAV in Monterey, California, as a potential interferer to first adjacent stations on either side of their frequency of 96.9 MHz. This prediction was because KWAV is a “grandfathered superpower” Class-B station, operating at an effective radiated power (ERP) of 18 kW and a height above average terrain (HAAT) of 747 meters. Normal ERP for a Class-B station at this height would be 1.3 kW. The predicted 60 dBu contour is 49.46 miles from the transmitter.

The FCC, over the years, had allowed at least one first adjacent station to overlap contours with KWAV. It was predicted that this other station would lose a good portion of its coverage area if KWAV put on an HD Radio signal. KWAV did put on an HD Radio signal, and there have been no reports of interference to this first adjacent station.

HD Radio will be changing the landscape on the AM and FM bands for broadcasters and listeners alike. Now that we know what the HD Radio signal is, let’s make it work, shall we?

¹ See, for example, “AM Nighttime Compatibility Study Report,” iBiquity Digital Corporation, May 23, 2003.