BACKGROUND OF THE INVENTION This invention relates to FM stereophonic broad- casting systems and, more particularly, an improved FM stereophonic broadcasting system which increases the broad¬ cast coverage area over that of current biphonic service yet is compatible with exisiting monophonic and biphonic re¬ ceivers.

The potential of FM sound broadcasting has long been recognized, and because of its relative immunity to electromagnetic interference and ^• its ability to provide full audio bandwidth with low noise, was also selected as the transmission method for television sound. Although FM radio was hardly a universal success in the commercial sense when stereophonic broadcasts were first authorized in 1961, it was not long before the attraction of two-channel high- fidelity sound elevated FM to the status it enjoys today. However, although FM-stereo adds a new acoustical dimension to radio reception, it does so only at the expense of serious degradation of another high fidelity parameter, namely, the signal-to-noise ratio.

The noise penalty in stereophonic broadcasting is well known; less obvious, however, is the restrictive in¬ fluence this phenomenon has on station coverage, which, for equivalent signal-to-noise ratio, typically may be only one fourth or one fifth the area of simple monophonic broad¬ casts. Several factors contribute to the higher noise levels and coverage losses resulting from multi-channel sound transmissions. When a broadcast station converts to biphonic service, monophonic coverage is reduced because

signal power must be divided among the various components of the more complex baseband signal. (The term "biphonic" will be used hereinafter in order to clearly differentiate two- channel broadcasting from other forms of stereophony such as triphonic and quadraphonic broadcasting.) The biphonic signal-to-noise ratio is lower than monophonic signal-to- ήoise ratio because of the wide band width of the composite signal having the familiar equation f(t) = M + p sin ω/2 t + S sin ωt Eq.(l) where M is the monophonic sum signal, p is the pilot, and S is the stereophonic difference signal. With a baseband spectrum extending to 53 kHz for biphonic transmissions-, the noise levei is particularly high because of the rising spectral characteristic due to frequency modulation. As shown in FIG. 1, the so-called "triangular" noise spectrum increases 6dB per octave with increasing frequency of the composite signal. Audio de-emphasis counteracts this some¬ what as shown in FIG. 1, but the noise problem is still severe. After demodulation, the noise components of the difference channel subcarrier are added, statistically in¬ dependent, to the noise already present in the monophonic -signal during audio dematrixing.

Any precise computation of the theoretical loss of signal-to-noise ratiomust take into account factors such as the effect of de-emphasis, the format of the audio test signal (which is assumed for the computation) , and inter¬ leaving. Interleaving is the interesting phenomenon where¬ by with certain audio signals the peak amplitude of the sum of the main channel signal and the sub-channel signal may be less than the sum of the peak amplitudes of these channels,

thus permitting the interleaved signals to be raised to full modulation, with a resultant improvement in- the signal-to- noise ratio. These factors have been studied by a number of researchers, and a calculation of the signal-to-noise deg- radation in biphonic broadcasting was published by N. Parker and D. W. Ruby in a 1962 paper entitled "Some notes on the calculation of the S/N Ratio For a FM System employing a double sideband AM multiplexer signal", IEEE Trans. Broad¬ cast Television Receivers (International Convention Issue) , vol. BTR-8, pp. 42-46, April 1962. The authors assumed the transmission of the peak monophonic power available, i.e., ho modulation of the subcarrier (L - R = 0) ; while their report of 23dB degradation has received widespread accept¬ ance, the figure is not entirely representative of typical programming. More recently, under EIA auspices, the subject has been studied in greater detail by the National Quadra¬ phonic Radio Committee (NQRC) and in its final report to the Federal Communications Commission, vol. II, chapter 1, sec¬ tion 1.4, November 1975, .by J. Gibson, et al, entitled, "Signal And Noise Levels In Proposed Multiplexed Systems For FM Broadcasting Of Quadraphonic Sound", reaffirmed the 23dB penalty for a monophonic test signal, but also, by virtue of using a wide variety of audio test signals, demonstrated that a penalty of 26dB is more representative of stereo- phonic programming with wide audio separation. For mono¬ phonic receivers, the NQRC data predict noise degradation of IdB to 7dB, depending 'On the particular type of test signal used.

Such losses of signal-to-noise ratio also cause a reduction in the effective area of coverage of a broadcast

_QMPI_

4 •

station; this effect for a representative set of trans¬ mission and reception conditions, based on NQRC data pub¬ lished January 15, 1976 as a Supplement to th ce afore- mentioned report of the NQRC to the FCC entitled, "Illus- trations to Relations Between Sigπal-To-Noise Ratio and Range in Existing FMBroadcast Services And Proposed Systems For FM Broadcasting of Quadraphonic Sound", is illustrated in FIG. 2. As a basis for this illustration, the NQRC used the FCC FM Engineering charts for the estimated field strength exceeded at 50% of the potential receiver locations for at least 50% of the time with a dipole receiving antenna height of 30 feet. The transmitter height was assumed to be 1,000 feet with a 10 kilowatt effective radiated power at 98 MHz, and the receiver was assumed to have a lOdB noise figure. For reception at a signal-to-noise ratio of 50dB, the limit of station coverage would extend to a radius of 128 miles when monophonic transmission only is employed. How¬ ever, with biphonic transmission, two-channel reception extends only to a 60-mile radius, and monophonic reception is reduced to 100 miles. Although in reality station service areas are often limited by co-channel and adjacent-channel interference rather than by noise, FIG. 2 represents a useful comparison of the theoretical limits.

A potential solution of the noise penalty problem is the use of companding systems which achieve noise reduction by compressing the dynamic range of an audio program before transmission and expanding it to its original dynamic range at the receiver, the effect of which is illustrated in FIG.3. The "original program" signal, with a wide dynamic range and a low noise level is represented at

the left of the figure, and in the center the program is shown compressed to approximately one-half its original dynamic range for transmission purposes. During trans¬ mission, additional noise is introduced at a level below that of the compressed program, but at a level which would have intruded on the program had it not been compressed. Finally, the "expanded program" is shown at the right re¬ constituted to its original dynamic range and with the transmission noise simultaneously reduced to an unobtrusive level. Companding systems exhibiting these properties have achieved success in various audio applications, including tape and disc recording. For its potential application to broadcasting, tests were conducted in Sweden in the early 1960's utilizing a companding system in the S channels of FM- AM and FM-FM transmission systems; favorable results were reported for the FM-FM transmissions, although the system was never fully implemented. The results are briefly described in Vol. X, Report 300-3, "Stereophonic Broad¬ casting," of the Xlllth Plenary Assembly, International Radio Consultative Committee (CCIR) , Geneva, 1974. Sig¬ nificant improvements in companding systems have been achieved during the last 20 years, and there is now renewed interest in the application of companding in broadcast systems as exemplified by the current examination by the Multichannel Sound Committee of the Broadcast Television Systems Committee of the potential application of com¬ panders to the S channel for television audio.

Given the recent advances in the art of audio companding, it is also appropriate to again examine its potential application to FM radio broadcasting. Currently,

some broadcasters utilize Dolby-type encoding to provide modest noise reduction in receivers equipped with appro¬ priate expanders and, relatively acceptable playback with simple receivers not having expanding capability. However, the requirement that compatibility with simple receivers must be maintained inhibits the potential for truly sig¬ nificant noise reduction in the other (expander-equipped) receivers.

SUMMARY OF THE INVENTION It is a primary object of the present invention to provide an improved biphonic FM radio broadcasting system that is fully compatible with existing receivers and which achieves better noise reduction than in prior systems.

A related object of the invention is to provide an improved FM radio broadcasting system that will extend the area of coverage of an FM broadcasting station.

In accordance with the present invention, a pair of stereophonically related audio frequency waves char¬ acterized as L (left) and R (right) are matrixed to obtain two signals exhibiting the matrix equations: (1) M =' L + R, and (2) S - L - R. The difference signal, S, is applied simultaneously to a modulator where it is used to amplitude- modulate a first sub-carrier signal and to a compressor which compresses its dynamic range to produce a compressed signal S'. The compressed signal S ¹ is used to amplitude- modulate a second sub-carrier signal of the same frequency but in quadrature phase relationship with the first. Sup- pressed-carrier, double-sideband modulation of each sub- carrier is employed, with the frequency of the subcarrier signals being sufficiently high as to assure a frequency gap below the lower sidebands of the modulated subcarrier sig¬ nals and the M signal. A conventional low-level phase reference pilot signal, lying within the aforementioned frequency gap, is employed for receiver detection purposes. The aforementioned M-signal, the two double-sideband sup- pressed-carrier signals, and the pilot signal are frequency modulated onto a high frequency carrier for transmission purposes.

NVIPO

The composite, frequency modulated carrier signal is transmitted to one or more remote receivers, which may be of the conventional monophonic or two-channel stereophonic type, or preferably a receiver constructed in accordance with the invention so as to include an expander for expanding the compressed S signal. The composition of the transmitted signal is such that a conventional monophonic receiver decodes only the sum signal M, and a two-channel receiver reproduces the transmitted M signal in both loudspeakers for monophonic reception, and the traditional stereophonic sig¬ nals for the biphonic mode. A new class of receiver, for providing substantially noise-free biphonic reception', re¬ produces the transmitted sum signal M and the expanded stereophonic difference signal.

O PI

BRIEF DESCRIPTION OF THE DRAWINGS The invention will be more fully appreciated from the following detailed description when considered in con¬ nection with the accompanying drawings in which: FIG. 1, to which reference has already been made, is a plot of the FM noise spectrum with de-emphasis;

FIG. 2, to which previous reference has been made, is a plot showing the reception limits for a 50 dB signal- to-noise ratio; FIG. 3, previously referred to, is a chart which illustrates the principle of signal companding;

FIG. 4 is a frequency diagram of the composite baseband signal developed in accordance with the principles of the present invention; FIG. 5 is a simplified block diagram of a trans¬ mitting terminal for generating the composite signal of FIG. 4;

FIG. 6 is a simplified block diagram of a re¬ ceiving terminal constructed in accordance with the in- vention;

FIG. 7 is a plot showing reception limits for a 50 dB signal-to-noise ratio in a FM radio broadcasting system according to the present invention; and

FIG. 8 is a plot showing signal-to-noise ratios at key contours.

DETAILED DESCRIPTION As background for the present invention, it will be useful to briefly review the basic principles of the existing two-channel stereo system approved by the FCC. In the current FM radio system, the stereophonically related signals that are added together, constitute a "monophonic sum channel" which consists of a (L + R) signal of 50 to 15,000 Hz, where L and R represent the left and right independent audio signals, each of which may also include a 0.7C "center" component. It is this combined signal that is reproduced by a standard monaural FM receiver, hence the descriptive term "monophonic channel" and the use herein of the letter M to identify this channel. To this is added a double-sideband suppressed 38 kHz sub-carrier signal S sin ωt, where S - (L - R) , along with a 19 kHz pilot signal. The composite modulation signal can be written as: e _m = M + p sin ω/2 t + S sin ωt Eq. (2) where ω = 2ττf _s and f _s = 38 kHz, and p is the amplitude of the 19 kHz pilot. Looking at the baseband spectrum, one finds a monophonic channel M from about 50 Hz to 15 kHz, a 19 kHz pilot, and a stereophonic channel S sinωt signal from 23 to 53 kHz. If SCA (Subsidiary Communication Authorization) is also being transmitted, there would be an SCA frequency modulated subcarrier band in the region above 53 kHz and below 75 kHz. In accordance with the present invention, a compressed S channel (hereinafter designated S ¹ ) is trans¬ mitted in addition tβ the just-described signal. The additional S' channel preferably is transmitted in quadra¬ ture with the conventional stereophonic subcarrier to pro- duce the composite baseband signal illustrated'in FIG.4 and

-£URE OMPI

represented by the following equation: e _m = M + p sin ω/2 t + S sinωt + S'cos ωt Eq.(3) where p represents the amplitude of the pilo<t signal and ω= 2ττ«38 kHz. The quadrature subcarrier requires no additional spectrum space, and, as will be shown herein¬ after, imposes only a negligible penalty in modulation potential.

The transmitter for generating this composite signal is illustrated in the block diagram of FIG. 5. For purposes of simplicity, some of the more conventional transmitter circuits (e.g., pre-emphasis networks, carrier frequency source, and carrier frequency modulator) have not been shown ^' in detail and will be mentioned only briefly, where necessary, in the following description. The two audio frequency signals L and R, derived from separate sources (not shown), are applied by pre-emphasis networks (not shown) to the inputs of a conventional matrix network 10 consisting, for example, of a network of summing am¬ plifiers arranged to produce at the output of the matrix the two audio signals M = (L + R) and S = (L - R) . The monophonic signal, M, is applied as one input to an adder 12, and the stereophonic difference signal (L - R) is applied to the input of a modulator 14, the output of which is applied to adder 12, and also to the input of a compressor 16 of a suitable noise-reducing companding system. An eminently suitable compressor is described in commonly assigned U.S. Patent No. 4,376,916 en-titled "Signal Compression and Ex¬ pansion System". Briefly, the compressor (and expander) described therein is of the type which includes a variable gain device for controlling the gain of the input signal and

• a circuit for generating a control signal for the variable gain device including a rectifier for producing respon- sively to the input signal a DC signal which substantially follows dynamic variations of the input signal. The control 5 signal generating circuit includes a network of at least three signal paths having differing time constants, the network having a common input terminal for receiving the rectified signal and a common output terminal connected to the variable gain stage for coupling the control signal

10 thereto. All except one of the signal paths each includes means for causing the respective path to be conductive only in response to ^' the rectified signal exhibiting a change in amplitude sufficient to exceed a predetermined voltage, and the said one signal path conducts in response to any differ-

15 ential in voltage between the common input and output terminals.

The compressed stereophonic difference signal from compressor 16, designated S', is applied to the input of a second modulator 18, the output of which is also

20 delivered to adder 12 where it is linearly combined ^' with the monophonic signal M and the signal from modulator 14.

The subcarrier and pilot signals are derived from a carrier generator 20 which is designed to provide an output sine wave signal having a frequency of 38 kHz and includes

It ^{t J} any one of the known arrangements for providing a 90° phase displacement between the subcarrier output signals applied to the respective modulators 14 and 18, as indicated in FIG. 5. The modulators 14 and 18 comprise suppressed- carrier amplitude modulators of known construction which serve to amplitude-modulate the two subcarriers with re-

- tJRE4 O PI

spective audio frequency signals so as to produce the two double-sideband, suppressed-carrier, amplitude-modulated subcarrier signals S sin ωt and S' cosωt. These twς> signals are then combined in adder 12 with the monophonic signal M and a sine wave pilot signal having a frequency of 19 kHz derived from carrier generator 20. The composite signal produced at the output of adder 12, set forth in Equation (3) above, is then applied to the FM exciter of the transmitter (not shown) and frequency modulated onto a high frequency carrier for transmission purposes.

A receiver according to the invention is shown in the block diagram of FIG. 6 and, again, for purposes of simplicity some of the more conventional FM receiver cir¬ cuits (e.g., RF and IF stages, discriminator, and de- emphasis networks) have not been shown and will be only briefly mentioned as necessary. In addition to having the capability of achieving maximum noise reduction, in the manner to be described, the receiver is fully compatible with conventional monophonic and two-channel (biphonic) stereophonic broadcasts. A received FM signal is amplified in the RF and IF stages (not shown) of a receiver/de¬ multiplexer 24, and demodulated in any of the known FM detection circuits (not shown) and demultiplexed to derive the audio signals contained in the received FM signal, namely, the signals M, S, and S'. When a monaural broadcast is being received, the output of the demultiplexer comprises only the monaural signal M consisting of (L + R) . This signal is applied as a first input to an adder 26 and to the positive input of a subtractor 28, and since no signal is applied to the second input of adder 26 nor the negative

O PI

input of subtractor 28, the monophonic M signal (i.e. , (L + R)~) appears at the output of each of adder 26 and sub- tractor 28, which are applied to the left <and right loudspeakers, respectivel _ For a received conventional two-channel stereo signal, the M and S signals will be derived at the output of the demultiplexer 24; as before, the M signal is applied to one input of each of adder 26 and subtractor 28, and the S signal, when a switch 32 is in the dotted line position, is subtracted from the M signal in subtractor 28. As a result, the output of adder 26 is the signal 2L, the amplitude of which is then reduced by one-half to obtain a signal L for application to the left loudspeaker (not shown) . Similarly, subtraction of the difference signal (L - R) from the mono- phonic signal yields a signal 2R which, again, is reduced in amplitude by one-half for reproduction by the right loud¬ speaker (not shown) of the two-channel system. All of the above is typical of the mode of operation of a conventional two-channel FM receiver. To achieve improved signal-to-noise ratio and extend the area of coverage Of the system, the compressed difference signal S ¹ is applied to the input of an expander 30 which exhibits complementarity with the compressor 16 in the transmitter for expanding the compressed signal to its original dynamic range, and this expanded signal, when switch 32 is in the position shown, is applied to the second inputs of adder 26 and.-subtractor 28. As in the case of the S signal, adder 26 and subtractor 28 produce the signals 2L and 2R at their respective outputs and these are then reduced in amplitude for reproduction by the left and right loud-

-15 -

speakers, respectively. Preferably the switch 32 is auto¬ matically actuated from its dotted line position πhen there is no compressed difference signal at the output of de¬ multiplexer 24 to the solid line position when the received signal contains a compressed difference signal. Such auto¬ matic switching can be achieved, for example, by either of the known techniques of modulating the pilot tone, or adding a separate identification signal, when a compressed differ¬ ence signal is transmitted; a detector in the receiver responsive to the identification signal produces a signal for actuating the switch 32 from the dotted line position to the position shown in FIG. 6.

The penalty in dB signal-to-noise provided by the addition of the companded-biphonic (C-biphonic) channel is compared with the performance of conventional monophonic and biphonic transmissions in the following table.

The table, based on NQRC calculations, shows the predicted performance for various combinations of the three trans¬ mission and receiving modes. The comparison is made tiith two

-gTPEAl

O PI

different test signals: a (L + R) signal equivalent to that used by Parker and Ruby, and L (or R) only, representative of most of the NQRC calculations. For either modulating signal, compressed-biphonic reception can be as good as equivalent monophonic reception if an ideal companding system is employed, i.e., sufficient noise reduction is achieved in the S' channel to allow the noise of the M channel to predominate.

FIG. 7 shows a prediction of the reception range limits for a 50 dB signal-to-noise ratio with companded biphonic transmission. The NQRC method and the 50, 50 charts contained in Vol. Ill, Part 73.333 of the FCC Rules and Regulations were used in developing this plot. It is seen that compared with conventional biphonic transmissions, the improved companded system causes a relatively insignificant reduction of monophonic reception from a 100 mile radius to a 96 mile radius, and a similarly modest reduction of biphonic reception from 60 miles to 56 miles. However, the improved companded biphonic service would extend all the way to the monophonic contour at 96 miles, which represents approximately a three-fold increase in coverage area over the existing biphonic service.

An alternative way of displaying the signal-to- noise of various transmission schemes, also suggested by the NQRC, is illustrated in FIG. 8 wherein signal-to-noise ratios are shown at the so-called "urban" contour (E ₀ = lmv/M) and at the "rur-al" contour (E ₀ = 50μv/M) . For the conditions assumed (which are the same as for the previous figures) all systems will exhibit the same signal-to-noise performance at the urban contour. With such high field

strength, reception characteristics here will be dictated only by receiver performance; typically, a signal-to-noise ratio of 65 to 70 dB may be realized. At the rural contour (70 mile radius), conventional biphonic receivers will exhibit a 43 dB signal-to-noise ratio, whereas companded biphonic receivers will achieve 62 dB.

It will be apparent from the foregoing that the improved companded service for FM radio broadcasting offers the potential for providing service nearly equivalent to that provided by monophonic receivers. Compatibility with existing receivers can be maintained by encoding the com¬ pressed signal in a new channel which is in quadrature with the conventional S channel. The necessary transmission equipment is relatively easy and inexpensive to implement, and technology for decoding the S ¹ channel has previously been developed for other applications such as AM stereo and FM quadraphonic reception.