News Column

Researchers Submit Patent Application, "Prediction-Based Fm Stereo Radio Noise Reduction", for Approval

September 11, 2014



By a News Reporter-Staff News Editor at Computer Weekly News -- From Washington, D.C., VerticalNews journalists report that a patent application by the inventors Purnhagen, Heiko (Sundbyberg, SE); Sehlstrom, Leif (Jarfalla, SE); Engdegard, Jonas (Stockholm, SE), filed on October 1, 2012, was made available online on August 28, 2014.

The patent's assignee is Dolby International Ab.

News editors obtained the following quote from the background information supplied by the inventors: "In an analog FM (frequency modulation) stereo radio system, the left channel (L) and right channel ® of the audio signal are conveyed in a mid-side (M/S) representation, i.e. as mid channel (M) and side channel (S). The mid channel M corresponds to a sum signal of L and R, e.g. M=(L+R)/2, and the side channel S corresponds to a difference signal of L and R, e.g. S=(L-R)/2. For transmission, the side channel S is modulated onto a 38 kHz suppressed carrier and added to the baseband mid signal M to form a backwards-compatible stereo multiplex signal. This multiplex baseband signal is then used to modulate the HF (high frequency) carrier of the FM transmitter, typically operating in the range between 87.5 to 108 MHz.

"When reception quality decreases (i.e. the signal-to-noise ratio over the radio channel decreases), the S channel typically suffers more during transmission than the M channel. In many FM receiver implementations, the S channel is muted when the reception conditions gets too noisy. This means that the receiver falls back from stereo to mono in case of a poor HF radio signal (typically referred to as a mono dropout).

"Even in case the mid signal M is of acceptable quality, the side signal S may be noisy and thus can severely degrade the overall audio quality when being mixed in the left and right channels of the output signal (which are derived e.g. according to L=M+S and R=M-S). When a side signal S has only poor to intermediate quality, there are two options: either the receiver chooses accepting the noise associated with the side signal S and outputs a real stereo signal comprising a noisy left and right signal, or the receiver drops the side signal S and falls back to mono.

"Parametric Stereo (PS) coding is a technique from the field of very low bitrate audio coding. PS allows encoding a 2-channel stereo audio signal as a mono downmix signal in combination with additional PS side information, i.e. the PS parameters. The mono downmix signal is obtained as a combination of both channels of the stereo signal. The PS parameters enable the PS decoder to reconstruct a stereo signal from the mono downmix signal and the PS side information. Typically, the PS parameters are time- and frequency-variant, and the PS processing in the PS decoder is typically carried out in a hybrid filterbank domain incorporating a plurality of Quadrature Mirror Filter (QMF) banks.

"It has been proposed in WO2011/029570, PCT/EP2011/064077 and PCT/EP2011/064084 to use PS encoding of a received FM stereo signal in order to reduce the noise comprised within the received FM stereo signal. The general principle of the Parametric Stereo (PS) based FM stereo radio noise reduction technology is to use parametric stereo parameters derived from the received FM stereo signal, in order to reduce the noise comprised in the received left and right signals. The disclosure of the above mentioned patent documents is incorporated by reference."

As a supplement to the background information on this patent application, VerticalNews correspondents also obtained the inventors' summary information for this patent application: "In the present document, a method and system for FM stereo radio noise reduction using a prediction-based framework is described. This prediction-based framework is an alternative approach to the Parametric Stereo (PS) based framework indicated above. As will be described in the present document, the prediction-based framework provides lower computational complexity. Furthermore, it has been observed that at the same time a prediction-based FM stereo radio noise reduction scheme achieves an improved audio quality compared to a PS-based FM stereo radio noise reduction scheme.

"According to an aspect an apparatus or system configured to reduce noise of a received multi-channel FM radio signal is described. The multi-channel FM radio signal may be a two channel stereo signal. In particular, the received multi-channel FM radio signal may be representable as or presentable as or indicative of a mid signal and a side signal. Furthermore, the side signal may be indicative of a difference between a left signal and a right signal of a stereo signal.

"In an embodiment, the apparatus comprises a parameter determination unit configured to determine one or more parameters indicative of a correlation and/or decorrelation between the received mid signal and the received side signal. The one or more parameters may be a prediction parameter a used to determine a correlated component of a noise-reduced side signal from the received mid signal and/or a decorrelation parameter b used to determine a decorrelated component of the noise-reduced side signal from a decorrelated version of the mid signal. Furthermore, the apparatus comprises a noise reduction unit configured to generate the noise-reduced side signal from the received mid signal using the one or more parameters. For this purpose, the noise reduction unit does not take into account the received side signal, e.g. the samples of the received side signal. In other words, the received side signal is not in the signal path for the determination of the noise-reduced side signal. In particular, the noise reduction unit may be configured to determine the noise-reduced side signal only from the received mid signal (e.g. the samples of the received mid signal) and the one or more parameters.

"As indicated above, the parameter determination unit may be configured to determine a prediction parameter a. The prediction parameter a may be indicative of a cross-correlation between the received mid signal and the received side signal. In particular, the parameter determination unit may be configured to determine the prediction parameter a based on an expectation value of a product of corresponding samples of the received mid signal and the received side signal. Even more particularly, the parameter determination unit may be configured to determine the prediction parameter a using the formula a=E[S*M]/E[M*M], wherein E[.cndot.] denotes the expectation operator, S denotes the received side signal and M denotes the received mid signal.

"In the case that the parameter determination unit provides a prediction parameter a, the noise reduction unit may be configured to generate the noise-reduced side signal (or a correlated component of the noise-reduced side signal) from the received mid signal using the prediction parameter a. The correlated component of the noise-reduced side signal may be determined as the product of the prediction parameter a and the received mid signal, i.e. a*M. This means that the correlated component of the noise-reduced side signal may be a weighted version of the received mid signal. In view of the fact that the prediction parameter a may be time variant and/or frequency variant, the weighting factor to the received mid signal may be time variant and/or frequency variant.

"The parameter determination unit may be configured to determine a decorrelation parameter b indicative of a decorrelation between the received mid signal and the received side signal. In particular, the parameter determination unit may be configured to determine the decorrelation parameter b based on the energy of a difference signal of the received side signal and a signal determined from the mid signal using the prediction parameter a. Even more particularly, the parameter determination unit may be configured to determine the decorrelation parameter b using the formula b=sqrt(E[D*D]/E[M*M]) with D=S-a*M being the difference signal. The operator 'sqrt( )' indicates the square root operation.

"In this case, the noise reduction unit may be configured to generate the noise-reduced side signal (or a decorrelated component of the noise-reduced side signal) from a decorrelated version of the received mid signal using the decorrelation parameter b. In particular, the decorrelated component of the noise-reduced side signal may be determined as b*decorr(M), with decorr(M) being the decorrelated version of the received mid signal. The decorrelated version of the received mid signal may be determined by filtering the received mid signal using an all-pass filter.

"If the received side signal comprises a significant amount of noise, it may be beneficial to reduce the impact of the decorrelated component of the noise-reduced side signal on the noise-reduced side signal. For this purpose, the parameter determination unit may be configured to determine an impact factor characteristic of (or indicative of) the spectral flatness of the received side signal. A high spectral flatness typically indicates a high degree of noise comprised within the side signal. As such, the decorrelation parameter b may be dependent on the impact factor. In particular, the decorrelation parameter b may decrease as the impact factor indicates an increasing degree of spectral flatness of the received side signal. By way of example, the impact factor is the SMF_impact_factor described in the present document and a modified decorrelation parameter b_new is determined as b_new=(1-SMF_impact_factor)*b, thereby forcing the decorrelation component of the noise-reduced side signal (i.e. b_new*decorr(M)) to zero, if the SMF-impact_factor tends towards '1'.

"As indicated above, the parameter determination unit may be configured to determine the one or more parameters (e.g. the prediction parameter a and/or the decorrelation parameter b) in a time variant manner. As such, for each of the one or more parameters, a sequence of the respective parameter for a corresponding sequence of time intervals may be determined. By way of example, for a first parameter (e.g. the prediction parameter a or the decorrelation parameter b) a sequence of first parameters for the sequence of time intervals is determined. The sequence of time intervals may be a sequence of signal frames (comprising e.g. 2048 signal samples). Typically, a particular first parameter of the sequence of first parameters for a particular time interval of the sequence of time intervals is determined using samples of the received mid signal and/or the received side signal which lie within the particular time interval. In cases, where the one or more parameters are time variant, the noise reduction unit may be configured to generate the noise-reduced side signal using the one or more time variant parameters.

"In order to ensure continuity between adjacent time intervals and in order to avoid audible discontinuities at the borders of adjacent time intervals, it may be beneficial to determine a sequence of interpolated first parameters by interpolating adjacent first parameters from the sequence of first parameters.

"In case of highly deteriorated reception conditions, FM receivers may force received FM radio signals to mono, i.e. the FM receivers may suppress the received side signal. The apparatus may be configured to detect such mono dropout, i.e. the apparatus may be configured to detect that the received multi-channel FM radio signal is a forced mono signal. This may be achieved by detecting a fast transition of the received side signal from high energy to low energy. In particular, an energy of the received side signal within a first time interval of the sequence of time intervals may be determined, and it may be determined that this energy is above a high threshold. Furthermore, a transition period of a number of following successive time intervals during which the energy of the side signal drops from a value above the high threshold to a value below a low threshold may be determined. Based on this information, it may be determined that the received multi-channel FM radio signal following the first time interval is a forced mono signal if the number of successive time intervals of the transition period is below an interval threshold. This interval threshold may be 1, 2, 3 or 4 time intervals following the first time interval.

"If it is detected that the received multi-channel FM radio signal in the time interval (directly) following the first time interval is a forced mono signal, the parameter determination unit may be configured to determine the one or more parameters for the time interval (directly) following the first time interval from the one or more parameters for the first time interval. In other words, the parameter determination unit may be configured to conceal the lack of parameters during a mono dropout by using the one or more parameters determined prior to the mono dropout.

"As described above, the parameter determination unit may be configured to determine the one or more parameters (e.g. the prediction parameter a and/or the decorrelation parameter b) in a frequency variant manner. This means that different parameters are determined for different subbands of the received mid and/or side signal. For this purpose, the apparatus may comprise a mid transform unit configured to generate a plurality of mid subband signals covering a corresponding plurality of frequency ranges from the received mid signal. Furthermore, the apparatus may comprise a side transform unit configured to generate a plurality of side subband signals covering the corresponding plurality of frequency ranges from the received side signal. In such cases, the parameter determination unit may be configured to determine the one or more parameters for each of the plurality of frequency ranges. In particular, for a second of the one or more parameters (e.g. the prediction parameter a and/or the decorrelation parameter b), a plurality of second subband parameters may be determined from the corresponding plurality of mid subband signals and the corresponding plurality of side subband signals. This may be done by applying the above mentioned formulas for determining the one or more parameters (e.g. the prediction parameter a or the decorrelation parameter b) to each of the plurality of frequency ranges.

"The noise reduction unit may be configured to generate the noise-reduced side signal using the one or more frequency variant parameters. In particular, the noise reduction unit may be configured to generate a plurality of noise-reduced side subband signals (only) from the corresponding plurality of mid subband signals and the corresponding plurality of subband parameters. Using an inverse transform unit, the noise-reduced side signal may be generated from the plurality of noise-reduced side subband signals.

"The mid transform unit and/or the side transform unit may be QMF filter banks and the inverse transform unit may be an inverse QMF filter bank. In view of the fact that the received mid signal is in the signal path (and the received side signal is not in the signal path), the side transform unit may meet lower requirements than the mid transform unit with regards to at least one of: frequency selectivity; frequency resolution; time resolution; and numerical accuracy.

"The received FM radio signal may be dominated by a noisy received side signal which has a higher energy level than the received mid signal. Such situations may lead to perceptually annoying artifacts when generating the noise-reduced side signal from the received mid signal using the one or more parameters. In order to cope with such situations, the parameter determination unit may be configured to limit the one or more parameters by applying to the one or more parameters a limitation factor c. In particular, the one or more parameters may be divided by the limitation factor c. In an embodiment, for c>1, the limitation factor c is proportional to the sum of the one or more squared parameters. In another embodiment, for c>1, the limitation factor c is proportional to the square root of the sum of the one or more squared parameters. Typically, the limitation factor c is selected such that the application of the limitation factor c does not increase the one or more parameters.

"It should be noted that the apparatus may comprise a delay unit configured to delay (a sample of) the received mid signal by an amount of time corresponding to computation time required to generate (a corresponding sample of) the noise-reduced side signal.

"In good reception conditions when the received side signal comprises little to no noise, it may be beneficial to use the received side signal for generating a stereo signal. For this purpose, the apparatus may comprise a combining unit configured to determine a modified noise-reduced side signal from the noise-reduced stereo signal and the received side signal using a quality indicator indicative of the quality of the received multi-channel FM radio signal. Depending on the quality of the received side signal the modified noise-reduced side signal may be blended between (or selected from or interpolated between) the noise-reduced side signal and the received side signal. For this purpose the combining unit may comprise a noise-reduced gain unit configured to weight the noise-reduced side signal using a noise-reduced gain; a bypass gain unit configured to weight the received side signal using a bypass gain; and a merging unit configured to merge (e.g. add) the weighted noise-reduced side signal and the weighted received side signal; wherein the noise-reduced gain and the bypass gain are dependent on the quality indicator. It should be noted that the combining unit may be configured to determine the modified noise-reduced side signal in a frequency selective manner.

"The apparatus may comprise a quality determination unit configured to determine the quality indicator which indicates the quality of the received side signal. This may be done by determining a power of the received mid signal, referred to as mid power, and a power of the received side signal, referred to as side power. A ratio of the mid power and the side power, i.e. a mid-to-side ratio may be determined and the quality indicator of the received FM radio signal may be determined based on at least the mid-to-side ratio. The present document describes various embodiments for determining a quality indicator .alpha..sub.HQ which indicates the quality of the received side signal in a reliable manner.

"The apparatus may further comprise an MS-to-LR converter configured to determined a noise-reduced left signal and a noise-reduced right signal from the received mid signal and the noise-reduced side signal (or the modified noise-reduced side signal). In particular, the MS-to-LR converter may be configured to determine the noise-reduced left signal from the sum of the received mid signal and the (modified) noise-reduced side signal; and the noise-reduced right signal from the difference of the received mid signal and the (modified) noise-reduced side signal.

"According to another aspect, a method for reducing noise of a received multi-channel FM radio signal is described. The received multi-channel FM radio signal may be presentable as a received mid signal and a received side signal. The method may comprise determining one or more parameters indicative of a correlation and/or decorrelation between the received mid signal and the received side signal; and generating a noise-reduced side signal from the received mid signal and not from the received side signal and using the one or more parameters.

"According to a further aspect, a software program is described. The software program may be adapted for execution on a processor and for performing the method steps outlined in the present document when carried out on a computing device.

"According to another aspect, a storage medium is described. The storage medium may comprise a software program adapted for execution on a processor and for performing the method steps outlined in the present document when carried out on a computing device.

"According to a further aspect, a computer program product is described. The computer program may comprise executable instructions for performing the method steps outlined in the present document when executed on a computer.

"It should be noted that the methods and systems including their preferred embodiments as outlined in the present patent application may be used stand-alone or in combination with the other methods and systems disclosed in this document. Furthermore, all aspects of the methods and systems outlined in the present patent application may be arbitrarily combined. In particular, the features of the claims may be combined with one another in an arbitrary manner.

DESCRIPTION OF DRAWINGS

"The invention is explained below by way of illustrative examples with reference to the accompanying drawings, wherein

"FIG. 1 illustrates a schematic example for a system for improving the stereo output of an FM stereo radio receiver;

"FIG. 2 illustrates an example of an audio processing apparatus based on the concept of parametric stereo;

"FIG. 3 illustrates an example of an audio processing apparatus based on the concept of prediction;

"FIG. 4 shows example power spectra for the mid and side signal of a noisy FM radio speech signal;

"FIG. 5 illustrates an example flow chart of a method for the processing of received FM radio signals using a quality indicator of the received FM radio signals; and

"FIG. 6 shows an example state machine used for the concealment of the prediction and decorrelation parameters."

For additional information on this patent application, see: Purnhagen, Heiko; Sehlstrom, Leif; Engdegard, Jonas. Prediction-Based Fm Stereo Radio Noise Reduction. Filed October 1, 2012 and posted August 28, 2014. Patent URL: http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=%2Fnetahtml%2FPTO%2Fsearch-adv.html&r=2503&p=51&f=G&l=50&d=PG01&S1=20140821.PD.&OS=PD/20140821&RS=PD/20140821

Keywords for this news article include: Dolby International Ab, Software.

Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC


For more stories covering the world of technology, please see HispanicBusiness' Tech Channel



Source: Computer Weekly News


Story Tools






HispanicBusiness.com Facebook Linkedin Twitter RSS Feed Email Alerts & Newsletters