GB2640266A

GB2640266A - Content-adaptive bass enhancement

Info

Publication number: GB2640266A
Application number: GB2405046.0A
Authority: GB
Inventors: Ben-Asher Matan; Neoran Itai; Shaashua Meir
Original assignee: Waves Audio Ltd
Current assignee: Waves Audio Ltd
Priority date: 2024-04-09
Filing date: 2024-04-09
Publication date: 2025-10-15
Also published as: WO2025215631A1

Abstract

A system and method for computerised audio processing having an input 11 and output 27. An audio signal is received and frequency bands are designated through a splitter or filter 12 into low/bass 14 and high 13 frequencies, with the bass band including a fundamental frequency of the audio signal. Harmonics of the fundamental frequency are then generated 17 using integer multiples. The system/method will dynamically measure the input and adapt energy 19 of the generated harmonics 29 based on the input. The harmonics at the dynamically adapted energy are then combined with a raw or processed input signal 23 to produce a bass-enhanced output. The method may comprise dynamically computing audibility of distortion based on measured input and adapt energy of the harmonics accordingly. The method may perform comparison of energy in the bass frequency to energy in the higher frequency and decrease generated harmonic energy accordingly, or do the same using a determined threshold.

Description

CONTENT-ADAPTIVE BASS ENHANCEMENT

BACKGROUND

1. Technical Field

Aspects of the present invention relate to dynamic psychoacoustic bass enhancement of 5 an audio signal

2. Description of Related Art

Low-end loudspeakers may present a significant amount of distortion due to physical limitations and cost constraints. When audio is rendered in small speakers, distortion may often be heard as "buzzing" artefacts. A low-end loudspeaker tends to be overloaded with bass frequencies that cannot be readily reproduced without audible distortion. A well-known solution is based on the psychoacoustic phenomenon known as the "missing fundamental" in which the human auditory system may perceive the fundamental frequency of a complex signal of higher harmonics of the missing fundamental. A listener may perceive the fundamental frequencies even though they are not physically present in the generated sound. US patent publications 5,930,373 and 11,102,577 disclose stereophonic bass enhancement including generating harmonics for the bass frequency band of an audio signal, summing the harmonics to the original higher frequency band of the original audio signal, and filtering out the bass frequency band, to yield a bass-enhanced signal.

Auditory masking is a well-known phenomenon in which the perception of one sound is affected by the presence of another sound. Auditory masking may be used for example in audio compression techniques such as mp3 encoding. Auditory masking is described in Moore, Brian CJ. An introduction to the psychology of hearing. Brill, 2012, Chapter 3.

As disclosed in international application publication W02020/222242, masking may be used to control distortion for dynamic reducing loudspeaker-induced harmonic distortion in sound generated by a loudspeaker responsive to a first audio signal by: attenuating a first band in the first audio signal including a fundamental frequency. The attenuation dynamically adjusts attenuation gain in accordance with a ratio between: i) an energy of the first band in the first audio signal; and ii) an energy of a second band in the first audio signal, the second band including one or more frequencies that is an integer multiple of the fundamental frequency.

In the article cited below, Xie, R, Wu, L., Sang, YF. et al. Correlation-aided method for identification and gradation of periodicities in hydrologic time series. Geosci. Lett. 8, 14 (2021). https://doi.org/10.1186/s40562-021-0M 83-x a method is developed for identifying periodicities of a time series, "Moving Correlation Coefficient Analysis" (MCCA). In the MCCA method, the correlation between the original time series and the periodic fluctuation is used as a criterion, aiming to seek out a periodic fluctuation that best fit the original time series, and to evaluate its statistical significance. Other common methods for identifying periodicities of a time series are reviewed and presented including: harmonic analysis method (HAM), power spectrum method (PSM) and maximum entropy method (MEM)

BRIEF SUMMARY

Various audio processors and audio processing methods are described herein. The audio processors include an audio input and an audio output. The audio input is configured to receive an input audio signal. A filter is configured to separate multiple frequency bands in the input audio signal. The frequency bands include a bass frequency band which includes one or more fundamental frequencies of the input audio signal. A harmonics generator is configured to generate audio harmonics of frequency which are integer multiples of the fundamental frequency. A measurement module is configured to dynamically measure an audio feature of the input audio signal and output accordingly a measured audio feature. A gain adjustment module is configured to input the measured audio feature of the input audio signal and determine at least in part therefrom, an adapted energy of the audio harmonics generated by the harmonics generator for use in bass enhancement. A combiner is configured to combine the generated audio harmonics at the adapted energy with the input audio signal or a processed version of the input audio signal, to produce a bass-enhanced output audio signal at the audio output. A predicted audibility of distortion at the audio output, may be dynamically computed in accordance with the measured feature of the input audio signal. The energy of the generated audio harmonics may be adapted according to the predicted audibility of distortion. The frequency bands may include a higher frequency band of the input audio signal including a frequency higher than the fundamental frequency. A comparator may be configured to perform a comparison of energy in the bass frequency band to energy in the higher frequency band. The adapted energy of the generated audio harmonics for use in bass enhancement may be dynamically adapted at least partly according to the comparison. The gain adjustment module may be configured to dynamically decrease the adapted energy of generated audio harmonics when the energy at the fundamental frequency in the low-frequency band of the input audio signal is higher in comparison with the energy of the original high-frequency band. The gain adjustment module may be configured to dynamically decrease the adapted energy of the generated audio harmonics when the energy in the high frequency band is higher than a previously determined threshold to tend to avoid overloading of the audio output. The gain adjustment module may be configured to dynamically decrease the adapted energy of the generated audio harmonics when the generated audio harmonics interact with high frequencies present in the high frequency band due to a difference in at least one of frequency and phase within a previously determined threshold. A periodicity measurement module may be configured to compute a degree of periodicity within the bass frequency band. The adapted energy of the generated audio harmonics may be dynamically adapted at least partly according to the periodicity. The energy of the generated audio harmonics may be dynamically increased when the measure of periodicity is increased.

Various computer-readable media comprising instructions are disclosed herein which, when executed by a computer system, cause the computer system to execute the method steps as disclosed herein.

These, additional, and/or other aspects and/or advantages of the present invention are set 10 forth in the detailed description which follows; possibly inferable from the detailed description; and/or learnable by practice of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is herein described, by way of example only, with reference to the accompanying drawings, wherein: Figure 1 illustrates schematically an audio processor, according to features of the present invention; Figure 2 illustrates schematically an audio processor, according to features of the present invention; Figure 3 illustrates schematically an audio processor, according to features of the present invention; Figure 4 illustrates schematically a prior art harmonic generator which may be used 5 according to features of the present invention; and Figure 5 is a flow diagram of a method of audio processing, according to features of the present invention.

The foregoing and/or other aspects will become apparent from the following detailed description when considered in conjunction with the accompanying drawing figures.

DETAILED DESCRIPTION

Reference will now be made in detail to features of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. The features are described below to explain the 5 present invention by referring to the figures.

By way of introduction, various embodiments of the present invention are directed to a method of enhancing bass (low) frequencies of an audio signal intended to be played over a loudspeaker, by way of example. The loudspeaker may be part of or connected to a computer system and not of highest quality and not capable of reproducing high quality bass frequencies without perceivable distortion. The bass-enhancement effect may include filtering out the bass frequency band, applying a harmonics generator to the bass frequency band and summing the generated harmonics to the original high frequencies of the original input audio signal (possibly with a gain or further post-processing), to yield a bass-enhanced signal.

Features of the present invention may be thus directed to evoke a psychoacoustic effect sometimes known as "missing-fundamental" in which harmonics of a fundamental bass frequency are present but the fundamental frequency from which the harmonics are generated has been filtered out and not input to the loudspeaker.

However, the high frequency harmonics being added to the input audio includes a non-linear transformation of the input audio signal and therefore adds high frequency distortion to the audio signal. In some cases, depending on the audio content, the added high frequency distortion may be audible and perceived as unpleasant. The added high frequency distortion is thus traded off against the original bass frequency distortion which tends to be avoided by filtering out the original low frequencies. Features of the present invention are directed to manage this trade-off so that depending on the audio content, bass enhancement may be more optimally applied to the audio signal.

The energy output of the harmonic generator may be controlled: set manually, use automatic gain (AGC), based on input level and/or may use dynamic range compression (US patent publication 5903373). According to features of the present invention, unpleasant distortion may be minimized by controlling the gain of generated harmonics dynamically, not only based on user control or on the input level, but also in adaptation to a measurable feature of the audio content. Features of the present invention are directed to lower the gain for bass-enhanced audio content expected to be perceived with more unpleasant high frequency harmonic distortion, and increase the gain for bass-enhanced audio content expected to be perceived with pleasant distortion.

Referring now to the drawings, reference is now made to Figure 1 which illustrates schematically a block diagram of an audio processor 10A according to features of the present invention. An audio input signal 11 may be input to a frequency splitting filter 12 configured to separate input audio signal 11 into at least a bass (low) frequency band 14 and a high frequency band 13. Low frequency band 14 may be used as an input 41 to a harmonic generator 17 configured to generate a series of harmonics, of integral order n=2....100, by way of example, for the fundamental frequencies that are found in low frequency band 14. Alternatively, an optional pre-processing module 18 may be used to pre-process low frequency audio signal 14 and output a pre-processed low frequency audio signal as input 41 to harmonic generator 17. In one embodiment, pre-processing module 18 may include a dynamic range compressor. Pre-processing module 18 may provide an optional input to periodicity measurement module 15.

Periodicity is a measure of how periodic is low frequency band 14 in the time domain. Periodicity may be determined using a normalised autocorrelation of low frequency band signal 14. In the case of discrete time windows i, autocorrelation may be expressed by the following formula, by way of example, equivalent to a correlation of signal x with a delayed copy of itself as a function of delay d.

I 0) x(i+d) )(1()2 The autocorrelation as shown above is a function of delay d and is normalized to a fraction between 0.0 and 1.0, where a value of 1.0 relates to a purely periodic input signal, and a value of 0.0 indicates a non-periodic or random signal. The delay d",", which maximises the autocorrelation is a measure of the period of the signal. The reciprocal of the period of the signal is a frequency which may represent a fundamental frequency I; -14"," in low frequency band signal 14. A highly periodic signal may be represented, in frequency domain as a sum of harmonics or integer multiples of a single fundamental frequency!. Hannonicity is a measure of the extent a signal may be represented by a sum of a series of harmonics of a single fundamental frequency,f/. Thus, periodicity and harmonicity are equivalent.

Referring back to Figure 1, periodicity or harmonicity measured in module 15 may be averaged over a time duration to prevent jitter over decisions and then used to control harmonic generator gain adjust 19, dynamically controlling the amount 29 of generated harmonics from harmonics generator 17 being used for bass enhancement. Gain adjust 19 may be performed directly to gain multiplier 21, or by controlling harmonic generator 17 over output 22, further details of which are presented in the description of Figure 4.

An optional high pass filter 25 may be used to filter out original low frequencies 14 and the resulting filtered output may be combined at combiner 23 with high frequency band 13 to produce audio output signal 27.

According to a feature of the present invention, the more low frequency band 14 is harmonic or periodic, bass-enhancement is likely perceived as pleasant and higher level energy 29 of the generated audio harmonics may be selected by gain adjust 19. On the other hand, non-harmonic or non-periodic inputs may result in unpleasant wide-band intermodulations when bass enhancement is applied and gain adjust 19 may dynamically reduce level energy 29 of the generated audio harmonics.

Reference is now made to Figure 2 which illustrates schematically a block diagram of an audio proccssor 10B according to features of the prcscnt invention. As in audio processor 10A (Figure 1), audio input signal 11 may be input to frequency splitting filter 12 configured to separate input audio signal 11 into at least bass (low) frequency band 14 and high frequency band 13. Low frequency band 14 may be used as an input to harmonic generator 17 configured to generate a series of harmonics, of integral order n-2....100, by way of example, for the fundamental frequencies that are found in low frequency band 14. Alternatively, an optional pre-processing module 18 may be used to pre-process low frequency audio signal 14 and output a pre-processed low frequency audio signal as input 41 to harmonic generator 17.

Low frequency band 14 may further be used as an input to an energy comparison module 16, which may compare energies respectively in high frequency band 13 and low frequency band 14 and provide a measure of auditory or frequency masking when harmonics are added to the audio. Pre-processing module 18 may provide an optional input to energy comparison module 16. Masking amount output from energy comparison 16 may be used to control harmonic generator gain dynamically controlling amount 29 of generated harmonics from harmonics generator 17 being used for bass enhancement. Gain adjust 19 may be performed directly to gain multiplier 21, and/or by controlling harmonic generator 17 over output 22, further details of which are presented in the description of Figure 4.

A number of auditory masking use cases may be considered: -When energy in high frequency band 13 is very low in comparison with energy in low frequency band 14, that is high frequency harmonics of fundamentals in low frequency band 14 are not already present in high frequency band 14, then bass-enhancement of low frequency band 14 may be audibly harsh with a large relative change of high frequency harmonics expected from the bass enhancement and gain output from gain adjust 19 may be correspondingly decreased to reduce harmonics output 29 from harmonics generator 17 being use for bass enhancement.

-When total energy in high frequency band 13 is already very high and adding further high frequencies output from harmonic generator 17 is not expected to effectively enhance bass response without tending to overload the output loudspeaker into its non-linear range, with too much high frequency energy, and causing loudspeaker distortion or rattle effects.

-Higher harmonics generated by harmonic generator 17 may interact with the original high frequencies present in the input signal high frequency band 13 due for example to small differences in frequency and/or phase and cause unpleasant beating. In this case, 20 gain output from gain adjust 19 may be correspondingly reduced.

Reference is now made to Figure 3, which illustrates an audio processor 10C according to features of the present invention. In audio processor IOC, features of audio processor 10A (Figure 1) and audio processor 10B (Figure 2) may be combined.

As in audio processor 10A (Figure 1), and audio processor 10B (Figure 2) audio input signal 11 may be input to frequency splitting filter 12 configured to separate input audio signal 11 into at least bass (low) frequency band 14 and high frequency band 13. Low frequency band 14 may be used as an input to harmonic generator 17 configured to generate a series of harmonics, of integral order n-2....100, by way of example, for the fundamental frequencies that are found in low frequency band 14.

Low frequency band 14 may be further used as an input to energy comparison module 16, which compares energies respectively in high frequency band 14 and low frequency band 13 and provides a measure of auditory or frequency masking when harmonics are added to the audio. Pre-processing module 18 may provide an optional input to energy comparison module 16 and/or to periodicity measurement module 15. Masking amount output from energy comparison 16 may be used to control the amount 29 of generated harmonics from harmonics generator 17 being used for bass enhancement. Gain adjust 19 may be performed directly to gain multiplier 21, and/or by controlling harmonic generator 17 over output 22, further details of which are presented in the description of Figure 4.

Gain adjust 19 with inputs from both periodicity measure 15 and from energy comparison module 16 may compute gain of harmonic generator 17 output, for example by multiplying inputs or otherwise using another function of both inputs such as a power 10 function of one input to the power of the second input.

High pass filter 25 may be used to filter out original low frequencies 14 and the resulting filtered output may be combined by combiner 23 with high frequency band 13 to produce audio output signal 27.

DYNAMIC IMPLEMENTATION OF BASS ENHANCEMENT

During implementation of content adaptive bass enhancement, according to features of the present invention, input audio signal 11 is varying dynamically and bass enhancement is updated in real time. It is therefore desirable to perform efficient real-time computations. Calculation of autocorrelation using the following formula, by way of example: x(i) x(i+d) I 402 suggests a computation over all time delays and all audio samples.

However, delays of interest are limited to periods found in low frequency band 14, e.g. 0.002 to 0.25 seconds, corresponding to 40 -500 Hertz, by way of example.

Moreover, the summations required in the autocorrelation as shown above are 25 computationally heavy. The summations may be replaced with more computationally efficient autoregressive infinite impulse filter (IIR) smoothing filters, one for every d value in the numerator, and another averaging the denominator as follows: IIRsmooth (x x ( + d I) IIRsmooth(x(i)2) I 0 with the maximum providing the period or delay d which is the inverse of a fundamental frequency present in low frequency band 14 and magnitude provides the degree of periodicity or harmonicity.

With use of IIR smoothing filters, attack and release times may be separately controlled, 5 and a different effective range of time sample r may be used when the audio envelope increases (attack), as opposed to when the audio envelope decreases (release).

HARMONIC GENERATOR 17 Reference is now also made to Figure 4 which illustrates schematically a portion of 10 harmonic generator 17, previously disclosed in US patent publication 5,930,373.

Harmonic generator 17 receives as input low frequency signal LFsignal 41, which may be output from optional pre-processing by block 18. Harmonic generator 17 outputs a harmonic PAsignal 42. LFSignal 41 may extend over a range of low frequencies j; . . . and PASignal 42 may extend above another frequency fi. Typically, but not necessarily f, f. For many applications,/ may be in the range/ <fi<f2.

For a specific example of enhancing an inexpensive speaker, which is inefficient below 120 Hertz, the frequencies may be set to /1=40 Hertz, f=120 Hertz,. 120 Hertz. Thus, in this specific example, LFSignal 41 which is predominantly between 40-120 Hz is replaced by a PASignal 42 that is mostly about UO Hertz.

PASignal 42 includes harmonics of the frequencies in LFSignal 41, and may comply with a similar loudness of LFSignal 41. Circuit 17 may integrate both functions the harmonic generation and loudness matching in a single non-linear recursive process.

The operation of harmonic generator 17 includes a non-linear recursive feedback loop 43. Generally, feedback loop 43 recursively multiplies input Lfsignal 41, thus generating 25 higher harmonics at each pass through feedback loop 43. Feedback gain of the loop includes three components: -A frequency dependent gain is applied by an FB-HighPassFilter 45.

-A constant (<1) gain is applied at an attenuator 46 following FB-HighPassFilter 45.

-A dynamic gain 47 is controlled by a compressor control logic block 48 which may 30 receive control parameters dynamically over control input 22 from gain adjust 19 as shown in Figures 1-3. Dynamic gain 47 depends on the energy-envelope of the generator's final output signal at sensor location 49.

An output high pass filter 50 controls which frequencies are allowed at the output of the harmonic generator 17. In our specific example, this would be frequencies greater than 120 Hz. LFSignal 41 including fundamental frequencies is multiplied at block 44 by a delayed feedback signal from delay block 51, thus generating the (NI/)'th order harmonics from any N'th harmonic in the feedback signal multiplication 44.

Note that multiplication 44 also has an effect of inherently increasing dynamic range associated with each harmonic. This increase in dynamic range may be compensated by 10 dynamic gain 47 in order to result with the desired residue expansion ratio.

Another side effect of the multiplication 44 is the generation of intermodulation and direct current (DC) components. Such components that fall below the frequency fi, are attenuated by the filter FB-HPF 45.

Mixer or combiner 52 injects LFSignal 41 into the feedback and maintains fundamental 15 frequencies in the feedback loop 43 at sufficient intensity for successively generating higher harmonics recursively As is well known to those versed in the art, harmonic generator circuit 17 as described herein is only one out of many possible variations (analogue, digital or combination thereof) for accomplishing the same result.

Reference is now also made to Figure 5 which illustrates a flow diagram of a method 6 0 of audio processing, according to features of the present invention. Audio input 11 of audio processor 10A/10B/10C is configured to receive (step 61) an input audio signal. Filter 12 is configured to separate (step 63) multiple frequency bands in the input audio signal. The frequency bands include a bass frequency band 14 which includes one or more fundamental frequencies of the input audio signal. Harmonics generator 17 is configured to generate (step 65) audio harmonics of frequency which are integer multiples of the fundamental frequency. Measurement module(s), e.g. periodicity measurement module 15 and/or energy comparison module 16 is/are configured to measure (step 66) an audio feature of the input audio signal and output correspondingly a measured audio feature. A gain adjustment module 19 is configured to input the measured audio feature of the input audio signal and adapt (step 67) at least in part therefrom, an adapted energy 29 of the audio harmonics generated by harmonics generator 17 for use in bass enhancement. A combiner 23 is configured to combine (step 69) the audio harmonics at the adapted energy with the input audio signal 13 or a processed version of the input audio signal, to produce a bass-enhanced output audio signal at audio output 27.

The embodiments of the present invention may comprise a general-purpose or special-purpose computer system including various computer hardware components, which are discussed in greater detail below. Embodiments within the scope of the present invention also include computer-readable media for carrying or having computer-executable instructions, computer-readable instructions, or data structures stored thereon. Such computer-readable media may be any available media, transitory and/or non-transitory which is accessible by a general-purpose or special-purpose computer system. By way of example, and not limitation, such computer-readable media can comprise physical storage media such as RAM, ROM, EPROM, flash disk, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic or solid state storage devices, or any other media which can be used to carry or store desired program code means in the form of computer-executable instructions, computer-readable instructions, or data structures and which may be accessed by a general-purpose or special-purpose computer system.

Terms as used herein including "processing", "computing", "comparing", "measuring", "adjusting", "determining", "calculating", "receiving", "mitigating", "reducing", "filtering", "extracting", "attenuating" or the like, refer to actions and/or processes of a computer system that manipulate and/or transform data representing physical properties, e.g. audio, electronic, physical objects and/or signals from physical objects.

The computer system as disclosed herein may include an audio driver with a microphone input, an audio processor configured to process input audio signals into a loudspeaker output. Audio processing may normally include sampling an analogue audio signal and performing analogue-to digital conversion to produce therefrom a digital audio signal. Audio inputs and outputs are generally of multiple channels such as stereophonic (two channels) or surround sound (e.g. 5 channels) by way of example. The audio input may include multiple input channels and the audio output may include corresponding multiple output channels. A linking module may be configured to normalise respective loudnesses of the output channels and to maintain interaural level difference (ILD) at the output audio signal at the audio output relative to interaural level difference (ILD) at the audio input.

The term "compare" as used herein is an evaluatation of two or more entities with respect to one another. The term "comparison" as used herein is defined as evaluation with respect to one another of two or more energies in the audio frequency band. The "comparison" may include mathematical operations such as a ratio and/or subtraction, may involve an analytical and/or a numerical computation. In the context of certain methods and systems described herein, any manner of "comparing" may be performed as part of the instructions of a computer program, and such instructions may also include electronic values stored in computer memory.

The term "dynamically" as used herein refers to updating according to changes in time 10 and may depend on dynamic changes of the audio content being input, processed, output, and optionally played, according to features of the present invention.

The term "audio" as used herein refers to a frequency range between 20 Hertz and 20,000 Hertz. The term "bass" as used herein refers generally to the lower portion of the audio frequency range up to about 500 Hertz, with most fundamental frequencies in the bass 15 frequency band between 40 Hertz and 500 Hertz.

The term "harmonicity" as used herein is a measure of the extent a signal may be represented by a sum of a series of harmonics of a single fundamental frequency. The terms "periodicity" and "harmonicity" are used herein interchangeably.

The term "high frequency band as used herein refers to audio frequencies higher than 400 20 Hertz.

The term "fundamental tone" or "fundamental" as used herein is the lowest frequency or bass frequency present of a sound waveform. The waveforms of all sounds, apart from a single sine wave, include the fundamental tone and many other tones of different frequencies. The fundamental tone is referred to as the first harmonic. A tone at twice the frequency of the first harmonic is a second harmonic. A tone played at three times the frequency of the first harmonic is called the third harmonic, and so on.

The term "partials" as used herein are non-fundamental tones with frequencies that are multiplied by fractional amounts, not whole numbers.

The term "distortion" as used herein may include a combination of both linear and 30 nonlinear distortions. The term "linear distortion" refers to a change in amplitude and/or phase without a change in frequency between the output and audio signals. The term "non-linear" distortion includes a change in frequency content between the output and input audio signals.

The term "predicted audibility of distortion" refers to one or more models known in the art which may be used to predict perceived quality of speech and/or music subjected to 5 linear and/or nonlinear distortion.

Such a model may be found in the following representative reference.

Moore BC, Tan CT, Zacharov N, Mattila VV. Measuring and predicting the perceived quality of music and speech subjected to combined linear and nonlinear distortion. Journal of the Audio Engineering Society. 2004 Dec 15;52(12):1228-44.

which reported results of experiments in which subjects rated the perceived quality of speech and music with added linear and nonlinear distortion, with a high correlation between the subjects.

The indefinite articles "a", "an" is used herein, such as "an audio input", "a fundamental frequency" have the meaning of "one or more" that is "one or more audio inputs", "one or 15 more fundamental frequencies".

All optional and preferred features and modifications of the described embodiments and dependent claims are usable in all aspects of the invention taught herein. Furthermore, the individual features of the dependent claims, as well as all optional and preferred features and modifications of the described embodiments are combinable and interchangeable with one another.

Although selected features of the present invention have been shown and described, it is to be understood the present invention is not limited to the described features.

Claims

CLAIMS1. A computerised method performable by an audio processor having an audio input and an audio output, the method comprising: receiving an input audio signal at the audio input; designating a plurality of frequency bands in said input audio signal, the frequency bands including a bass frequency band including a fundamental frequency of the input audio signal; generating audio harmonics of frequency being integer multiples of the fundamental frequency; dynamically measuring a feature of the input audio signal; dynamically adapting energy of the generated audio harmonics, according to the measured feature of the input audio signal; and combining the generated audio harmonics at the dynamically adapted energy with the input audio signal or a processed version of the input audio signal, thereby producing a bass-enhanced output audio signal at the audio output.
2. The computerised method of claim 1, further comprising: dynamically computing a predicted audibility of distortion at the audio output in accordance with the measured feature of the input audio signal.
3. The computerised method 2, further comprising: performing said dynamically adapting energy of the generated audio harmonics according to the predicted audibility of distortion.
4. The computerised method according to any of claims 1 to 3, wherein the designated frequency bands include a higher frequency band of the input audio signal with frequencies higher than the fundamental frequency, the method further comprising: performing a comparison of energy in the bass frequency band to energy in the higher frequency band.
5. The computerised method of claim 4, further comprising: said dynamically adapting energy of the generated audio harmonics at least partly according to the comparison.
6. The computerised method of claim 4, further comprising: dynamically decreasing the energy of the generated audio harmonics when the energy at the fundamental frequency in the bass frequency band of the input audio signal is increased in comparison with the energy of the high frequency band.
7. The computerised method of claim 4, further comprising: dynamically decreasing the adapted energy of the generated audio harmonics when the energy in the high frequency band is higher than a previously determined threshold, and thereby tending to avoid overloading the audio output.
8. The computerised method of claim 4, further comprising: dynamically decreasing the adapted energy of the generated audio harmonics when the generated audio harmonics interact with high frequencies present in the high frequency band due to a difference in at least one of frequency and phase within a previously determined threshold.
9. The computerised method according to any of claims 1 or 2, further comprising: computing a degree of periodicity within the bass frequency band; and said dynamically adapting the adapted energy of the generated audio harmonics at least partly according to said periodicity.
10. The computerised method of claim 9, further comprising: dynamically increasing the energy of the generated audio harmonics when the degree of periodicity is increased.
11. A computer-readable medium comprising instructions which, when executed by a computer system, cause the computer system to execute the method steps according to any of claims 1 to 3.
12. An audio processor comprising: an audio input and an audio output, wherein the audio input is configured to receive an input audio signal; a filter configured to separate a plurality of frequency bands in the input audio signal, the frequency bands including a bass frequency band including a fundamental frequency of the input audio signal; a harmonics generator configured to generate audio harmonics of frequency being integer multiples of the fundamental frequency; a measurement module configured to dynamically measure an audio feature of the input audio signal and output accordingly a measured audio feature; a gain adjustment module configured to input the measured audio feature of the input audio signal and determine at least in part therefrom, an adapted energy of the audio harmonics; and a combiner configured to combine the audio harmonics at the adapted energy with the input audio signal or a processed version of the input audio signal, to produce a bass-enhanced output audio signal at the audio output.
13. The audio processor of claim 12, further configured to dynamically compute a predicted audibility of distortion at the audio output in accordance with the measured feature of the input audio signal.
14. The audio processor of claim 13, further configured to: dynamically adapt energy of the generated audio harmonics according to the predicted audibility of distortion.
15. The audio processor according to any of claims 12 or 13, wherein the frequency bands include a higher frequency band of the input audio signal with frequency higher than the fundamental frequency, the audio processor further comprising: a comparator configured to perform a comparison of energy in the bass frequency band to energy in the higher frequency band and wherein respective energies of the generated audio harmonics are dynamically adapted at least partly according to the compari son.
16. The audio processor of claim 15, wherein the gain adjustment module is configured to dynamically decrease the adapted energy of the generated audio harmonic at the higher frequency when the energy at the fundamental frequency in the bass frequency band of the input audio signal is increased in comparison with the energy of the high frequency band.
17. The audio processor of claim 15, wherein the gain adjustment module is configured to dynamically decrease the adapted energy of the generated audio harmonics when the energy in the high frequency band is higher than a previously determined threshold, to tend to avoid overloading of the audio output.
18. The audio processor of claim 15, wherein the gain adjustment module is configured to dynamically decrease the adapted energy of the generated audio harmonics when the generated audio harmonics interact with high frequencies present in the high frequency band due to a difference in at least one of frequency and phase within a previously determined threshold.
19. The audio processor of claim 12, further comprising: a periodicity measurement module configured to compute a degree of periodicity within said bass frequency band wherein the adapted energy of the generated audio harmonics is dynamically adapted at least partly according to the periodicity.
20. The audio processor of claim 19, further configured to dynamically increase the energy of the generated audio harmonics when the degree of periodicity is increased.