DE102015110938B4 - Method for modifying an impulse response of a sound transducer - Google Patents

Abstract

A method for modifying an impulse response (201) of a digital sound transducer (103), wherein from the impulse response (201) by means of Fourier transform (202) a complex frequency response (203) is determined, wherein the complex frequency response (203) in an amplitude response (204 ) and a phase response (205), wherein a digital low pass filter (206) is applied to the amplitude response (204) and a filtered amplitude response (207) is obtained, wherein the filtered amplitude response (207) and the phase response (205) are modified Complex frequency response (208) is determined, and wherein from the modified complex frequency response (208) by means of inverse Fourier transform (209) a modified impulse response (210) is determined. wherein an IIR filter is used as the digital low-pass filter (206), characterized in that a first IIR filter is applied from the low to the high frequencies of the amplitude response and a second IIR filter from the high to the low frequencies of the Amplitude response is applied.

Description

The present invention relates to a method for modifying an impulse response of a sound transducer and to a sound transducer suitable for carrying out the method.

The US 2014/0044273 A1 relates to a direct sound pickup device comprising: a spectrum transform unit that transforms an input signal having a reverberation signal in an input signal on which a Fourier transform has been performed into a first amplitude spectrum signal Lfa; a low pass filter unit that performs a low pass filtering process for the first amplitude spectrum signal Lfa for each frequency to obtain a second amplitude spectrum signal Lfa1; a first subtraction unit that calculates a third amplitude spectrum signal by subtracting the second amplitude spectrum signal Lfa1 from the first amplitude spectrum signal Lfa; and an inverse Fourier transform unit that generates a direct sound signal Lfd by means of a frequency spectrum signal calculated based on a phase spectrum signal and the third amplitude spectrum signal.

In the US 2011/0064244 A1 For example, a method and apparatus for processing audio data will be described. At least a first spectrum with at least one first spectral property is removed from the spectrum of audio data, the resulting spectrum of the audio data is transformed after removal of the at least one first spectrum and the at least one first spectrum or at least one of the first spectrum and / or at least one second spectrum with at least one second spectral property of the transformed spectrum are impressed.

The DE 38 56 269 T2 relates to a digital equalizer comprising a digital filter for frequency compensation of an audio signal which has been converted to a digital code sampling form. More particularly, the invention relates to such an apparatus employing a finite impulse response (FIR) digital filter and allows for independent adjustment of the amplitude / frequency and phase / frequency response characteristics of the filter.

In the DE 697 00 084 T2 is a method of transforming a periodic signal, a method of transforming sound and a method of analyzing a signal, and more particularly, a method of transforming a periodic signal to transform sound, a method of transforming sound, and a method of analyzing a signal for analyzing sound.

The US 2013/0019739 A1 relates to a sound analysis and associated sound synthesis method comprising: (a) receiving a first input sound signal; (b) analyzing the acoustic input signal to determine its impulse response representative of the timbre of the input sound signal; (c) receiving a second input sound signal; (d) processing the second input sound signal into a form to which the impulse response can be applied, the processing comprising generating a pink noise equivalent frequency spectrum of the second input sound signal; and (e) applying the impulse response to the processed second input sound signal to produce an output signal that includes at least portions of the timbre of the first input sound signal.

In electric stringed instruments, in particular electric guitars (electric guitars), electric basses (electric basses) and electric violins (electric violins), the vibrations of the strings with pickups are converted into electrical signals. The electrical signals are transmitted by means of a stringed instrument amplifier, e.g. of a guitar amplifier, amplified and air-vibrated by means of a stringed instrument loudspeaker, i. Sound, transformed.

The frequency response of the combination stringed instrument amplifiers and speakers is typically nonlinear, so that the particular stringed instrument amplifiers and speakers greatly contribute to the sound perceived by a listener located near the stringed instrument speaker.

For example, Guitar players deliberately use different string instrument amplifiers and speakers with different frequency responses to affect the sound that the listener hears. In particular, they make strong use of distortions to the sound design.

The different frequency responses of different stringed instrument amplifiers and loudspeakers are deliberately used by instrumentalists, especially e-guitar players, to influence the sound.

If the sound is to be made accessible to a larger group of people, the sound is typically converted again with a microphone located near the loudspeaker into electrical signals which, if necessary, after an intermediate recording (CD, MP3), via a substantially linear music amplifier and in the Substantially linear music speakers are converted back into sound.

The sound of the instrument speaker, even with essentially linear characteristics of the music amplifier and the music loudspeaker, may differ quite considerably from the sound which the listener perceives of the sound emitted by the music loudspeaker. Namely, the sound of the music loudspeaker has additional so-called colorations, which may be due to the Abmikrophonieren.

The listeners of a concert or listener of a music recording typically only know the sound provided with these additional colors. As a rule, only the instrumentalist, e.g. the electric guitar player, the direct sound of the string instrument speaker.

The provision of a stringed instrument amplifier and loudspeaker for concerts is associated with a high logistical effort. Frequently, string instrument amplifiers based on tube amplifiers are used which have a high weight. In addition, the change of a stringed instrument amplifier and / or loudspeaker during a concert regularly requires a complex rewiring and / or new tuning of the music amplifier.

A sound transducer, which is to replace the well-known string instrument amplifiers tube-based, is initially in the US 5,789,689 A been described. Next is in the EP 1 883 064 A1 a method has been proposed for adapting a sound transducer to a reference sound transducer, the sound transducer having a first vibration profile with a first frequency response, a second vibration profile having a second frequency response and a trivial non-linearity connected between the first and second vibration profiles, and wherein the sound transducer has a self-profile which corresponds to the frequency response resulting from the combination of the first oscillation profile, the trivial nonlinearity and the second oscillation profile, the method comprising the steps of determining a first reference profile of the reference transducer at a low input level and a second reference profile of the reference transducer at a high input level; the first and the second frequency response of the sound transducer, so that at low input level, in which the intrinsic profile of the sound transducer is not influenced by the nonlinearity, the Pr odukt the first and the second frequency response corresponds to and regulating the first and the second frequency response of the sound transducer, so that at high input level, in which the intrinsic profile of the sound transducer is influenced by the non-linearity, the difference between the intrinsic profile of the sound transducer and the second reference profile is minimized and the product of the first and second frequency responses still corresponds to the first reference profile.

With the known method, a sound transducer, in particular its impulse response, can be adapted to a specific string instrument amplifier / speaker combination, so that when playing the electronic string instrument connected to the sound transducer via the music amplifier and loudspeaker, the sound corresponds to that which the listener also hears in the classical signal chain, perceive electric string instrument string instrument amplifiers / speakers microphone music amplifiers / speakers. The impulse response may correspond, for example, to the above-mentioned second oscillation profile. The frequency response of the sound transducer can combine the sound characteristics of the combination string instrument speaker and microphone.

In contrast to the classical signal chain, the instrumentalist now perceives the same sound as the listeners, in particular the colorings caused inter alia by the demodulating. An instrumentalist who has become accustomed to the direct sound of the string instrument speaker may find the tone with the colorations unbalanced and uncomfortable, although this sound is the predominant sound in recordings and larger live concerts.

On this basis, the object of the present invention is to specify a method for modifying the impulse response of a digital sound transducer and a digital sound transducer with a modified impulse response, so that the sound perceived by the instrumentalist of the digital sound converter more closely resembles the usual sound of the classical string instrument amplifiers / loudspeakers. Combination corresponds.

For this purpose, a method according to the main claim 1 and a digital sound transducer according to claim 10 is proposed. Advantageous embodiments of the method are described in the dependent claims 2 to 9 and can be applied in an equivalent manner to the digital sound transducer according to claim 10.

A method for modifying an impulse response of a digital sound transducer, wherein a complex frequency response is determined from the impulse response by means of Fourier transform, wherein the complex frequency response is decomposed into an amplitude response and a phase response, wherein a digital low-pass filter is applied to the amplitude response and a filtered amplitude response is, taking from the filtered amplitude response and the phase response is a modified complex frequency response is determined, and wherein from the modified complex frequency response by inverse Fourier transform a modified impulse response is determined, it may allow an instrumentalist even when using a digital sound converter his usual analogue stringed instruments / -lautsprechern sound image for To make available.

As a Fourier transformation, a discrete Fourier transform, in particular an FFT (fast Fourier transform) can be used. The use of the FFT can make use of existing program functions and optimized hardware components.

In a first embodiment of the method for modifying an impulse response of a digital sound transducer, a FIR filter is used as the digital low-pass filter. An FIR filter (finite impulse filter) can have the advantage that quantization errors affect the filter coefficients only to a small extent.

In another embodiment, a linear phase FIR filter is used. The use of a linear phase FIR filter can avoid a shift in the amplitude response.

According to one embodiment of the method for modifying an impulse response of a digital sound transducer, an IIR filter is used as the digital low-pass filter. The use of an infinite impulse response filter (IIR) for the digital low pass filter may allow for a lower order filter stage compared to an FIR filter, thus limiting the computational overhead.

Furthermore, an embodiment of the method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ) that a first IIR filter be applied from the low to the high frequencies of the amplitude response, and a second IIR filter be applied from the high to the low frequencies of the amplitude response. By applying a first IIR filter from the low to the high frequencies and a second IIR filter from the high to the low frequencies, a frequency shift can be avoided.

According to a further exemplary embodiment of the method for modifying an impulse response of a digital sound transducer, it is provided that the use frequency of the digital low-pass filter is adapted for each magnitude of the amplitude response.

According to another embodiment of the method for modifying an impulse response of a digital sound transducer, the frequency of use of the digital low-pass filter is adjusted for each magnitude of the amplitude response by 1 / F, where F is the frequency associated with the magnitude or its frequency index. In this way, aurally correct modification of the impulse response can be achieved.

In accordance with a further embodiment of the method for modifying an impulse response of a digital sound converter, the frequency of use of the digital low-pass filter for frequencies greater than 2 kHz, in particular greater than 5 kHz, is adjusted more strongly than with 1 / F, interfering colorations can be reduced even more without to negatively influence the sound character of the sound transducer.

According to another embodiment of the method for modifying an impulse response of a digital sound transducer, the absolute use frequency of the digital low-pass filter can be predetermined by a user. In this way, the digital sound transducer can be adapted to the preferences of the instrumentalist.

In the drawing shows

1 a concert situation;

2 a flowchart of a method for modifying modify the impulse response of a digital sound transducer;

3 an exemplary impulse response of a digital sound transducer;

4 a linear representation of the impulse response corresponding amplitude response;

5 a logarithmic representation of the amplitude response;

6 a linear representation of the unfiltered and filtered amplitude response;

7 a logarithmic representation of the unfiltered and filtered amplitude response; and

8th the unfiltered and filtered amplitude response corresponding unfiltered and modified impulse response

1 shows the example of a guitarist 101 with an electric guitar 102 an instrumentalist with a stringed instrument attached to a digital sound transducer 103 connected. With the digital sound converter 103 with its own impulse response transformed sound is via a music amplifier / speaker 104 with essentially linear reinforcement the concert audience 105 sonication.

For example, in a larger concert, a public address system (PA) is used as the music amplifier / speaker to reproduce the speech and / or music. With a PA system it is possible to sound large areas substantially evenly.

Likewise, one is the guitarist 101 facing reinforcement system 106 , which is also called monitor system, provided. This allows the guitarist 101 the perception and adaptation of his own guitar playing. Unlike traditional systems where the guitarist 101 uses a stringed instrument amplifier and speaker and the sound is microphoned, the guitarist takes 101 now almost the same sound as the concert audience 105 , In particular, the guitarist listens 101 also for him unfamiliar and therefore disturbing colorations, which are caused by the Abmikrophonieren the stringed instrument amplifier and speaker.

In the 2 the sequence of a method for modifying the impulse response of a digital sound transducer is shown by way of example. The starting point is the present impulse response 201 of the digital sound converter. This impulse response 201 may have been obtained, for example, by adapting the digital sound transducer to a reference sound transducer, as in the above-mentioned EP 1 883 064 A1 has been described.

The impulse response 201 is first using Fourier transform 202 in a complex frequency response 203 converted and then in amplitude response 204 and phase response 205 disassembled. The amplitude response is subsequently a low-pass filtering 206 subjected, so a filtered amplitude response 207 is obtained. The filtered amplitude response 207 becomes with the (unchanged) phase response 205 to a modified complex frequency response 208 merged, which by means of inverse Fourier transformation 209 in a modified impulse response 210 is converted.

A sound transducer with a pulse response modified in this way 210 It may allow an instrumentalist to perceive the sound of his stringed instrument with diminished, often annoying, staining and heighten confidence in his own playing.

In the 3 is an example of an impulse response of a digital sound transducer shown. The sound transducer may have been adapted to a reference sound transducer as described above. On the abscissa axis the time is plotted in samples and on the ordinate axis the associated discrete amplitude values.

4 shows a linear representation of the impulse response according to 1 corresponding, typical amplitude response, which can be obtained by means of Fourier transform. After the transformation, the amplitudes are equally distributed in the frequency domain. Clearly visible is an increasing ripple of the amplitude response over 2 kHz, in particular over 5 kHz, which can be perceived by listeners as disturbing. In the 5 the magnitudes are again plotted over a logarithmic frequency axis. In the logarithmic, ie aurally correct plot, it also becomes clear that the ripple becomes denser with increasing frequency. The above-described ripple can hardly be mitigated in a conventional recording studio situation in a conventional Abmikrophonierung even with downstream equalizers because too many individual resonances would have to be recorded in an analysis and regulated accordingly.

The effect of the digital low-pass filter on the amplitude response is in the 6 and 7 shown. The magnitudes of the filtered amplitude response are each connected to a solid line and contrasted with the unfiltered amplitude response shown as a broken line. It can be seen that, especially for frequencies greater than four thousand hertz, the filtering leads to a sharp decrease in the ripple. At the same time, however, the magnitudes in the lower frequency range, which make up the essential sound character of the digital sound transducer, are only slightly affected.

In the 8th is the modified impulse response of the digital sound transducer shown, which is obtained after combining the (original) phase response and the filtered amplitude response to a modified complex frequency response followed by an inverse Fourier transform. Recognizable here is a significant smoothing with increasing number of samples.

Claims (7)

Method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ), where from the impulse response ( 201 ) by means of Fourier transformation ( 202 ) a complex frequency response ( 203 ) is determined where the complex frequency response ( 203 ) into an amplitude response ( 204 ) and a phase response ( 205 ), using a digital low-pass filter ( 206 ) on the amplitude response ( 204 ) and a filtered amplitude response ( 207 ), wherein from the filtered amplitude response ( 207 ) and the phase response ( 205 ) a modified complex frequency response ( 208 ), and where from the modified complex frequency response ( 208 ) by inverse Fourier transformation ( 209 ) a modified impulse response ( 210 ) is determined. being a digital low-pass filter ( 206 ) an IIR filter is used, characterized in that a first IIR filter is applied from the low to the high frequencies of the amplitude response and a second IIR filter is applied from the high to the low frequencies of the amplitude response. Method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ) according to claim 1, characterized in that the frequency of use of the digital low-pass filter ( 206 ) is adjusted for each magnitude of the amplitude response. Method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ) according to one of the preceding claims 1 or 2, characterized in that the use frequency of the digital low-pass filter ( 206 ) is adjusted for each magnitude of the amplitude response with 1 / F. Method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ) according to claim 2, characterized in that the use frequency of the digital low-pass filter ( 206 ) is adjusted for frequencies greater than 2 kHz, in particular greater than 5 kHz, stronger than with 1 / F. Method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ) According to claim 4, characterized in that the cutoff frequency of the digital low pass filter ( 206 ) is adjusted for frequencies greater than 5 kHz, stronger than with 1 / F. Method for modifying an impulse response ( 201 ) of a digital sound transducer ( 103 ) according to one of the preceding claims, that the absolute input frequency of the digital low-pass filter is predetermined by a user. Digital sound converter for carrying out a method according to one of the claims 1 to 6.

Images