CN112584277B - Indoor audio frequency equalizing method - Google Patents
Indoor audio frequency equalizing method Download PDFInfo
- Publication number
- CN112584277B CN112584277B CN202011424945.8A CN202011424945A CN112584277B CN 112584277 B CN112584277 B CN 112584277B CN 202011424945 A CN202011424945 A CN 202011424945A CN 112584277 B CN112584277 B CN 112584277B
- Authority
- CN
- China
- Prior art keywords
- head
- transfer function
- listening
- indoor
- function
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R3/00—Circuits for transducers, loudspeakers or microphones
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R29/00—Monitoring arrangements; Testing arrangements
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Acoustics & Sound (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Otolaryngology (AREA)
- Stereophonic System (AREA)
Abstract
The application relates to a method for indoor audio playing equalization, which is executed by a processor and comprises the following steps: playing the first audio signal through a loudspeaker in an indoor test state; when a listener changes the listening direction along with time, acquiring the listening direction through a head follower, acquiring a second audio signal corresponding to the listening direction through a microphone, and calculating an indoor transfer function corresponding to the listening direction according to the first audio signal and the second audio signal; traversing indoor listening directions to obtain an indoor transfer function set; according to the indoor transfer function set, taking the personalized head-related transfer function set as an optimization target, calculating an equalization function corresponding to the listening position, and obtaining a set of equalizers corresponding to all the listening positions; when in a listening state, acquiring an indoor listening position in real time through a head follower; acquiring an equalization function corresponding to the listening position from the set of equalizers; passing the first audio signal through an equalizer; and inputting the equalized first audio signal to a loudspeaker for playing.
Description
Technical Field
The application relates to the field of audio playing, in particular to an indoor audio balancing method.
Background
In an audio playback system, the acoustic characteristics of audio played in a limited room space are distorted by the system response of speakers, and the impact responses of sound reflection, sound absorption, and the like of indoor walls and internal objects. Such distortion affects the timbre and timbre of the audio signal, causing the listener to perceive the audio negatively. In some playback systems, the distortion caused by the above-mentioned indoor transfer function to the audio can be corrected by spectrally modifying the input signal to the loudspeaker. This process of spectral modification is known as indoor equalization. Before performing the indoor equalization, the characteristics of the indoor transfer function need to be calculated. However, in the process of testing the indoor transfer function by the traditional method, the traditional microphone can only measure the impact response of one or a plurality of test points, and the impact response of all positions to be listened in the test chamber is difficult to traverse; even if the impact response of the listening position is tested, the microphone placed by the traditional method cannot accurately record the sound pressure of the eardrums of the ears of the listener at the listening position, and the test of the accuracy of the listening position is difficult to perform. Unlike audio played by speakers that a listener listens to at a fixed location, audio played by speakers that a listener listens to while moving needs to satisfy the effect of perceptual balance of timbre and sound energy at different locations.
In practical situations, the traditional indoor audio equalizer cannot perform individual-precise audio signal perception optimization due to different individual binaural positions, orientations relative to speakers, auricle filtering characteristics and the like of the listener, and it is more difficult to ensure that the sound quality and the sound color of the audio perceived by the listener do not change along with the moving position of the listener when the listener walks.
Summary of the invention
In order to solve the above problem, an embodiment of the present application provides a method for indoor audio equalization, which is executed by a processor, and includes: playing a first audio signal through the loudspeaker in an indoor test state; acquiring a first listening orientation by a head follower, acquiring a second audio signal corresponding to the first listening orientation by a microphone disposed in an ear of a listener, and the head follower being disposed on a head of the listener, when the listener changes the listening orientation over time; determining an indoor transfer function corresponding to the first listening position according to the first audio signal and the second audio signal; traversing a first indoor listening position to obtain an indoor transfer function set; according to the indoor transfer function set, taking a head-related transfer function set as an optimization target, and obtaining a balance function corresponding to the first listening position; obtaining a set of equalizers corresponding to all the first listening positions according to the equalization functions corresponding to the first listening positions; the head follower, the loudspeaker and the microphone are in communication relation with the processor; acquiring an indoor second listening orientation in real time by the head follower while a listener wearing the head follower is in a listening state; the second listening orientation comprises a distance, an azimuth, and an elevation of the speaker relative to a head position of the listener; obtaining an equalization function corresponding to the second listening orientation from the set of equalizers; equalizing the first audio signal through the equalization function; and inputting the equalized first audio signal to the loudspeaker for playing.
In one embodiment, the method further comprises: playing the first audio signal through the speaker while a listener wearing the head follower is in a listen-while-test state; acquiring a listening position that varies with time by the head follower, a second audio signal corresponding to the listening position acquired by the microphone, wherein the microphone is disposed in an ear of the listener; calculating an indoor transfer function corresponding to the listening position according to the first audio signal and the second audio signal; calculating an equalizer corresponding to the real-time listening orientation by taking the head-related transfer function corresponding to the real-time listening orientation as an optimization target; equalizing the first audio signal through the equalizer; and feeding back the equalized first audio signal to the loudspeaker for playing.
In one embodiment, the second audio signal corresponding to the first listening position acquired by the microphone includes left and right ear audio signals corresponding to the first listening position acquired by the microphone.
In one embodiment, said determining from said first audio signal, said second audio signal, an indoor transfer function corresponding to said first listening position comprises: obtaining a listening position corresponding to the time interval according to the listening position changing along with the time; segmenting the first audio signal according to a set time interval to obtain a short-time input signal; segmenting the second audio signal according to a set time interval to obtain a left ear short-time sound transmission signal and a right ear short-time sound transmission signal; and according to the short-time input signal and the short-time sound transmission signal, calculating to obtain an indoor transfer function corresponding to the listening direction.
In one embodiment, the equalization function corresponding to the first listening position is obtained by taking the set of head-related transfer functions as an optimization target according to the set of indoor transfer functions; obtaining a set of equalizers corresponding to all first listening orientations according to the equalization function corresponding to the first listening orientation, including: acquiring an omnidirectional head-related transfer function corresponding to personalized parameters of an auricle, a head and shoulders of a listener from a head-related transfer function library to obtain a head-related transfer function set; performing frequency spectrum smoothing on the omnidirectional head related transfer function set to obtain left and right ear head related transfer function frequency spectrums; performing time domain truncation on the indoor transfer function set, removing later reverberation, and performing frequency spectrum smoothing to obtain a left ear response smoothing function set and a right ear response smoothing function set; performing time domain truncation on the indoor transfer function set, removing later reverberation, and performing frequency spectrum smoothing to obtain a left ear response smoothing function set and a right ear response smoothing function set; calculating a public peak value part and personalized parts corresponding to all listening directions according to the left ear response smoothing function set and the right ear response smoothing function set; balancing the public peak part by taking the left and right ear head related transfer function frequency spectrums as an optimization target to obtain a public balancing function; the personalized part is balanced through a calculation cost function to obtain a personalized balance function; convolving the public equalization function and the personalized equalization function in a time domain or multiplying the public equalization function and the personalized equalization function in a frequency domain to obtain an equalizer corresponding to the at least one listening position; convolving the public equalization function and the personalized equalization function in a time domain or multiplying the public equalization function and the personalized equalization function in a frequency domain to obtain an equalizer corresponding to the at least one listening position; traversing all listening positions in the room, obtaining a set of equalizers for all first listening positions.
In one embodiment, the equalization function corresponding to the first listening position is obtained by taking the set of head-related transfer functions as an optimization target according to the set of indoor transfer functions; obtaining a set of equalizers corresponding to all first listening orientations according to the equalization function corresponding to the first listening orientation, including: acquiring an omnidirectional head-related transfer function corresponding to personalized parameters of an auricle, a head and shoulders of a listener from a head-related transfer function library to obtain a head-related transfer function set; performing frequency spectrum smoothing on the omnidirectional head related transfer function set to obtain left and right ear head related transfer function frequency spectrums; performing time domain truncation on the indoor transfer function set, removing later reverberation, and performing frequency spectrum smoothing to obtain a left ear response smoothing function set and a right ear response smoothing function set; taking the left and right ear head related transfer function frequency spectrums as optimization targets, iteratively calculating a second cost function in at least one listening position according to left and right ear response smooth function sets in at least one listening position, and when the second cost function is smaller than a threshold value, obtaining an equalization function corresponding to the at least one listening position; traversing all listening positions in the room, obtaining a set of equalizers for all first listening positions.
In one embodiment, the obtaining an omnidirectional head-related transfer function corresponding to personalized parameters of an auricle, a head and a shoulder of a listener from a head-related transfer function library to obtain a set of head-related transfer functions includes: and performing template matching on the personalized parameters of the auricle, the head and the shoulders of the listener and an existing head-related transfer function database, and extracting the omnibearing head-related transfer function of the listener to obtain a head-related transfer function set.
In one embodiment, the obtaining an omnidirectional head-related transfer function corresponding to personalized parameters of an auricle, a head and a shoulder of a listener from a head-related transfer function library to obtain a set of head-related transfer functions includes: measuring a first head-related transfer function of at least one set listening orientation of the listener in a full anechoic room by a sparseness measure; extracting second head-related transfer functions of other listening orientations of the listener from a head-related transfer function database according to the corresponding priori knowledge and personalized parameters of the auricle, the head and the shoulders of the listener; combining the at least one first head-related transfer function and the second head-related transfer function to obtain a set of omnidirectional head-related transfer functions.
In one embodiment, the equalizing the common peak part with the left and right ear head related transfer function spectrums as an optimization target to obtain a common equalization function; and the personalized part is balanced by calculating a cost function to obtain a personalized balance function, and the method comprises the following steps: calculating a first cost function of at least one listening position by taking the left and right ear head related transfer function frequency spectrums as a balance target; regularizing the first cost function to obtain an amplitude value of a personalized equalization function; and performing minimum phase reconstruction on the amplitude of the personalized equalization function to obtain the personalized equalization function.
In one embodiment, the iteratively calculating the second cost function in the at least one listening orientation further comprises setting a frequency-dependent weighting factor according to a peak distribution of the set of indoor transfer functions, and iteratively calculating the second cost function in the at least one listening orientation according to the frequency-dependent weighting factor.
By the indoor audio equalization method, a listener wears the head follower and the probe microphone during indoor equalization, so that audio signals at eardrums of two ears can be accurately acquired in real time at different listening positions, and indoor transfer functions of the loudspeaker and the two ears can be accurately measured; the personalized indoor transfer function is balanced in real time through an indoor balancing function, and the filtering characteristics of personalized features such as a binaural auricle, a head, a shoulder and the like to sound are reserved when the indoor balancing function is designed, so that the hearing experience of two ears of a listener is optimized; the listener can acquire the optimal audio hearing experience in a precise direction, and can still perceive the audio consistency when moving randomly indoors. The method enables a listener to obtain the best auditory perception experience when listening to the audio played by the loudspeaker indoors.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments disclosed in the present specification, the drawings needed to be used in the description of the embodiments will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments disclosed in the present specification, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.
Fig. 1 is a schematic diagram of a method for indoor audio equalization according to an embodiment of the present application;
fig. 2 is a flowchart illustrating a method for calculating an indoor transfer function according to an embodiment of the present invention;
fig. 3 is a specific flowchart of generating an audio equalizer corresponding to each listening position in the method for indoor audio equalization according to the embodiment of the present application;
fig. 4 is a flowchart of a method of indoor audio equalization provided by an embodiment of the present application;
fig. 5 is a flowchart of an embodiment of calculating an equalization function corresponding to each listening position in a method for indoor audio equalization according to an embodiment of the present application;
fig. 6 is a flowchart of another embodiment of calculating an equalization function corresponding to each listening position in a method for indoor audio equalization according to an embodiment of the present application.
Detailed Description
The technical solution of the present application is further described in detail by the accompanying drawings and examples.
The concept of the present application is introduced first. The head follower is a device capable of tracking the moving direction of a listener, and can be used for acquiring the direction information of the indoor loudspeaker to be equalized relative to the listener in real time. The microphone is a sound transmission device, and a probe type microphone can be respectively inserted into two ears of a listener so as to record an audio signal close to an eardrum when the listener moves in real time. In an indoor audio playing system, when a listener moves, a head follower and a probe microphone are jointly used to measure the indoor transfer functions of the two ears of a loudspeaker, the indoor transfer functions related to the listening position are tracked and measured in real time, and the parameters of the related equalization functions are updated in real time, so that the audio signals input to the loudspeaker can be dynamically equalized in real time.
When the speaker-binaural indoor transfer function is balanced at a plurality of positions of an indoor sound field, noise peaks and valleys in the indoor transfer function caused by indoor and speaker frequency responses need to be removed, key peaks and valleys of the indoor transfer function caused by personalized parameters such as auricles, heads and shoulders of a listener are reserved, meanwhile, consistency balancing is carried out on inherent zero poles of the plurality of positions of the indoor sound field, the moving track of the listener is dynamically tracked, so that the accurate personalized indoor balancing is achieved, and the audio signals perceived by the listener can be kept consistent even if the listener is located at different indoor positions.
Therefore, when in indoor equalization, a listener wears the head follower and the probe microphone, so that the audio signals at the eardrums of the ears can be accurately acquired in real time at different listening positions, the indoor transfer functions of the two ears of the speaker can be accurately measured, the indoor equalization function is constructed, the individualized indoor transfer function is equalized in real time through the indoor equalization function, the filtering characteristics of the individualized characteristics of the ear pinna, the head, the shoulder and the like of the two ears of the listener are reserved when the indoor equalization function is designed, and the hearing experience at the two ears of the listener is optimized.
In a test state, the loudspeaker fixes a sound reproduction direction, a listener wears the head follower and the probe type binaural microphone to move in a room, all positions to be listened in the room are traversed, loudspeaker-binaural room transfer functions corresponding to all the positions to be listened in the test room are arranged into an indoor transfer function library personalized by the listener. Personalized indoor transfer functions may be identified from different listeners. Based on the function library, carrying out zero-pole analysis related to room characteristics, acoustic filtering characteristic analysis related to physiological characteristics such as human head auricles and the like, and spectrum characteristic distortion analysis related to a loudspeaker, and calculating an indoor equalization algorithm; in a listening state, a head follower is used for tracking the moving direction of a listener, and indoor equalization parameters are adjusted in real time by using an equalization algorithm which is calculated and stored in a testing state, so that the balanced perception effect of the sound color of the listener when the listener moves is optimized.
In a test state while listening, a head follower is utilized to track the moving direction of a listener, an indoor transfer function is calculated by utilizing an audio signal input to a loudspeaker and a signal received by a microphone in a corresponding time period, an indoor equalization target is constructed according to the head related transfer function in the corresponding direction, the indoor equalization function in the corresponding direction is calculated, and the audio signal to be played is input to the loudspeaker for playing after being equalized by the indoor equalization function.
The application provides a method for indoor audio equalization, which is executed by a processor connected with a head follower, a microphone and a loudspeaker. The head follower is arranged on the head of a listener and used for tracking the moving track of the listener; the processor can acquire the orientation information of the indoor loudspeaker to be equalized relative to the listener in real time according to the movement track of the listener. The microphone is arranged at the ear of a listener and used for recording an audio signal close to the eardrum when the listener moves in real time, the processor can accurately measure the indoor transfer functions of the loudspeaker and the ears by using the audio signal, construct an indoor equalization function and generate an equalizer, real-time equalization of the played audio signal along the moving track of the listener is realized by the equalizer, and the equalizer generated by the method can effectively enhance the audio perception effect when the listener moves freely indoors.
The method for indoor audio equalization provided by the application needs to perform the following preparatory operations in the testing stage: the sound reproduction direction of the loudspeaker to be balanced is fixed in a room, so that a listener wears a head follower, microphones are respectively placed in two ears, and the microphones can be probe microphones, so that the microphones are as close to the tympanic membrane positions of left and right auditory canals as possible.
Fig. 1 is a flowchart illustrating a method for indoor audio equalization according to an embodiment of the present application. As shown in fig. 1, S101, the processor inputs an audio signal to a speaker to be equalized, and plays the audio signal indoors through the speaker, where the audio signal may be recorded as a first audio signal. S102, enabling a listener to move indoors, traversing all indoor positions, simultaneously recording the movement track of the listener along with time through a head follower by a processor, obtaining the listening position along with time, and recording the listening position as a first listening position; the audio signal, denoted as the second audio signal, is recorded by the microphone as a time-varying audio signal near the eardrums of both ears of the listener. The movement trajectory corresponds in time to the first audio signal and the second audio signal. S103, the processor calculates the indoor transfer function corresponding to each listening position according to the audio signal of the loudspeaker, the audio signal recorded by the microphone and each listening position, and obtains the indoor transfer function set corresponding to each indoor listening position. And S104, the processor acquires the personalized omnidirectional head-related transfer function set corresponding to each listening direction from the head-related transfer function library. S105, according to the indoor transfer function set corresponding to each listening position, taking the head-related transfer function in the personalized omnidirectional head-related transfer function set as an optimization target, and calculating an equalization function corresponding to each listening position; and obtaining the audio equalizer corresponding to each listening position.
Specifically, in the step S101, when obtaining the personalized omnidirectional head-related transfer function set of the listener, the processor may perform template matching on the personalized physiological parameter characteristics of the auricle, the head and the shoulders of the listener and the existing head-related transfer function database to obtain personalized head-related transfer functions HRTFs of respective listening orientations of the listener, or predict the personalized head-related transfer functions of the listener in omnidirectional directions by measuring only the head-related transfer functions HRTFs of a few directions of the listener in an omnidirectional sound-damping room by a sparse measurement method.
In executing S103, the processor may calculate according to the procedure of calculating the indoor transfer function provided in fig. 2 to obtain the indoor transfer function set. As shown in fig. 2, S1031, the processor obtains a time-varying listening orientation from the time-varying movement trajectory of the listener recorded by the head follower. S1032 divides the first audio signal input to the speaker and the second audio signal recorded by the binaural probe microphone according to the time periods corresponding to the different listening orientations, respectively, so that each audio signal of each time period corresponds to the listening orientation of the time period. S1033, according to the first audio signal input to the loudspeaker and the second audio signal recorded by the microphone corresponding to each listening position, the indoor transfer functions of the loudspeaker and the ears in the time period corresponding to different listening positions in the moving track are calculated by using a Normalized Least Mean Square adaptive filtering method (NLMS). In the time domain, the audio signal of the input loudspeaker of each short time period and the time domain convolution result of the indoor transfer function of the corresponding listening position of the same time period approach the audio signal recorded by the probe microphone of the corresponding time period. S1034, recording and sorting the indoor transfer functions corresponding to different listening positions according to the listening positions corresponding to the time intervals and the indoor transfer functions corresponding to the listening positions to obtain an indoor transfer function set.
In executing S105, audio equalizers corresponding to the respective listening positions may be generated according to the flowchart provided in fig. 3. As shown in fig. 3, S1051 performs time domain truncation on the indoor transfer function set, removes late reverberation, and then performs spectrum smoothing. The removed reverberant sound includes the noise response caused by the room and speaker frequency responses. And S1052, analyzing and calculating the frequency response common peak part of the indoor transfer function and the personalized part corresponding to the listening direction. The common peaks represent typical acoustic transfer characteristics in the room, as well as key peaks of the room transfer function caused by the subject's pinna, head, shoulders, etc. The personalized part is a peak value corresponding to different listening directions in the indoor transfer function set. And S1053, performing frequency spectrum smoothing on the head-related transfer function set to obtain a typical peak-valley structure of a frequency domain. S1054, taking the typical peak-valley structure of the head-related transfer function corresponding to the listening position in the frequency domain as an optimization target, spreading and balancing the common peak part of the frequency response of the indoor impact response function to obtain a public balancing function, spreading and balancing the personalized part to obtain a personalized balancing function. And S1055, the public equalization function and the personalized equalization function are cascaded in time domain convolution or frequency domain multiplication to obtain the equalizer corresponding to each listening position, the equalizers corresponding to different listening positions are stored, and the set of the omnibearing equalizers is obtained.
In the actual moving listening stage of the user, the user wears a head follower, and the indoor listening direction is obtained in real time through the head follower; recording the listening orientation as a second listening orientation comprising a distance, an azimuth, and an elevation of the speaker relative to a head position of the listener; acquiring an equalizer corresponding to a listening position from a set of omnidirectional equalizers; equalizing the audio signal through the equalizer; and inputting the equalized audio signal into a loudspeaker for playing.
By the indoor audio balancing method, the listener can obtain the optimal audio hearing experience in the accurate direction, and can still perceive the consistency of audio when moving randomly indoors. The method enables a listener to obtain the best auditory perception experience when listening to the audio played by the loudspeaker indoors.
An embodiment of the present application provides a method for indoor audio equalization, which matches an individualized omnidirectional head-related transfer function set based on physiological characteristic parameters of a listener, uses the function set as an equalization optimization target, identifies speaker-binaural indoor transfer functions in different listening orientations through head follower and microphone measurement, calculates a common polar point model of the indoor transfer functions in all listening orientations, designs a uniform equalization function for the common polar point model, performs equalization on the individualized model of each indoor transfer function by using the head-related transfer function corresponding to each listening orientation as an equalization target, and concatenates the equalization functions of the two models to form an equalizer corresponding to each listening orientation. The method not only realizes accurate equalization of the listening position, but also ensures that the perceived audio frequency is kept unchanged when the listener moves.
Firstly, a preliminary operation is carried out, in a room to be equalized, a listener wears a head follower, and microphones are respectively placed in two ears, and the microphones are as close to the tympanic membrane positions of left and right auditory canals as possible.
Fig. 4 is a flowchart of a method of indoor audio equalization according to an embodiment of the present application. As shown in fig. 4, an audio signal S is input to a speaker, and is played through the speaker indoors S201. The audio signal may be denoted as a first audio signal.
S202, the listener moves indoors and traverses all indoor positions. In the process, the head follower records the moving track m (t) ═ (x (t), y (t), z (t)) and (x, y, z) of the listener along with the time t, wherein the moving track m (t) ═ x (t), y (t), z (t)) is the spatial position information of the listener; the microphone records audio signals near the binaural tympanic membrane, denoted as second audio signals, including a left ear audio signal Record _ L and a right ear audio signal Record _ R.
S203, according to the moving track M (t) recorded by the head follower, determining the orientation information of the loudspeaker relative to the listener along with the time change, and recording the orientation information as the listening orientation
Wherein the content of the first and second substances,
respectively, representing distance, azimuth and elevation information of the speaker relative to the listener's head position.
S204, the audio signal S input to the loudspeaker and the left ear audio signal Record _ L and the right ear audio signal Record _ R recorded by the binaural probe type microphone are segmented according to the time intervals corresponding to different listening orientations, so that the ith time interval tiCorresponding to audio signalsA certain listening position
i is an index value, i ═ 1,2, … n, for example:
i-th period tiThe corresponding audio signal is S (t)i) And is recorded as a short-time input signal; the left ear audio signal is Record _ L (t)i) The right-ear audio signal is Record _ R (t)i) Recording as short-time sound transmission signals of the left ear and the right ear; corresponding to a listening position of
And S205, calculating an indoor transfer function corresponding to the listening position by using an NLMS algorithm according to the short-time input signal, the short-time sound transmission signal and the listening position corresponding to the time interval.
Specifically, the audio signal S (t) corresponding to the speaker is input according to each short periodi) Left ear audio signal Record _ L (t)i) And a right ear audio signal Record _ R (t)i) Calculating indoor transfer functions corresponding to different listening positions by using NLMS algorithm, including left and right ear indoor transfer functions BRIR _ L (t)i) And BRIR _ R (t)i). So that in the time domain representation, the loudspeaker of each short period inputs the audio signal S (t)i) Left and right ear indoor transfer function BRIR _ L (t) of listening position corresponding to respective time periodi) And BRIR _ R (t)i) Approaches the left and right ear audio signals Record by the binaural microphone in corresponding time periods Record _ L (t)i) And Record _ R (t)i)。
The NLMS algorithm has the calculation formula as follows:
wherein the indoor transfer function w (n) ═ w0(n)...wN(n)]TW (N) is a column vector of order N, N represents the estimated acoustic path order, e (N) represents the prediction error signal, w (-) is iterated according to equations (1) and (2) until convergence, with the condition that e (N) is less than a specified threshold.
For an audio signal input to the loudspeaker (denoted as first audio signal),
d (n) is the audio signal recorded by the probe microphone, denoted as the second audio signal, and μ is the convergence factor.
The audio signals d (n) recorded by the microphone include left and right ear audio signals Record _ L and Record _ R according to the input signal of the loudspeaker
And Record _ L and Record _ R can calculate the ith time period tiIs BRIR _ L (t)i) And BRIR _ R (t)i)。
S206, recording the indoor transfer functions corresponding to the i time periods in different listening directions to obtain an indoor transfer function set comprising the left ear transfer function set corresponding to the i time periods
Set of transfer functions to the right ear
S207, according to the indoor transfer function set, taking the head-related transfer function in the head-related transfer function set as an optimization target, and calculating an equalization function corresponding to the listening direction; an equalizer is obtained for each listening position.
Specifically, according to the indoor transfer function set, carrying out analysis calculation, and calculating a common peak part in a frequency domain of all indoor responses, wherein the common peak part represents typical acoustic transfer characteristics in the room; dividing each indoor transfer function into two parts, wherein one part is a frequency response common peak value part of all indoor transfer function responses and is recorded as a common pole model; the other part is the other part in each room response and is marked as a personalized model. Calculating a common equalization function by taking 1 as an optimization target for the common polar point model, and calculating individualized equalization functions corresponding to different listening positions by taking head-related transfer functions in a head-related transfer function set as an optimization target for the individualized model, so that the equalization target of the indoor transfer function approaches to a typical peak-valley structure of the HRTF; a set of equalizers corresponding to each different listening position is obtained.
S208, storing the equalizer corresponding to each listening position to obtain a set of omnibearing equalizers:
when the listener actually moves in the listening stage, the listener wears the head follower, and the head follower tracks and acquires the listening position of the current listener in real time
Selecting equalizer corresponding to listening position from the set of omnibearing equalizers
Equalization is performed on the input audio signal of the loudspeaker in the room.
In the above steps, the flow of one embodiment of S207 of calculating the equalization function corresponding to each listening position is as shown in fig. 5, S2071, the processor performs time domain truncation on the left and right ear indoor transfer function sets, removes the late reverberation, and performs the frequency spectrum smoothing, which is recorded as the left and right ear response smoothing function sets
And
s2072, to
And
constructing an all-pole model:
the all-pole model in equations (2) and (3) includes two models, one model being a common-pole model of all indoor responses, labeled Hc(z), z representing a z-transform of the indoor transfer function; another model is a personalized model in each indoor response except for the common pole, and the personalized model comprises a left ear personalized model
Personalized model of right ear
S2073, calculating a common peak value of the all-pole model to obtain a common-pole model:
in the formula (5), Hc(z) is a common pole model, P represents the order, and a (n) is a coefficient. The optimization objective for the common pole model is 1 to remove the peak-valley, H, in the indoor transfer function caused by the frequency response of the indoor and loudspeakerc(z)Fc(z) ═ 1, according to the common-pole model Hc(z) obtaining a common equalization function Fc(z) is:
s2074, obtain a personalized omni-directional head related transfer function set corresponding to the listener.
Specifically, head-related transfer functions close to the personalized physiological parameters of the auricle, the head and the shoulders of the listener are selected from the existing head-related transfer function database of multiple listeners, the head-related transfer functions surrounding all the orientations of the head of the listener are included, and a personalized omnidirectional head-related transfer function set for the listener is formed. Wherein the personalized omnidirectional head related transfer function set comprises a left ear omnidirectional head related transfer function set and a right ear omnidirectional head related transfer function set
And
wherein the content of the first and second substances,
representing the distance r, azimuth theta and elevation angle of the loudspeaker with respect to the listener's head position
The Head Related Transfer Function (HRTF) is defined as the frequency domain acoustic Transfer Function from the sound source to the ears in free field, which expresses the combined filtering effect of the listener's ear physiological structure on sound waves:
wherein, PL、PRRespectively generating frequency domain complex sound pressure in the left ear and the right ear by a simple harmonic point sound source; p0Is the rear point of head removalThe sound source has a frequency domain complex sound pressure at the center of the original head, and f is the frequency.
S2075, performing spectrum smoothing on the omnidirectional head related transfer function set to obtain typical peak-valley structures of frequency domains, and recording as left and right ear head related transfer function spectrums
And
setting the personalized equalizer corresponding to the personalized model as Fi(z), total equalizer f (z), then:
F(z)＝Fi(z)Fc(z) (9)
calculating a cost function by taking the left and right ear head related transfer function frequency spectrums in the frequency domain f as a balance target
The frequency domain error of equation (10) is minimized.
Due to Hc(z)Fc(z) ═ 1 thus
Thus cost function
Can be simplified into the following form
Where | represents the amplitude and f represents the frequency. The personalized equalization function F of the room transfer function is visualized by equation (11)i(z) taking the smoothed head-related transfer function HRTF asAnd (6) balancing the target. Cost function
Is recorded as a first cost function. The first cost function measures the difference between the frequency response to be equalized and the target response in a certain frequency band of interest.
Personalized equalization function F for solving ith listening position on personalized modeliThe frequency spectrums of the related transfer functions of the left and the right ears are taken as a balance target and pass through a cost function Emag(. to) minimize the frequency domain error of equation (11). In a personalized equalization function FiIn the solving process, in order to prevent the occurrence of ill-condition, the cost function E is required to be usedmag(. cndot.) is normalized as shown in the following equation:
in equation (12), β (f) is the regularization factor, pair
Derivation is carried out to enable partial derivative to be zero, and the regularized optimal personalized equalization function F is obtainediThe amplitude of (d) is:
S2076, the public equalization function and the personalized equalization function corresponding to each listening position are cascaded through convolution in time domain or multiplication in frequency domain to obtain the equalization function corresponding to each listening position
I.e. the equalizer for each listening position:
another embodiment of the present application provides a method for indoor audio equalization, which is based on sparseness measurement to measure head-related transfer functions of a few directions of a listener, the few directions being at least one direction of a setting, and recovering a listener's full-space personalized head-related transfer function HRTF set by using HRTFs of non-measured directions with prior knowledge of a high spatial resolution HRTF database. The personalized head related transfer function HRTF set is used as a balance optimization target, indoor transfer functions of the loudspeaker and the ears in different directions are measured and identified through a head follower and a microphone, a common error weight coefficient is set for the corresponding frequency point of a common peak value model when a balance error cost function is designed, and the common error weight coefficient is enabled to be larger, so that a common polar point model for balancing indoor response is emphasized. The method not only realizes accurate equalization of the listening position, but also ensures that the tone color and tone quality of the audio perceived by the listener when the listening position is moved are kept unchanged.
Specifically, in a test state, firstly, a sound reproduction direction is fixed on an indoor loudspeaker to be balanced, a listener wears a head follower, microphones are respectively arranged in two ears, and the microphones can be probe microphones, so that the microphones are as close to the tympanic membrane positions of left and right auditory canals as possible. Wherein the general flow chart of the method can be referred to fig. 5.
S301, the audio signal is processed
The audio signal is input to a speaker through which the audio signal is played indoors. The audio signal may be denoted as a first audio signal.
S302, the listener moves indoors and traverses all the indoor positions. Recording spatial position information (x, y, z) of a listener along with time t by a head follower worn by the listener to obtain a moving track M (t) (x (t), y (t), z (t)); meanwhile, the probe microphones in both ears Record audio signals d (n) near the eardrums of both ears, and it is noted that the second audio signals include a left-ear audio signal Record _ l (n) and a right-ear audio signal Record _ r (n).
S303, according to the moving track M (t) recorded by the head follower, calculating to obtain the track of the orientation of the loudspeaker relative to the listener changing along with the time, and recording as
Wherein the content of the first and second substances,
respectively, the distance, azimuth and elevation of the speaker with respect to the listener's head position.
S304, extracting time periods corresponding to different listening directions according to the moving track of the loudspeaker relative to the listener along with the time variation of the directions, and carrying out comparison on the audio signals
Record _ L (n) and Record _ R (n) are time-sliced to make the ith short time period tiCorresponding to a certain listening position of a listener
i is an index value, i 1, 2. Wherein, the following steps: ith short period tiIs S (t)i) And is recorded as a short-time input signal; the audio signals of the left and right ears are Record _ L (t)i)，Record_R(ti) Recording as short-time sound transmission signals of the left ear and the right ear; corresponding to a listening position of
S305, according to the short-time input signal and the short-time sound transmission signal corresponding to each short time period and the listening position corresponding to the time period, calculating indoor transfer functions corresponding to different listening positions by using an NLMS algorithm, wherein the indoor transfer functions comprise a left-ear indoor transfer function BRIR _ L (t) and a right-ear indoor transfer function BRIR _ L (t)i) And BRIR _ R (t)i). So that, in the time domain, the speaker of each short period inputs the audio signal S (t)i) Left and right ear indoor transfer function BRIR _ L (t) with corresponding time periodi) And BRIR _ R (t)i) Approximation of the left-ear audio signal Record _ L (t) recorded by the binaural microphone for the respective time periodi) And a right ear audio signal Record _ R (t)i)。
S306, recording the indoor transfer functions corresponding to different listening positions in i time periods to obtain an indoor transfer function set, wherein the indoor transfer function set comprises a left ear transfer function set and a right ear transfer function set
And
s307, according to the indoor transfer function set, taking the head-related transfer function in the omnidirectional head-related transfer function set as an optimization target, and calculating a balance function corresponding to each listening direction; and obtaining the audio equalizer corresponding to each listening position. The difference from S207 is in S3071-S3074:
as shown in fig. 6, S3071, the processor performs time-domain truncation on the left and right ear transfer function sets respectively, removes late reverberation, and performs frequency spectrum smoothing to obtain the left and right ear response smoothing function sets
And
s3072, obtaining the personalized head-related transfer function set of the listener based on sparseness measurement.
Specifically, free-field head-related transfer functions (HRTFs) in a few directions are measured in a full anechoic chamber based on a sparseness measurement method, the few directions are set to be at least one direction, and the HRTFs in the directions which are not measured restore the full-angle space personalized HRTFs of a listener by means of the prior knowledge of the existing HRTF database with high spatial resolution. The personalized head related transfer functions corresponding to the left ear and the right ear are respectively
And
wherein the content of the first and second substances,
respectively, representing the distance, azimuth and elevation of the loudspeaker relative to the listener's head position.
S3073, performing frequency spectrum smoothing on the personalized head related transfer function set to obtain a typical peak-valley structure of a frequency domain, and recording the typical peak-valley structure as the typical peak-valley structure
And
s3074, with the typical peak-valley structure of the frequency domain as an optimization target, calculating a cost function to solve the equalizer corresponding to each listening position. And setting different frequency band weight coefficients in the cost function, adopting a unified weight coefficient for a common peak part, and adopting respective weight coefficients for personalized parts at other positions.
Specifically, the left and right ear response smoothing function sets are subjected to spectrum analysis, and a left ear head related transfer function of f frequency points is used
For optimizing the target, for listening position
Is as follows
Carrying out equalization to obtain the right ear head related transfer function with f frequency point
For optimizing the aim, for different listening orientations
Is as follows
Carrying out equalization, wherein i is 1, 2.. n, and equalization of the left ear and the right ear adopts the same equalization function, and is set as
Calculating different listening orientations
Cost function for lower equalization optimization
Comprises the following steps:
cost function in equation (15)
The second cost function is used for measuring the difference between the frequency response to be equalized and the target response in a certain concerned frequency band. n isjAnd nkRespectively indicating listening position
Starting frequency n of audio signal to be equalizedjAnd a termination frequency nk(ii) a W (-) is a weight coefficient changing with frequency change f, a larger and common weight coefficient W (-) is adopted for the frequency points corresponding to the common peak value of the frequency domain of the indoor transfer function set, the weight coefficient W (-) is adopted for the personalized model part of the frequency domain of the indoor transfer function set, and the weight coefficient is distributed according to the peak value distribution condition.
S3075, carrying out random iteration on the formula (15) until convergence, wherein the convergence condition is
And when the value is smaller than the specified threshold value, obtaining a specific function expression of the equalization function Eq.
S308, storing the equalizer corresponding to each listening position to obtain a set of omnibearing equalizers:
and under the actual moving listening state of the listener, the listener wears the head follower, the head follower tracks and records the listening direction of the listener in real time, and the equalizer corresponding to the listening direction is selected from the set of the omnibearing equalizers to perform equalization on the input audio signal of the indoor loudspeaker.
The method for indoor audio equalization provided in another embodiment of the present application may further be performed in a test state of a listener wearing the head follower while listening, and specifically, may play a first audio signal through a speaker, acquire a listening direction changing with time through the head follower, and acquire a second audio signal corresponding to the listening direction through the microphone, where the microphone is disposed in the ear of the listener; calculating an indoor transfer function corresponding to the listening position according to the first audio signal and the second audio signal; calculating an equalizer corresponding to the listening position by taking a head-related transfer function corresponding to the real-time listening position as an optimization target; equalizing the first audio signal through the equalizer; and feeding back the equalized first audio signal to the loudspeaker for playing.
According to the indoor audio equalization method, the head follower is used for tracking the movement track of the listener to carry out indoor equalization, the improvement of an indoor equalization technology can be promoted, and the balance between the auditory perception of the listener when listening to the audio indoors and the audio perception at different listening positions can be accurately optimized.
It will be further appreciated by those of ordinary skill in the art that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether these functions are performed in hardware or software depends on the particular application of the solution and design constraints. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied in hardware, a software module executed by a processor, or a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The above-mentioned embodiments, objects, technical solutions and advantages of the present application are described in further detail, it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present application, and are not intended to limit the scope of the present application, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present application should be included in the scope of the present application.
Claims (9)
1. A method for indoor audio equalization performed by a processor, the method comprising:
playing the first audio signal through a loudspeaker in an indoor test state; acquiring a first listening orientation by a head follower, acquiring a second audio signal corresponding to the first listening orientation by a microphone disposed in an ear of a listener, and the head follower being disposed on a head of the listener, when the listener changes the listening orientation over time; determining an indoor transfer function corresponding to the first listening position according to the first audio signal and the second audio signal; traversing a first indoor listening position to obtain an indoor transfer function set;
according to the indoor transfer function set, taking a head-related transfer function set as an optimization target, and obtaining a balance function corresponding to the first listening position; obtaining a set of equalizers corresponding to all first listening orientations according to the equalization function corresponding to the first listening orientation, including:
acquiring an omnidirectional head-related transfer function corresponding to personalized parameters of an auricle, a head and shoulders of a listener from a head-related transfer function library to obtain a head-related transfer function set; performing frequency spectrum smoothing on the omnidirectional head related transfer function set to obtain left and right ear head related transfer function frequency spectrums; performing time domain truncation on the indoor transfer function set, removing later reverberation, and performing frequency spectrum smoothing to obtain a left ear response smoothing function set and a right ear response smoothing function set; calculating a public peak value part and personalized parts corresponding to all listening directions according to the left ear response smoothing function set and the right ear response smoothing function set; balancing the public peak part by taking the left and right ear head related transfer function frequency spectrums as an optimization target to obtain a public balancing function; the personalized part is balanced through a calculation cost function to obtain a personalized balance function; the public equalization function and the personalized equalization function are convoluted in a time domain or multiplied in a frequency domain to obtain an equalizer corresponding to at least one listening position; traversing all indoor listening positions to obtain a set of equalizers of all first listening positions;
the head follower, the loudspeaker and the microphone are in communication relation with the processor;
acquiring an indoor second listening orientation in real time by the head follower while a listener wearing the head follower is in a listening state; the second listening orientation comprises a distance, an azimuth, and an elevation of the speaker relative to a head position of the listener; obtaining an equalization function corresponding to the second listening orientation from the set of equalizers; equalizing the first audio signal through the equalization function; and inputting the equalized first audio signal to the loudspeaker for playing.
2. The method for indoor audio equalization of claim 1, further comprising: playing the first audio signal through the speaker while a listener wearing the head follower is in a listen-while-test state; acquiring a listening position that varies with time by the head follower, a second audio signal corresponding to the listening position acquired by the microphone, wherein the microphone is disposed in an ear of the listener; calculating an indoor transfer function corresponding to the listening position according to the first audio signal and the second audio signal; calculating an equalizer corresponding to the listening position by taking the head-related transfer function corresponding to the listening position as an optimization target; equalizing the first audio signal through the equalizer; and feeding back the equalized first audio signal to the loudspeaker for playing.
3. The method of indoor audio equalization of claim 1, wherein the second audio signal acquired by the microphone corresponding to the first listening orientation comprises a left ear audio signal and a right ear audio signal acquired by the microphone corresponding to the first listening orientation.
4. The method of indoor audio equalization of claim 1, wherein said determining an indoor transfer function corresponding to the first listening position from the first audio signal and the second audio signal comprises:
obtaining a listening position corresponding to a set time period according to the listening position changing along with time;
segmenting the first audio signal according to a set time interval to obtain a short-time input signal;
segmenting the second audio signal according to a set time interval to obtain a left ear short-time sound transmission signal and a right ear short-time sound transmission signal;
and according to the short-time input signal and the short-time sound transmission signal, calculating to obtain an indoor transfer function corresponding to the listening direction.
5. The method for indoor audio equalization according to claim 1, wherein the equalization function corresponding to the first listening position is obtained according to the indoor transfer function set with a head-related transfer function set as an optimization target; obtaining a set of equalizers corresponding to all first listening orientations according to the equalization function corresponding to the first listening orientation, including:
acquiring an omnidirectional head-related transfer function corresponding to personalized parameters of an auricle, a head and shoulders of a listener from a head-related transfer function library to obtain a head-related transfer function set;
performing frequency spectrum smoothing on the omnidirectional head related transfer function set to obtain left and right ear head related transfer function frequency spectrums;
performing time domain truncation on the indoor transfer function set, removing later reverberation, and performing frequency spectrum smoothing to obtain a left ear response smoothing function set and a right ear response smoothing function set;
taking the left and right ear head related transfer function frequency spectrums as optimization targets, iteratively calculating a second cost function in at least one listening position according to left and right ear response smooth function sets in at least one listening position, and when the second cost function is smaller than a threshold value, obtaining an equalization function corresponding to the at least one listening position;
traversing all listening positions in the room, obtaining a set of equalizers for all first listening positions.
6. The method for indoor audio equalization according to claim 1 or 5, wherein the obtaining an omnidirectional head-related transfer function corresponding to personalized parameters of auricle, head and shoulder of a listener from a head-related transfer function library to obtain a personalized head-related transfer function set comprises:
and performing template matching on the personalized parameters of the auricle, the head and the shoulders of the listener and an existing head-related transfer function database, and extracting the omnibearing head-related transfer function of the listener to obtain a head-related transfer function set.
7. The method for indoor audio equalization according to claim 1 or 5, wherein the obtaining an omnidirectional head-related transfer function corresponding to personalized parameters of auricles, head and shoulders of a listener from a head-related transfer function library to obtain a head-related transfer function set comprises:
measuring a first head-related transfer function of at least one set listening orientation of the listener in a full anechoic room by a sparseness measure;
extracting second head-related transfer functions of other listening orientations of the listener from a head-related transfer function database according to the corresponding priori knowledge and personalized parameters of the auricle, the head and the shoulders of the listener;
combining the at least one first head-related transfer function and the second head-related transfer function to obtain a set of omnidirectional head-related transfer functions.
8. The method for equalizing indoor audio frequency according to claim 1, wherein the equalizing the common peak portion with the left and right ear-related transfer function spectrums as optimization targets to obtain a common equalizing function; and the personalized part is balanced by calculating a cost function to obtain a personalized balance function, and the method comprises the following steps:
calculating a first cost function of at least one listening position by taking the left and right ear head related transfer function frequency spectrums as a balance target;
regularizing the first cost function to obtain an amplitude value of a personalized equalization function;
and performing minimum phase reconstruction on the amplitude of the personalized equalization function to obtain the personalized equalization function.
9. The method for indoor audio equalization of claim 5, wherein said iteratively calculating a second cost function in said at least one listening orientation further comprises setting a frequency-dependent weighting factor according to a peak distribution of said set of indoor transfer functions, and iteratively calculating a second cost function in said at least one listening orientation according to said frequency-dependent weighting factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011424945.8A CN112584277B (en) | 2020-12-08 | 2020-12-08 | Indoor audio frequency equalizing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011424945.8A CN112584277B (en) | 2020-12-08 | 2020-12-08 | Indoor audio frequency equalizing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112584277A CN112584277A (en) | 2021-03-30 |
| CN112584277B true CN112584277B (en) | 2022-04-22 |
Family
ID=75128022
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011424945.8A Active CN112584277B (en) | 2020-12-08 | 2020-12-08 | Indoor audio frequency equalizing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112584277B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114171048A (en) * | 2021-11-30 | 2022-03-11 | 雷欧尼斯（北京）信息技术有限公司 | Indoor multi-position frequency domain audio frequency equalization method, device and equipment |
| CN114598974B (en) * | 2022-03-11 | 2024-02-27 | 广州大学 | Bone conduction earphone equalization method based on distortion product otoacoustic emission |
| CN114662663B (en) * | 2022-03-25 | 2023-04-07 | 华南师范大学 | Sound playing data acquisition method of virtual auditory system and computer equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102055425A (en) * | 2009-11-02 | 2011-05-11 | 哈曼贝克自动系统股份有限公司 | Audio system phase equalizion |
| CN108449688A (en) * | 2018-03-19 | 2018-08-24 | 长沙世邦通信技术有限公司 | Studio broadcast audio-frequency processing method, apparatus and system |
| CN110021306A (en) * | 2018-01-07 | 2019-07-16 | 创新科技有限公司 | Method for generating Custom Space audio using head tracking |
| CN111107482A (en) * | 2018-10-25 | 2020-05-05 | 创新科技有限公司 | System and method for modifying room characteristics for spatial audio rendering through headphones |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413547B (en) * | 2013-07-23 | 2016-03-02 | 大连理工大学 | A kind of method that room reverberation is eliminated |
| DE102014210215A1 (en) * | 2014-05-28 | 2015-12-03 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Identification and use of hearing room optimized transfer functions |
-
2020
- 2020-12-08 CN CN202011424945.8A patent/CN112584277B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102055425A (en) * | 2009-11-02 | 2011-05-11 | 哈曼贝克自动系统股份有限公司 | Audio system phase equalizion |
| CN110021306A (en) * | 2018-01-07 | 2019-07-16 | 创新科技有限公司 | Method for generating Custom Space audio using head tracking |
| CN108449688A (en) * | 2018-03-19 | 2018-08-24 | 长沙世邦通信技术有限公司 | Studio broadcast audio-frequency processing method, apparatus and system |
| CN111107482A (en) * | 2018-10-25 | 2020-05-05 | 创新科技有限公司 | System and method for modifying room characteristics for spatial audio rendering through headphones |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112584277A (en) | 2021-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112584277B (en) | Indoor audio frequency equalizing method | |
| US9918179B2 (en) | Methods and devices for reproducing surround audio signals | |
| US9552840B2 (en) | Three-dimensional sound capturing and reproducing with multi-microphones | |
| JP4343845B2 (en) | Audio data processing method and sound collector for realizing the method | |
| US20080137870A1 (en) | Method And Device For Individualizing Hrtfs By Modeling | |
| CN106664497B (en) | Audio reproduction system and method | |
| EP1816895B1 (en) | Three-dimensional acoustic processor which uses linear predictive coefficients | |
| US20080306720A1 (en) | Hrtf Individualization by Finite Element Modeling Coupled with a Corrective Model | |
| Sakamoto et al. | Sound-space recording and binaural presentation system based on a 252-channel microphone array | |
| JP2009512364A (en) | Virtual audio simulation | |
| JP7208365B2 (en) | Apparatus and method for adapting virtual 3D audio into a real room | |
| CN112492445B (en) | Method and processor for realizing signal equalization by using ear-covering type earphone | |
| Thiemann et al. | A multiple model high-resolution head-related impulse response database for aided and unaided ears | |
| Lee et al. | A real-time audio system for adjusting the sweet spot to the listener's position | |
| Gupta et al. | Parametric hear through equalization for augmented reality audio | |
| CN112492446B (en) | Method and processor for realizing signal equalization by using in-ear earphone | |
| KR100862663B1 (en) | Method and apparatus to localize in space position for inputting signal. | |
| CN114586378A (en) | Partial HRTF compensation or prediction for in-ear microphone arrays | |
| Deng et al. | Modeling and estimating acoustic transfer functions of external ears with or without headphones | |
| JPH09191500A (en) | Method for generating transfer function localizing virtual sound image, recording medium recording transfer function table and acoustic signal edit method using it | |
| Schwark et al. | Data-driven optimization of parametric filters for simulating head-related transfer functions in real-time rendering systems | |
| US10887717B2 (en) | Method for acoustically rendering the size of sound a source | |
| Fejgin et al. | BRUDEX database: Binaural room impulse responses with uniformly distributed external microphones | |
| Delebecque et al. | BinauRec: A dataset to test the influence of the use of room impulse responses on binaural speech enhancement | |
| Klunk | Spatial Evaluation of Cross-Talk Cancellation Performance Utilizing In-Situ Recorded BRTFs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |