CN112511966A - Self-adaptive active frequency division method for vehicle-mounted stereo playback - Google Patents
Self-adaptive active frequency division method for vehicle-mounted stereo playback Download PDFInfo
- Publication number
- CN112511966A CN112511966A CN202011230703.5A CN202011230703A CN112511966A CN 112511966 A CN112511966 A CN 112511966A CN 202011230703 A CN202011230703 A CN 202011230703A CN 112511966 A CN112511966 A CN 112511966A
- Authority
- CN
- China
- Prior art keywords
- frequency
- frequency division
- loudspeaker
- low
- program
- 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.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04S—STEREOPHONIC SYSTEMS
- H04S7/00—Indicating arrangements; Control arrangements, e.g. balance control
- H04S7/30—Control circuits for electronic adaptation of the sound field
- H04S7/307—Frequency adjustment, e.g. tone control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R5/00—Stereophonic arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04S—STEREOPHONIC SYSTEMS
- H04S7/00—Indicating arrangements; Control arrangements, e.g. balance control
- H04S7/30—Control circuits for electronic adaptation of the sound field
- H04S7/302—Electronic adaptation of stereophonic sound system to listener position or orientation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R2499/00—Aspects covered by H04R or H04S not otherwise provided for in their subgroups
- H04R2499/10—General applications
- H04R2499/13—Acoustic transducers and sound field adaptation in vehicles
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04S—STEREOPHONIC SYSTEMS
- H04S2420/00—Techniques used stereophonic systems covered by H04S but not provided for in its groups
- H04S2420/07—Synergistic effects of band splitting and sub-band processing
Abstract
The invention discloses a self-adaptive active frequency division method for vehicle-mounted stereo reproduction. The method comprises the following steps: setting the type of the active frequency division network; acquiring objective parameters of a loudspeaker; according to objective parameters, an active frequency division network suitable for two-frequency division or three-frequency division is generated by combining an active frequency division algorithm; and carrying out feature extraction and classification on the input signal source, and carrying out self-adaptive matching on the frequency division mode by using a classification result. The invention is suitable for the sound reproduction of different program signal sources, realizes the automatic detection of the characteristics of the program sources, and limits the detection types to three types, thereby ensuring the accuracy of the detection result; the invention is suitable for the vehicle stereo active frequency division network of the layout of the automobile frequency division loudspeaker and the characteristics of the program source, and can realize a certain space sound field or virtual sound image effect on the basis of ensuring the tone quality of the reproduced sound signal.
Description
Technical Field
The invention relates to the field of automobile sound reproduction, in particular to a self-adaptive active frequency division method for vehicle-mounted stereo reproduction.
Background
Stereo playback in the acoustic space of a car (car stereo playback system for short) faces three difficulties: (1) the automobile acoustic space has the characteristics of small volume, complex structure and the like, so that the sound field conditions in the automobile are more complex compared with the traditional indoor sound field, and the stereo playback is not facilitated; (2) after sound reproduction design, the problem of phase and amplitude connection exists near frequency dividing points, and particularly, the sound signal radiation near the frequency dividing points is remarkably influenced, so that the reproduction effect (including the influence on tone and timbre) is influenced; (3) separation of low, mid, and high frequency spatial locations (including separation in direction and distance) presents further difficulties in processing the stereo and surround sound signals.
In the case of crossover technology for automotive sound reproduction systems, there are usually two-crossover or three-crossover speaker systems. The key technology of frequency division mainly comprises two aspects, wherein the type of frequency division is active frequency division or passive frequency division; the second is the selection of the frequency division point, i.e. the frequency corresponding to the frequency division point of the second frequency division or the third frequency division.
■ distinction between active and passive frequency division
Passive frequency division means that an audio signal is amplified by a power amplifier circuit and then subjected to frequency division by a passive frequency divider, and then fed to a corresponding high-frequency or low-frequency loudspeaker unit, wherein the principle is that a high-frequency sound signal is filtered by an inductance circuit, and the remaining low-frequency sound signal is fed to the low-frequency loudspeaker unit; the low frequency acoustic signals are filtered out by the electrolytic capacitor and the remaining high frequency acoustic signals are fed to the high frequency speaker unit. In general, the passive frequency division method is adjusted by a variable resistor, so that the frequency division point is relatively blurred and greatly affected by the precision and nonlinearity of the semiconductor device.
The active frequency division means that the frequency division is carried out on the audio signal of the host in a central processing unit of the host before the audio signal is amplified by a power amplifier circuit. The principle is that an audio signal enters a Central Processing Unit (CPU) of a host, the audio signal is divided into high, medium and low frequency signals according to a frequency response range in the CPU of the host, the divided signals are respectively input to a power amplifier to be amplified independently, the power amplifier only plays a role of amplifying the signals, a frequency divider is omitted, and loss in the transmission process is reduced. The frequency division is digital. The frequency division directly by digital signals (electronic frequency dividers) has much less loss than the frequency division of analog signals, and can reproduce good sound quality.
In general, the passive frequency division system is simpler, has higher cost performance and is easily accepted by more vehicle owners; the active frequency division system has high cost and high technical requirements, but has better tone quality effect, is more suitable for replaying and reproducing a space sound field, and has richer tone color details.
■ selection of frequency division points
The selection of the division point is different in different brands and different vehicle types. The frequency of the frequency dividing point is related to the layout mode and objective parameters of the vehicle-mounted loudspeaker on the one hand. For example, the crossover parameters and the like need to be set according to the layout mode and objective parameters of the vehicle-mounted speakers, the crossover network needs to perform signal processing such as time delay and attenuation according to the direction asymmetry, distance difference and the like of different speaker units from the positions of target listeners, and the playback effect needs to be debugged based on multiple measurement results. On the other hand, the frequency of the frequency dividing point should also be related to the signal category, and is designed to be adaptive active frequency dividing for different programs, which is one of the problems to be solved by the present invention. Some common dividing point selections are shown in table 1, and it can be seen that even if the dividing point is divided by three, there is a significant difference in the setting of the dividing point. It should be noted that, from the data listed in table 1, in practice, "quasi-four-division" has been used for the on-board sound reproduction, i.e., one auxiliary bass path regardless of the direction of the sound source, plus three division paths. For uniformity, they are commonly referred to collectively as divide-by-three.
Table 1 lists some of the existing car soundsDivide-by-three approach for playback*
References:
[1]Cecchi S,Palestini L,Peretti P,et al.Advanced Audio Algorithms for a Real Automotive Digital Audio System[C]//AES 125th Convention,San Fransisco,CA,USA.2012.
[2]Mathias Johansson,et al.Sound Field Control Using a Limited Number of Loudspeakers[C]//Audio Engineering Society(AES)36th International Conference,June 2009.2009.
[3]Herbert Hetze,Harald
New Methods of Adaptive Sound Reinforcement in Car Environment[C]//Audio Engineering Society(AES)92nd International Conference，March 1992.1992.
[4]Marcin Kalinowski,Matteo Nucci.Crossover filters pre and post limiters for basic loudspeaker protection:a case study[C]//the Conference on Automotive Audio,Neuburg an der Donau,Germany.2019.
■ the main problem of the existing active frequency division method
(1) Due to the limitation and complexity of the acoustic space of the automobile and the flexibility and variety of the sizes and installation modes of the vehicle-mounted loudspeaker units (the high, middle and low loudspeaker units have large position dispersion degree, and the horizontal position and the vertical position are different), the control of the stereo image is difficult greatly. Therefore, different allocation modes should take account of different characteristics of sound sources: mainly considering sound of sound image, the loudspeaker unit with high elevation angle is mainly fed; emphasizing the rendering of sound of a surround sound field, loudspeaker units with a large radiated sound pressure level should be fed mainly. However, this problem is not currently appreciated.
(2) The frequency division mode in a certain type of vehicle is single, the frequency division point is fixed and not changed, and the difference is not facilitatedSound effect of program source Fruit presentation. For example, the formants influencing tone colors in the phonography usually appear in a frequency band of 2-4kHz, the existing mainstream frequency division points are all between 3-4kHz, and the formants are just destroyed, so that the speech sound of sand is likely to appear; generally, the sound energy of the music signal is basically concentrated in the middle and low frequency range below 1.8-2kHz, and the low frequency division point of 300-400Hz leads to the music signal being fed to the middle and high frequency unit, which is not favorable for the expressive power of the music signal.
(3) Due to the complexity of the automotive sound reproduction system, as well as the positional sensitivity of the sound reproduction system arrangement, the setting of the division point is often a one-time operation. There is therefore little work involved in changing the divide point setting for car playback, even to support multiple divide modes, and to adaptively switch divide modes.
(4) Since stereo localization is mainly medium and low frequency localization, a frequency range below about 1-2kHz can perform effective stereo image control. Therefore, the method of pulling sound image by the high-frequency loudspeaker unit is proposed in the prior art, and the frequency division point above 3-4kHz obviously has certain problems. Therefore, the frequency division problem of car stereo playback remains to be further studied.
Disclosure of Invention
The invention provides a self-adaptive active frequency division method for vehicle-mounted stereo reproduction, aiming at overcoming the defects of the conventional vehicle-mounted stereo reproduction active frequency division technology. The invention can set frequency dividing parameters according to the layout of the vehicle-mounted loudspeaker and the objective parameters of the loudspeaker, and can adaptively switch the active frequency dividing network in three preset frequency dividing modes along with the change of a program source when a stereo signal is played back, so that the frequency dividing network can meet different requirements of comprehensive playback effects of human voice, color, sound image, music scene and the like, thereby optimizing the playback effect. The invention is to select the active frequency division mode according to the characteristics of the program source, thereby ensuring the integrity of the sound presentation process of the played back program source as much as possible, avoiding the distortion of the played back signal caused by improper frequency division, and improving the effect of vehicle-mounted stereo playback on the premise of not losing the sound quality.
The purpose of the invention is realized by at least one of the following technical solutions.
An adaptive active frequency division method for vehicle stereo playback, comprising the steps of:
s1, setting the type of the active frequency division network;
s2, collecting objective parameters of the loudspeaker;
s3, generating an active frequency division network suitable for two-frequency division or three-frequency division according to the objective parameters in the step S2 and by combining an active frequency division algorithm;
and S4, carrying out feature extraction and classification on the input signal source, and carrying out self-adaptive matching on the frequency division mode by using the classification result.
Further, in step S1, the active frequency dividing network is set to three types, which are: the frequency division mode is suitable for pure music programs, the frequency division mode is suitable for singing programs and the frequency division mode is suitable for pure voice programs; the dividing mode of the three types of programs just covers various combinations of music and voice, so that the pure voice type programs can replay a clear human voice sound image in front of a target listening position and close to a horizontal plane where two ears are located, the pure music (background music) type programs can replay a better sound field surrounding sense or surround sound effect, and the replay effect of the singing type programs can give consideration to the human voice sound image and background music sound field effects;
frequency-division mode suitable for pure speech-like programs: human sound images at proper positions (close to the horizontal plane where two ears are positioned and concentrated in front of a target listening position) can be presented in vehicle stereo reproduction;
frequency division mode suitable for pure music-like programs: background music presents better surrounding feeling;
frequency division mode suitable for singing class programs: the vocal sound and image of singing and the surrounding sense of background music are considered, and the vocal formants (between 2 and 4 kHz) of the singing person are not damaged.
Further, in step S2, the objective parameters of the speaker to be collected mainly include: the frequency range under the constraints of the crossover mode and the loudspeaker directivity is as follows:
obtaining the lower limit frequency of the frequency response range of the medium-high frequency loudspeaker unit from the frequency response characteristic curve
Lower limit and upper limit frequency of frequency response range of middle and low frequency loudspeaker unit
And
upper limit frequency of frequency response range of low frequency loudspeaker unit
Further, the lower limit frequency of the frequency response range of the middle-high frequency speaker unit
From a forward angle relative to the mid-high frequency loudspeaker of
The lower limit and the upper limit of the frequency response range of the medium and low frequency speaker unit
And
from a forward angle relative to the mid-low frequency loudspeaker of
Determining the upper limit frequency of the frequency response range of the woofer unit
From a forward angle of inclusion with respect to the woofer
Is determined.
Further, in step S2, according toThe redundancy coefficient gamma needs to be selectedh,γm,γlThe value ranges of the redundancy coefficients are all 0-1, the value ranges are determined by reducing the effective radiation sound pressure level range, the sound pressure level reduction amplitude is not more than 3dB, and then the upper limit of cut-off frequency when the sound pressure level deviates from the forward radiation of the loudspeaker is properly expanded by using the coefficient (1+ the redundancy coefficient).
Further, in step S3, the frequency response is sequentially from high to low for the middle and high frequency speaker unit, the middle and low frequency speaker unit, and the left and right speaker division points are the same, and the parameter settings and corresponding algorithms of the three division modes are as follows:
first frequency dividing point FMLThe expression of (i, j) is:
second frequency-dividing point FHMThe expression of (i, j) is:
in the formula (1) and the formula (2), the sequence number i of the frequency division mode is 1,2 and 3 respectively correspond to a pure voice mode, a pure music mode and a singing mode; the loudspeaker layout mode j is 1,2 and 3 respectively corresponding to the frequency division of two, the frequency division of three A and the frequency division of three B.
Further, the frequency response ranges of the medium-high frequency, medium-low frequency and low-frequency loudspeaker units under the directivity constraint are respectively
And
redundancy coefficient gammah、γmAnd gammalRespectively suitable for medium-high frequency, medium-low frequency and low-frequency loudspeaker units.
Further, the two-frequency division is usually that a middle and high frequency speaker unit is positioned on the A column, and the middle and low frequency speaker unit is positioned on the side door; the three-crossover a generally means that the medium-high and medium-low frequency speaker units are both located on the a pillar, and the low frequency speaker unit is located on the side door; the B-triple-crossover generally means that the mid-high frequency speaker unit is located in the a-pillar, the mid-low frequency speaker unit is located in the side door, and the low frequency speaker unit is located in the floor below the seat.
Further, in step S4, the program sources are divided into: pure voice type programs with relatively more frames containing low short-time energy; pure music type programs with relatively small number of fragments containing higher short-time average zero-crossing rate standard deviation, relatively more higher fundamental tone frequency frame number and relatively higher second sub-band energy ratio; and singing programs with relatively more segments containing higher short-time average zero-crossing rate standard deviation;
preprocessing a program source, measuring and calculating short-time energy, short-time average zero-crossing rate, fundamental tone frequency and sub-band energy ratio of the program source, setting a proper threshold value, and counting to obtain the number of low short-time energy frames, the number of higher short-time average zero-crossing rate standard deviation fragments and the number of higher fundamental tone frequency frames; then, comparing with a preset threshold to obtain a judgment result of the program source signal type, which is as follows:
extracting short-time energy characteristic parameters of the program source, and defining the parameters as the parameters { anN is 1,2, …, n is the number of programs in the program library, anSetting a proper threshold value for the frame number of the nth program which is lower than the short-time energy threshold value to identify a pure voice program; extracting the characteristic parameter of the energy ratio of the second sub-band, and defining the characteristic parameter as a parameter { bn,n＝1,2,…}，bnExtracting the pitch frequency characteristic parameter as the parameter { cn,n＝1,2,…}，cnExtracting the short-time average zero-crossing rate characteristic parameter for the frame number higher than the fundamental tone frequency threshold in the nth program, and defining the parameter as { dn,n＝1,2,…}，dnSetting a proper threshold value for the number of audio segments higher than the short-time average zero-crossing rate standard deviation threshold value in the nth program, and distinguishing and identifying the singing type program and the pure music type program;
then, according to the type of the extracted program, a frequency division mode of the active frequency division network is set, and self-adaptive matching is completed.
Further, when the adaptive active frequency dividing function cannot be judged or turned off, the type of the default program source signal is the singing type program.
Compared with the prior art, the invention has the advantages and beneficial effects that:
1. the parameters of the frequency division network are set according to the layout mode of the loudspeakers in the vehicle and the objective parameters of the loudspeakers, so that the frequency division network can meet different requirements for emphasizing human sound and image and background music sound fields, and the comprehensive playback effects of human sound and image, music scenes and the like are presented.
2. The method is suitable for sound reproduction of different program signal sources, automatic detection of the program source characteristics is realized, and the detection types are limited to three types, so that the accuracy of the detection result is ensured;
3. the vehicle stereo active frequency division network suitable for the layout of the automobile frequency division loudspeaker and the characteristics of the program source can realize a certain space sound field or virtual sound image effect on the basis of ensuring the tone quality of the reproduced sound signal.
Drawings
FIG. 1 is a schematic diagram of an adaptive active frequency division method for car stereo playback according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an algorithm design of an adaptive active frequency division network according to an embodiment of the present invention;
fig. 3a is a schematic diagram of a frequency division layout corresponding to j ═ 1 in an embodiment of the present invention;
fig. 3b is a schematic layout of a divide-by-three a corresponding to j-2 in an embodiment of the present invention;
fig. 3c is a schematic layout of a divide-by-three B topology corresponding to j-3 in an embodiment of the present invention;
fig. 4 is a schematic diagram of the directivity characteristics of the speaker unit according to the embodiment of the present invention at radiation angles of 0 ° and 30 °;
FIG. 5 is a flowchart of a program source feature extraction and classification algorithm in an embodiment of the present invention;
FIG. 6 is a scatter diagram of audio recognition results according to an embodiment of the present invention.
Detailed Description
The present invention will be further described with reference to the following examples and drawings, but the embodiments of the present invention are not limited thereto.
Example (b):
an adaptive active frequency division method for car stereo playback, as shown in fig. 1, includes the following steps:
s1, setting the type of the active frequency division network;
as shown in fig. 2, the active frequency-dividing network is set to three types, which are: the frequency division mode is suitable for pure music programs, the frequency division mode is suitable for singing programs and the frequency division mode is suitable for pure voice programs, the division mode of the three types of programs can just cover various combinations of music and voice, so that the pure voice programs can play back clear human sound images in front of a target listening position and close to the horizontal plane where two ears are located, the pure music (background music) programs can play back better sound field surrounding sense or surround sound effect, and the playback effect of the singing programs can give consideration to both the human sound images and the background music sound field effect;
in this embodiment, the design examples of each frequency division mode are as follows:
the frequency division mode suitable for pure music programs basically does not consider the influence of frequency division on sound images, and mainly considers the timbre and surround effect of music. The frequency division mode of the program emphasizes the creation of surround feeling, and focuses the reproduced sound energy on the mid-bass loudspeaker unit so as to widen the sound field range and highlight the surround effect of pure music. When creating the sense of envelopment, it can be realized by signal processing methods such as inter-path decorrelation.
The frequency division mode suitable for singing programs can effectively concentrate the voice of a singer in the front of the singer on the premise of not damaging the vowel formant, and the perception elevation angle can be close to the horizontal plane;
for crossover modes suitable for pure speech-like programs, which primarily emphasize reproduction from hundreds of Hz to about 4kHz, the lower frequency limit of the mid-high frequency speaker units needs to be lowered appropriately in order to get the human voice as close as possible to the height of the listener's horizontal plane. Furthermore, such programs should focus on the complete rendering of the speech spectrum and virtual sound image control, regardless of the low frequency signal and surround effects. The specific numerical value of the frequency dividing point needs to be optimized according to the specific vehicle type and the loudspeaker unit.
S2, as shown in fig. 2 and 3, the collecting objective parameters of the speaker mainly includes: the frequency range under the constraints of the crossover mode and the loudspeaker directivity is as follows:
obtaining the lower limit frequency of the frequency response range of the medium-high frequency loudspeaker unit from the frequency response characteristic curve
Lower limit and upper limit frequency of frequency response range of middle and low frequency loudspeaker unit
And
upper limit frequency of frequency response range of low frequency loudspeaker unit
Lower limit frequency of frequency response range of medium-high frequency loudspeaker unit
From a forward angle relative to the mid-high frequency loudspeaker of
The lower limit and the upper limit of the frequency response range of the medium and low frequency speaker unit
And
from a forward angle relative to the mid-low frequency loudspeaker of
Determining the upper limit frequency of the frequency response range of the woofer unit
From a forward angle of inclusion with respect to the woofer
Is determined.
Selecting redundancy coefficient gamma according to requirementh,γm,γlThe value ranges of the redundancy coefficients are all 0-1, the value ranges are determined by reducing the effective radiation sound pressure level range, the sound pressure level reduction amplitude is not more than 3dB, as shown in figure 4, and then the upper limit of cut-off frequency deviating from the forward radiation of the loudspeaker is properly expanded by using the coefficient (1+ the redundancy coefficient).
In the embodiment, two typical modes of mounting the vehicle-mounted loudspeaker are taken as examples, wherein one mode is that a middle-high unit is mounted on an A column, and a low-sound unit is mounted on the lower side of a vehicle door; the other is that a high-middle high-frequency loudspeaker unit is arranged on the A column, a middle-low frequency loudspeaker unit is arranged on the side door, and a low-frequency loudspeaker unit is arranged at the bottom of the seat. In order to obtain a good frequency-division reproduction effect, the target region is set to the driving position, i.e., the measurement point of the objective parameter is near the center position of the driver's head. Respectively measuring the frequency response right in front of each loudspeaker unit and the frequency response near the center of the target listening position head, thereby calculating the frequency response ranges of the medium-high frequency loudspeaker unit, the medium-low frequency loudspeaker unit and the low-frequency loudspeaker unit under the directivity constraint
And
redundancy coefficient gamma under constraint condition of attenuation 3dBh、γmAnd gammalRespectively suitable for medium-high frequency, medium-low frequency and low-frequency loudspeaker units.
S3, generating an active frequency division network suitable for two-frequency division or three-frequency division according to the objective parameters in the step S2 and by combining the active frequency division algorithm shown in the figure 2;
the frequency response is sequentially a medium-high frequency loudspeaker unit, a medium-low frequency loudspeaker unit and a low-frequency loudspeaker unit from high to low, the frequency dividing points of the left loudspeaker and the right loudspeaker are the same, and the parameter settings of the three frequency dividing modes and corresponding algorithms are as follows:
first frequency dividing point FMLThe expression of (i, j) is:
second frequency-dividing point FHMThe expression of (i, j) is:
in the formula (1) and the formula (2), the sequence number i of the frequency division mode is 1,2 and 3 respectively correspond to a pure voice mode, a pure music mode and a singing mode; the loudspeaker layout mode j is 1,2 and 3 respectively corresponding to the frequency division of two, the frequency division of three A and the frequency division of three B.
As shown in fig. 3a, the divide-by-two is usually a middle and high frequency speaker unit located in the a pillar, and a middle and low frequency speaker unit located in the side door; as shown in fig. 3b, crossover a generally means that both the mid and mid high frequency and low frequency speaker units are located in the a pillar, with the low frequency speaker unit located in the side door; as shown in fig. 3c, crossover B generally means that the mid-high frequency speaker unit is located in the a-pillar, the mid-low frequency speaker unit is located in the side door, and the low frequency speaker unit is located in the floor below the seat.
S4, performing feature extraction and classification on the input signal source, and performing self-adaptive matching on the frequency division mode by using a classification result;
preprocessing a program source, measuring and calculating short-time energy, short-time average zero-crossing rate, fundamental tone frequency and sub-band energy ratio of the program source, setting a proper threshold value, and counting to obtain the number of low short-time energy frames, the number of higher short-time average zero-crossing rate standard deviation fragments and the number of higher fundamental tone frequency frames; then, comparing with a preset threshold to obtain a judgment result of the program source signal type, which is as follows:
extracting short-time energy characteristic parameters of the program source, and defining the parameters as the parameters { anN is 1,2, … }, n isNumber of programs in the program library, anSetting a proper threshold value for the frame number of the nth program which is lower than the short-time energy threshold value to identify a pure voice program; extracting the characteristic parameter of the energy ratio of the second sub-band, and defining the characteristic parameter as a parameter { bn,n＝1,2,…}，bnExtracting the pitch frequency characteristic parameter as the parameter { cn,n＝1,2,…}，cnExtracting the short-time average zero-crossing rate characteristic parameter for the frame number higher than the fundamental tone frequency threshold in the nth program, and defining the parameter as { dn,n＝1,2,…}，dnSetting a proper threshold value for the number of audio segments higher than the short-time average zero-crossing rate standard deviation threshold value in the nth program, and distinguishing and identifying the singing type program and the pure music type program;
then, according to the type of the extracted program, a frequency division mode of the active frequency division network is set, and self-adaptive matching is completed.
In this embodiment, as shown in fig. 5, first, an audio of a certain signal duration is intercepted from each song, the audio of the duration is subjected to framing processing, a short-time energy characteristic parameter is extracted, then a suitable short-time energy threshold is set, the number of frames in the audio of the segment that are lower than the short-time energy threshold is counted, a suitable frame number threshold is set, the audio that is greater than the frame number threshold is identified as a pure-voice program, and if the audio is lower than the frame number threshold, the next stage is entered; intercepting the audio with a certain signal duration, performing fast Fourier transform on the audio with the duration, calculating a second sub-band energy ratio of the audio, setting a proper sub-band energy ratio threshold value, identifying the audio higher than the threshold value as a pure music program, and entering a next characteristic parameter judgment stage if the audio is lower than the frame number threshold value; intercepting audio with a certain signal duration, performing endpoint detection on the audio, extracting fundamental tone frequency in the audio, setting a proper fundamental tone frequency threshold, counting the number of frames in the audio, which are higher than the fundamental tone frequency threshold, in the section of audio, setting a proper frame number threshold, identifying the audio, which is higher than the frame number threshold, as a pure music program, and entering a next characteristic parameter judgment stage if the audio is lower than the frame number threshold; intercepting an audio frequency with a certain signal duration, carrying out segmentation processing on the audio frequency with the duration, framing the segmented audio frequency, extracting short-time average zero-crossing rate characteristic parameters, calculating the short-time average zero-crossing rate standard deviation of an audio frequency segment, setting a proper short-time average zero-crossing rate standard deviation threshold value, counting the number of audio frequency segments in the audio frequency with the duration, which are higher than the short-time average zero-crossing rate standard deviation threshold value, setting a proper segment number threshold value, identifying the audio frequency which is higher than the segment number threshold value as a singing program, and identifying the audio frequency which is lower than the segment number threshold value as a pure music program. Taking 60 songs as an example, 20 pure voice type programs, singing type programs and pure music type programs respectively, and fig. 6 is a program source classification result scatter diagram. It can be seen that in the audio recognition process, the correct recognition rate is 98.33%, which is generally considered to meet the requirements of practical application.
In this embodiment, as an example of a certain type of vehicle, the layout of the speakers in the vehicle is a front row three-frequency division mode, the high and middle speaker units are installed on the a pillar, the low speaker unit is installed on the vehicle door, the rear row is two-frequency division mode, the middle and low speaker units are both located at the vehicle door, and the ultra-low speaker is located in the tail compartment of the vehicle. Firstly, obtaining frequency response curves of all loudspeakers, and selecting a proper frequency dividing point region according to the frequency response and the loudspeaker layout mode mentioned in the example; meanwhile, according to the characteristics of different types of signal sources, under the condition that the frequency response of the working frequency band of each loudspeaker is kept as flat as possible, a frequency dividing point is selected in a preset frequency dividing point region, percussion instruments such as drums in pure music can generate shocking sound effects, meanwhile, signal energy generated by various instruments is mainly concentrated in a low frequency band, a specific frequency dividing point can be obtained through a formula in the step 3, after fine adjustment is carried out, the frequency dividing point of three frequency divisions can be set to be 100Hz, 1000 Hz and 3000Hz, frequency band signals below 100Hz are replayed by a super-low-pitch loudspeaker, and the percussion feeling of the low-frequency instruments in the pure music and the surrounding feeling of the pure music are fully presented; regarding singing type programs, a user mainly focuses on the vocal effect of a singing song, the vocal energy is mainly distributed in 300-4000 Hz, meanwhile, the singing type song contains pure music, according to a formula, frequency dividing points of three frequency division are set to be 100, 600 and 4000Hz, and the surround feeling of vocal tone, sound image and pure music is considered; for pure voice programs, users mainly pay attention to sound image distribution of human voice and language programs, the working range of middle-high frequency and middle-low frequency loudspeakers can be enlarged to improve the virtual sound image played by the loudspeakers by combining a formula, and the frequency dividing points of three frequency divisions are set to be 100, 300 and 4000 Hz; the default type is set as singing type program, and the frequency dividing points of the three frequency divisions are set as: 100. 300 and 4000 Hz. Each crossover speaker uses a FIR filter designed with a higher order window function to guarantee a linear phase. Then, using MLS signal as measuring signal, using miniature microphone to measure single-point impulse response at head double-ear center position of driver in car, then using cross-correlation principle to calculate delay point number, at the same time amplitude response can be obtained by using impulse response after MLS deconvolution to obtain delay network parameter, the delay network mainly makes regulation to delay and amplitude of every channel audio signal.
It should be noted that the parameters of the active frequency division network not only include frequency division points, types and orders of filters, but also include the number of delay points and magnitude terms of each speaker in the delay network. The selection of specific delay points and the adjustment of the amplitude should be set and verified in combination with subjective evaluation. However, for a specific vehicle model, respective frequency division modes can be preset and optimally debugged for different types of signal sources, so as to ensure a better sound reproduction effect.
It can be seen from the above example that a car stereo active crossover network system adapted to car crossover speaker layout and program source characteristics can be adapted to stereo playback from different signal sources. The system can adjust the parameters of the frequency division network according to the layout mode and the objective parameters of the vehicle-mounted loudspeaker, adaptively selects the frequency division mode to match different signal sources, and improves the playback effect of the vehicle-mounted stereo without losing tone quality. Compared with the traditional single-mode active/passive frequency division, the method improves the playback effect and optimizes the hearing experience in the vehicle.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (10)
1. An adaptive active frequency division method for vehicle stereo playback, comprising the steps of:
s1, setting the type of the active frequency division network;
s2, collecting objective parameters of the loudspeaker;
s3, generating an active frequency division network suitable for two-frequency division or three-frequency division according to the objective parameters in the step S2 and by combining an active frequency division algorithm;
and S4, carrying out feature extraction and classification on the input signal source, and carrying out self-adaptive matching on the frequency division mode by using the classification result.
2. The adaptive active frequency dividing method for car stereo reproduction according to claim 1, wherein in step S1, the active frequency dividing network is set to three types, which are: the frequency division mode is suitable for pure music type programs, the frequency division mode is suitable for singing type programs, and the frequency division mode is suitable for pure voice type programs.
3. The adaptive active frequency dividing method for car stereo reproduction according to claim 1, wherein in step S2, the objective parameters of the speakers to be collected mainly include: the frequency range under the constraints of the crossover mode and the loudspeaker directivity is as follows:
obtaining the lower limit frequency of the frequency response range of the medium-high frequency loudspeaker unit from the frequency response characteristic curve
Lower limit and upper limit frequency of frequency response range of middle and low frequency loudspeaker unit
And
upper limit frequency of frequency response range of low frequency loudspeaker unit
4. The adaptive active crossover method of claim 3, wherein the lower frequency limit of the mid-high frequency speaker unit frequency response range
From a forward angle relative to the mid-high frequency loudspeaker of
The lower limit and the upper limit of the frequency response range of the medium and low frequency speaker unit
And
from a forward angle relative to the mid-low frequency loudspeaker of
Determining the upper limit frequency of the frequency response range of the woofer unit
From a forward angle of inclusion with respect to the woofer
Is determined.
5. The adaptive active frequency division method for car stereo reproduction according to claim 3, wherein in step S2, the redundancy coefficient γ is selected as requiredh,γm,γlThe value ranges of the redundancy coefficients are all 0-1, the value ranges are determined by reducing the effective radiation sound pressure level range, the sound pressure level reduction amplitude is not more than 3dB, and then the upper limit of cut-off frequency when the radiation deviates from the forward direction of the loudspeaker is expanded by using the coefficient (1+ the redundancy coefficient).
6. The adaptive active crossover method for car stereo reproduction of claim 4, wherein in step S3, the frequency response is sequentially from high to low for the mid-high frequency speaker unit, the mid-low frequency speaker unit and the low frequency speaker unit, the left and right speaker crossover points are the same, and the parameter settings and corresponding algorithms for the three crossover modes are as follows:
first frequency dividing point FMLThe expression of (i, j) is:
second frequency-dividing point FHMThe expression of (i, j) is:
in the formula (1) and the formula (2), the sequence number i of the frequency division mode is 1,2 and 3 respectively correspond to a pure voice mode, a pure music mode and a singing mode; the loudspeaker layout mode j is 1,2 and 3 respectively corresponding to the frequency division of two, the frequency division of three A and the frequency division of three B.
7. The adaptive active crossover method of claim 6, wherein the mid-high, mid-low and low-frequency speaker units have frequency response ranges under directivity constraints of respectively
And
redundancy coefficient gammah、γmAnd gammalRespectively suitable for medium-high frequency, medium-low frequency and low-frequency loudspeaker units.
8. The adaptive active crossover method of car stereo reproduction of claim 6, wherein the divide-by-two is a mid-high frequency speaker unit located in the a-pillar and a mid-low frequency speaker unit located in the side door; the three-frequency division A means that the middle-high frequency and middle-low frequency loudspeaker units are positioned on the column A, and the low frequency loudspeaker unit is positioned on the side door; the three-frequency division B means that the middle and high frequency loudspeaker unit is positioned on the column A, the middle and low frequency loudspeaker unit is positioned on the side door, and the low frequency loudspeaker unit is positioned on the bottom plate below the seat.
9. The adaptive active frequency dividing method for car stereo reproduction according to claim 1, wherein in step S4, the program source is preprocessed to measure the short-time energy, the short-time average zero-crossing rate, the pitch frequency and the sub-band energy ratio of the program source, set the appropriate threshold, and statistically obtain the low short-time energy frame number, the higher short-time average zero-crossing rate standard deviation segment number, and the higher pitch frequency frame number; then, comparing with a preset threshold to obtain a judgment result of the program source signal type, which is as follows:
extracting short-time energy characteristic parameters of the program source, and defining the parameters as the parameters { anN is 1,2, …, n is the number of programs in the program library, anSetting a proper threshold value for the frame number of the nth program which is lower than the short-time energy threshold value to identify a pure voice program; extracting the characteristic parameter of the energy ratio of the second sub-band, and defining the characteristic parameter as a parameter { bn,n＝1,2,…}，bnExtracting the pitch frequency characteristic parameter as the parameter { cn,n＝1,2,…}，cnExtracting the short-time average zero-crossing rate characteristic parameter for the frame number higher than the fundamental tone frequency threshold in the nth program, and defining the parameter as { dn,n＝1,2,…}，dnSetting the number of audio frequency segments in the nth program which are higher than the short-time average zero-crossing rate standard deviation threshold valueDistinguishing and identifying the singing type programs and the pure music type programs by proper threshold values;
then, according to the type of the extracted program, a frequency division mode of the active frequency division network is set, and self-adaptive matching is completed.
10. The adaptive active frequency dividing method for car stereo reproduction according to claim 9, wherein the default program source signal category is singing type program when the adaptive active frequency dividing function cannot be determined or turned off.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011230703.5A CN112511966B (en) | 2020-11-06 | 2020-11-06 | Self-adaptive active frequency division method for vehicle-mounted stereo playback |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011230703.5A CN112511966B (en) | 2020-11-06 | 2020-11-06 | Self-adaptive active frequency division method for vehicle-mounted stereo playback |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112511966A true CN112511966A (en) | 2021-03-16 |
| CN112511966B CN112511966B (en) | 2021-12-17 |
Family
ID=74955411
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011230703.5A Active CN112511966B (en) | 2020-11-06 | 2020-11-06 | Self-adaptive active frequency division method for vehicle-mounted stereo playback |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112511966B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114979932A (en) * | 2022-04-20 | 2022-08-30 | 中国第一汽车股份有限公司 | Method and system for approximately realizing audition of reverberant sound field of concert hall in vehicle-mounted sound reproduction |
| CN116437268A (en) * | 2023-06-14 | 2023-07-14 | 武汉海微科技有限公司 | Adaptive frequency division surround sound upmixing method, device, equipment and storage medium |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4031321A (en) * | 1973-11-06 | 1977-06-21 | Bang & Olufsen A/S | Loudspeaker systems |
| JPH10290378A (en) * | 1997-04-16 | 1998-10-27 | Sony Corp | Clock generator |
| US6148243A (en) * | 1996-04-05 | 2000-11-14 | Canon Kabushiki Kaisha | Sound Processing method and system |
| CN102523542A (en) * | 2011-12-14 | 2012-06-27 | 南京琅声声学科技有限公司 | Quasi adaptive acoustic equalization method and apparatus thereof |
| CN202841493U (en) * | 2012-09-27 | 2013-03-27 | 郑全录 | Electronic frequency-halving audio signal processing circuit |
| CN104469603A (en) * | 2013-09-25 | 2015-03-25 | 联想(北京)有限公司 | Electronic device, frequency dividing method and frequency dividing device |
| CN106303820A (en) * | 2015-06-10 | 2017-01-04 | 联想(北京)有限公司 | A kind of frequency division of the frequency method and system |
| CN107064626A (en) * | 2016-12-28 | 2017-08-18 | 安徽问天量子科技股份有限公司 | The frequency calculation method and frequency meter of adaptive wide frequency ranges |
| CN107396274A (en) * | 2017-07-07 | 2017-11-24 | 广州飞达音响股份有限公司 | The method, apparatus and system of active linear array sound equipment sound field adjustment |
| US9955261B2 (en) * | 2016-01-13 | 2018-04-24 | Vlsi Solution Oy | Method and apparatus for adjusting a cross-over frequency of a loudspeaker |
| CN110851109A (en) * | 2019-05-15 | 2020-02-28 | 音王电声股份有限公司 | Sound quality processor based on room impulse response measurement |
-
2020
- 2020-11-06 CN CN202011230703.5A patent/CN112511966B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4031321A (en) * | 1973-11-06 | 1977-06-21 | Bang & Olufsen A/S | Loudspeaker systems |
| US6148243A (en) * | 1996-04-05 | 2000-11-14 | Canon Kabushiki Kaisha | Sound Processing method and system |
| JPH10290378A (en) * | 1997-04-16 | 1998-10-27 | Sony Corp | Clock generator |
| CN102523542A (en) * | 2011-12-14 | 2012-06-27 | 南京琅声声学科技有限公司 | Quasi adaptive acoustic equalization method and apparatus thereof |
| CN202841493U (en) * | 2012-09-27 | 2013-03-27 | 郑全录 | Electronic frequency-halving audio signal processing circuit |
| CN104469603A (en) * | 2013-09-25 | 2015-03-25 | 联想(北京)有限公司 | Electronic device, frequency dividing method and frequency dividing device |
| CN106303820A (en) * | 2015-06-10 | 2017-01-04 | 联想(北京)有限公司 | A kind of frequency division of the frequency method and system |
| US9955261B2 (en) * | 2016-01-13 | 2018-04-24 | Vlsi Solution Oy | Method and apparatus for adjusting a cross-over frequency of a loudspeaker |
| CN107064626A (en) * | 2016-12-28 | 2017-08-18 | 安徽问天量子科技股份有限公司 | The frequency calculation method and frequency meter of adaptive wide frequency ranges |
| CN107396274A (en) * | 2017-07-07 | 2017-11-24 | 广州飞达音响股份有限公司 | The method, apparatus and system of active linear array sound equipment sound field adjustment |
| CN110851109A (en) * | 2019-05-15 | 2020-02-28 | 音王电声股份有限公司 | Sound quality processor based on room impulse response measurement |
Non-Patent Citations (4)
| Title |
|---|
| LINDA LIANG: "Auditory distance perception cues in small enclosed spaces: numerical simulation and psychoacoustical evaluation", 《THE 49TH INTERNATIONAL CONGRESS AND EXPOSITION ON NOISE CONTROL 》 * |
| SUNIL BHARITKAR: "Optimization of crossover frequency and crossover region response for multichannel acoustic applications", 《2005 13TH EUROPEAN SIGNAL PROCESSING CONFERENCE》 * |
| 余光正: "近场头相关传输函数的多声源快速测量系统设计与验证", 《声学学报》 * |
| 高亮: "自适应声场均衡控制器的设计", 《中国优秀硕士学位论文全文数据库 信息科技辑 》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114979932A (en) * | 2022-04-20 | 2022-08-30 | 中国第一汽车股份有限公司 | Method and system for approximately realizing audition of reverberant sound field of concert hall in vehicle-mounted sound reproduction |
| CN116437268A (en) * | 2023-06-14 | 2023-07-14 | 武汉海微科技有限公司 | Adaptive frequency division surround sound upmixing method, device, equipment and storage medium |
| CN116437268B (en) * | 2023-06-14 | 2023-08-25 | 武汉海微科技有限公司 | Adaptive frequency division surround sound upmixing method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112511966B (en) | 2021-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2486737B1 (en) | System for spatial extraction of audio signals | |
| US9264834B2 (en) | System for modifying an acoustic space with audio source content | |
| JP5149968B2 (en) | Apparatus and method for generating a multi-channel signal including speech signal processing | |
| US8160282B2 (en) | Sound system equalization | |
| EP0476790B1 (en) | Stereo enhancement system | |
| CN112511966B (en) | Self-adaptive active frequency division method for vehicle-mounted stereo playback | |
| JP4844622B2 (en) | Volume correction apparatus, volume correction method, volume correction program, electronic device, and audio apparatus | |
| CN102972047A (en) | Method and apparatus for reproducing stereophonic sound | |
| US10645513B2 (en) | Stereophonic sound reproduction method and apparatus | |
| CN101120412A (en) | A system for and a method of mixing first audio data with second audio data, a program element and a computer-readable medium | |
| JPWO2009144781A1 (en) | Audio playback device | |
| CN111556425A (en) | Tone equalization method for virtual sound reproduction of loudspeaker | |
| CN110789478B (en) | Vehicle-mounted sound parameter auxiliary adjusting method and device and audio processor | |
| CN113645531A (en) | Earphone virtual space sound playback method and device, storage medium and earphone | |
| US9240208B2 (en) | Recording apparatus with mastering function | |
| US8086448B1 (en) | Dynamic modification of a high-order perceptual attribute of an audio signal | |
| KR101745019B1 (en) | Audio system and method for controlling the same | |
| KR102443055B1 (en) | Method and apparatus for 3D sound reproducing | |
| KR102217832B1 (en) | Method and apparatus for 3D sound reproducing using active downmix | |
| US20220329957A1 (en) | Audio signal processing method and audio signal processing apparatus | |
| Liang et al. | Improvement of Sound Quality in Car Based on the Third-Party Sound Effect | |
| JP2023012347A (en) | Acoustic device and acoustic control method | |
| NAKAMICHI et al. | Development of Premium Double Tweeter Sound System for Dealer Option | |
| KR20210034564A (en) | Method and apparatus for 3D sound reproducing | |
| DK201300471A1 (en) | System for dynamically modifying car audio system tuning parameters |
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 |