CN102148033B - Method for testing intelligibility of speech transmission index - Google Patents

Method for testing intelligibility of speech transmission index Download PDF

Info

Publication number
CN102148033B
CN102148033B CN2011100825142A CN201110082514A CN102148033B CN 102148033 B CN102148033 B CN 102148033B CN 2011100825142 A CN2011100825142 A CN 2011100825142A CN 201110082514 A CN201110082514 A CN 201110082514A CN 102148033 B CN102148033 B CN 102148033B
Authority
CN
China
Prior art keywords
signal
transmission index
sti
octave
value
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.)
Expired - Fee Related
Application number
CN2011100825142A
Other languages
Chinese (zh)
Other versions
CN102148033A (en
Inventor
彭健新
吴东海
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN2011100825142A priority Critical patent/CN102148033B/en
Publication of CN102148033A publication Critical patent/CN102148033A/en
Application granted granted Critical
Publication of CN102148033B publication Critical patent/CN102148033B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

The invention relates to a method for testing the intelligibility of a speed transmission index. The method comprises the following steps: receiving a test signal transmitted by the tested speed transmission index at a position to be tested by adopting a microphone; transmitting the received test signal to a signal processing unit by the microphone; calculating a modulation transfer function (MTF) in an octave frequency band of 250-4000Hz by the signal processing unit; calculating a speech transmission index (STI) value according to the MTF; and judging the intelligibility of the speech transmission index according to the STI value. By adopting the method provided by the invention, five wider octave frequency bands (250-4000Hz) are contributed to the intelligibility of Chinese speech for weighting only, thus evaluation on the intelligibility of Chinese speech of a Chinese speech transmission index is ensured to be accurate, and the test speed is improved.

Description

A kind of testing intelligibility of speech transmission index
Technical field
The present invention relates to the fields such as communication, voice alarm system, specifically refer to the method for testing of speech articulation in a kind of speech communication transmission system.
Background technology
For the speech transmission index lamprophonia whether, general language transmission index (STI) value that adopts IEC 60268-16 (2003) its transmission channel of canonical measure, be defined as STI as China's standard GB/T/T 16851-1997 to emergent sound system and be more than or equal to 0.65.The STI that measures Public Address system is always very consuming time.A complete measurement, need to obtain and analyze 98 groups of modulation transfer functions (MTF).For Public Address system, in order to save Measuring Time, IEC 60268-16 (2003) adopts STIPA to carry out objective evaluation.But the measurement of STI-PA is very high to the requirement of ground unrest, can not contain pulse characteristics in ground unrest, its test macro can not be introduced stronger nonlinear distortion.And (all parts of identical sound pressure level and system all are activated) should be in case of emergency measured in the STI-PA of voice alarm system measurement.
Summary of the invention
In order to overcome the prior art above shortcomings, a kind of testing intelligibility of speech transmission index of the present invention, concrete technical scheme is as follows.
A kind of testing intelligibility of speech transmission index, it adopts microphone to receive the test signal through tested speech transmission index transmission in position to be measured, the test signal that microphone will receive is transferred to signal processing unit, signal processing unit is by calculating the modulation transfer function of 250Hz to 4000Hz octave band, according to modulation transfer function computational language transmission index STI value, according to the sharpness of language transmission index STI value judgement speech transmission index.
In above-mentioned method of testing, described test signal is envelope modulated signal, and described signal processing unit carries out described calculating to the envelope modulated signal that receives after by octave band filtering, envelope detected, low-pass filtering again.
In above-mentioned method of testing, described processing unit calculates the modulation transfer function of five octave bands of 250Hz to 4000Hz, then according to modulation transfer function computational language transmission index STI value.
In above-mentioned method of testing, the carrier wave of test signal is the octave narrow band noise by 250,500,1000,2000, centered by 4000Hz; Between each narrow band noise, leave isolation strip, the envelope modulated signal of test signal is the simple harmonic wave of extremely low frequency, and frequency is from 0.63Hz to 12.5Hz, and every 1/3oct gets 1; Its frequency is respectively 0.63Hz, 0.8Hz, 1.0Hz, 1.25Hz, 1.6Hz, 2.0Hz, 2.5Hz, 3.15Hz, 4.0Hz, 5.0Hz, 6.3Hz, 8.0Hz, 10.0Hz, 12.5Hz, totally 14 modulating frequencies
In above-mentioned method of testing, signal processing unit carries out described calculating and comprises:
Obtain 5 MTF curves of octave separation in 250~4000Hz frequency range, every curve carries out spectrum analysis in the 0.63-12.5Hz modulation frequency range with 1/3 octave band; By the m that measures 5 * 14 data that obtain K, fMatrix can be converted to respectively 70 apparent signal to noise ratio snr K, fNumerical value, k are that octave band, f are modulating frequency,
SNR k , f = 10 lg m k , f 1 - m k , f dB - - - ( 1 )
Then apparent signal to noise ratio snr K, fBe limited in ± ask transmission index TI in the 15dB scope K, f, and make 0≤TI K, f≤ 1,
TI k , f = SNR k , f + 15 30 - - - ( 2 )
Transmission index TI by same octave band K, fOn average to obtain modulation transmissions index M TI k:
MTI k = 1 14 Σ f = 1 14 TI k , f - - - ( 3 )
Computational language transmission index STI:
STI = Σ n = 1 5 ω n MTI n - - - ( 4 )
ω nAs shown in table 1.
In above-mentioned method of testing, can also adopt MLS (maximal-length sequence) signal or swept-frequency signal to measure the impulse response of the described transmission channel of test, the m of 5 * 14 data of snr computation when then application of formula (5) is according to impulse response and test K, fMatrix, calculate the STI value according to formula (1)~(4),
m ( f ) = | ∫ 0 ∞ e - j 2 πft p 2 ( t ) dt | ∫ 0 ∞ p 2 ( t ) dt · ( 1 + 10 ( - S / N ) / 10 ) - 1 - - - ( 5 ) .
In above-mentioned method of testing, adopt acoustic software (as Dirac, Winmls etc.) to measure room impulse response, and application software calculate the MTI of 250~4000Hz octave band kValue, then calculate the STI value according to (4) formula.
Compared with prior art, the present invention has following advantage and technique effect: the STI value that the present invention adopts the weighted value (GBT 15485-1995) of Chinese language articulation index to calculate the Chinese language transmission system is estimated its speech articulation, it is only to (250~4000Hz) carry out weighted to 5 larger octave bands of Chinese language sharpness contribution, to guarantee the accuracy that the Chinese language sharpness of Chinese language transmission system is estimated, improved test speed.
The accompanying drawing explanation
Fig. 1 is the test macro block diagram.
Fig. 2 is STI measuring principle block diagram.
Fig. 3 is for measuring the MTF block diagram.
Fig. 4 is the test signal figure after envelope modulation.
Fig. 5 is the impulse response figure measured.
Graph of a relation between the STI value that Fig. 6 obtains for kind of different weighted methods.
Fig. 7 a~Fig. 7 c is respectively STI value that 3 kinds of different weighted methods adopting IEC60268-16 (1988) standard, IEC60268-16 (2003) standard and the inventive method obtain and the match graph of a relation between subjective Chinese articulation score.
Embodiment
Below in conjunction with accompanying drawing, specific embodiment of the invention is described further, but enforcement of the present invention and protection domain are not limited to this.
As Fig. 1, the audio-frequency test signal (modulation signal or random signal) relevant with measuring STI passes through speaker playback, the test signal that adopts receiving element (microphone) to receive in position to be measured after tested speech transmission index transmission, then by signal processing unit (can adopt dsp chip to realize), calculate the modulation transfer function MTF of 250~4000Hz octave band, calculate the STI value according to MTF.Realize that block diagram as shown in Figure 2.
The method of testing detailed step is as follows:
1. measuring system should be configured by Fig. 2, and wherein the distance of measuring sound source and speaker-microphone is 0.5m.
2. the carrier wave of measuring-signal is the octave narrow band noise by 250,500,1000,2000, centered by 4000Hz, and its intensity with the language spectrum quite; Between each narrow band noise, reserve isolation strip, its envelope modulated signal is the simple harmonic wave of extremely low frequency, and its frequency is from 0.63Hz to 12.5Hz, and every 1/3oct chooses 1 modulating frequency, totally 14.
3. obtain 5 MTF curves of octave separation in 250~4000Hz frequency range.Measure the block diagram of MTF as shown in Figure 3.Every curve carries out spectrum analysis in the 0.63-12.5Hz modulation frequency range with 1/3 octave band.By the m that measures 5 * 14 data that obtain K, fMatrix can be converted to respectively 70 apparent signal to noise ratio snr K, f(k is that octave band, f are modulating frequency) numerical value.
SNR k , f = 101 g m k , f 1 - m k , f dB - - - ( 1 )
4. then apparent signal to noise ratio snr K, fBe limited in ± ask transmission index TI in the 15dB scope K, f, and make 0≤TI K, f≤ 1.
TI k , f = SNR k , f + 15 30 - - - ( 2 )
5. by the transmission index TI of same octave band K, fOn average to obtain modulation transmissions index M TI k:
MTI k = 1 14 Σ f = 1 14 TI k , f - - - ( 3 )
6. computational language transmission index STI:
STI = Σ n = 1 5 ω n MTI n - - - ( 4 )
ω nBe worth as shown in table 1.
The different STI computing method of table 1 weighted value ω n
Figure BDA0000053582530000035
7. this test macro also can adopt the impulse response that MLS signal or swept-frequency signal are measured the test transmission channel, the m of 5 * 14 data of snr computation when then application of formula (5) is according to impulse response and test K, fMatrix, calculate the STI value according to formula (1)~(4).This process can adopt existing commercial acoustic Survey Software such as Dirac, Winmls 2004 etc. to measure the MTI of 250~4000Hz octave band kValue, then calculate the STI value according to (4) formula.
m ( f ) = | ∫ 0 ∞ e - j 2 πft p 2 ( t ) dt | ∫ 0 ∞ p 2 ( t ) dt · ( 1 + 10 ( - S / N ) / 10 ) - 1 - - - ( 5 )
Practical measuring examples: measure the speech articulation of certain venue Public Address system at certain receiving position.
Method 1: in the time of growing, the language spectrum signal is after envelope modulation (shown in Figure 4), by Public Address system, play, in position to be measured, adopt microphone to receive envelope modulated signal, signal processing unit carries out after octave band filtering, envelope detected, low-pass filtering calculating the m of 5 * 14 data to signal K, fMatrix, calculate the STI value according to formula (1)~(4).
Method 2: adopt acoustic measurement software Dirac to measure (if used the index Frequency Sweeping Method) from Public Address system to the impulse response of accepting position as shown in Figure 5, and (250~4000Hz) signal and noise level calculate 5 * 14 m by Dirac software to measure each octave band of receiving position K, f(table 2), calculate the value of STI according to formula (1)~(4).
Table 2m K, f, MTI, STI
Figure BDA0000053582530000042
The present invention is by without reverberation with have under the reverberation condition, Whole frequency band Chinese language signal is carried out to filtering, obtain having the Speech articulation test signal of different frequency scope, by subjective Chinese language sharpness, estimate, obtain having the speech articulation score of the Chinese language signal of different frequency scope, experimental result shows: in order to reach Chinese language sharpness preferably, the low-limit frequency scope of required Chinese language signal is 300-6000Hz.(250~4000Hz) carry out having very high correlativity (seeing Fig. 6), coefficient R=0.958 between STI value that STI value that weighted obtains and employing I EC criterion calculation obtain to present embodiment by adopting 5 octave bands larger to the contribution of Chinese language sharpness.Fig. 7 a~Fig. 7 c is the STI value that obtains of 3 kinds of different weighted methods and the match relation between subjective Chinese articulation score, and table 1 is corresponding weighted value, coefficient R and standard deviation.This shows, the Chinese language sharpness can better be predicted and estimate to the STI value that the weighted method that the weighted adopted in IEC60268-16 (2003) and the application adopt obtains all.But, adopt IEC60268-16 (2003) thus standard need to be measured the modulation transmissions index of 7 octave bands and obtains the STI value, and the present invention only need measure the modulation transmissions index of 5 octave bands and can obtain STI, can save Measuring Time like this.

Claims (7)

1. testing intelligibility of speech transmission index, it is characterized in that adopting in position to be measured microphone to receive the test signal through tested speech transmission index transmission, the test signal that microphone will receive is transferred to signal processing unit, signal processing unit is by calculating the modulation transfer function of 250Hz to 4000Hz octave band, according to modulation transfer function computational language transmission index STI value, according to language transmission index STI value, estimate the sharpness of speech transmission index.
2. method of testing according to claim 1, is characterized in that described test signal is envelope modulated signal, and described signal processing unit carries out described calculating to the envelope modulated signal that receives after by octave band filtering, envelope detected, low-pass filtering again.
3. method of testing according to claim 1, is characterized in that described signal processing unit calculates the modulation transfer function of five octave bands of 250Hz to 4000Hz, then according to modulation transfer function computational language transmission index STI value.
4. method of testing according to claim 3, the carrier wave that it is characterized in that test signal is the octave narrow band noise by 250,500,1000,2000, centered by 4000Hz; Between each narrow band noise, leave isolation strip, the envelope modulated signal of test signal is the simple harmonic wave of extremely low frequency, its frequency is respectively 0.63Hz, 0.8Hz, 1.0Hz, 1.25Hz, 1.6Hz, 2.0Hz, 2.5Hz, 3.15Hz, 4.0Hz, 5.0Hz, 6.3Hz, 8.0Hz, 10.0Hz, 12.5Hz, totally 14 modulating frequencies.
5. method of testing according to claim 4 is characterized in that signal processing unit carries out described calculating and comprises:
Obtain 5 MTF curves of octave separation in 250~4000Hz frequency range, every curve carries out spectrum analysis in the 0.63-12.5Hz modulation frequency range with 1/3 octave band; By the m that measures 5 * 14 data that obtain k,fMatrix can be converted to respectively 70 apparent signal to noise ratio snr k,fNumerical value, k are that octave band, f are modulating frequency,
SNR k , f = 10 lg m k , f 1 - m k , f dB - - - ( 1 )
Then apparent signal to noise ratio snr k,fBe limited in ± ask transmission index TI in the 15dB scope k,f, and make 0≤TI k,f≤ 1,
TI k , f = SN R k , f + 15 30 - - - ( 2 )
Transmission index TI by same octave band k,fOn average to obtain modulation transmissions index M TI k:
MTI k = 1 14 Σ f = 1 14 TI k , f - - - ( 3 )
Computational language transmission index STI:
STI = Σ n = 1 5 ω n MTI n - - - ( 4 ) ω in formula nFor each octave band weighted value, corresponding to 250Hz~4000Hz octave band, ω nBe respectively 0.072,0.144,0.218,0.327 and 0.234.
6. method of testing according to claim 5, it is characterized in that the impulse response of adopting maximal length sequence signal or swept-frequency signal to measure the transmission channel of test language transmission system, the signal to noise ratio (S/N ratio) S/N when then application of formula (5) is according to impulse response p (t) and test calculates the m of 5 * 14 data k,fMatrix m (f), calculate the STI value according to formula (1)~(4),
m ( f ) = | ∫ 0 ∞ e - j 2 πft p 2 ( t ) dt | ∫ 0 ∞ p 2 ( t ) dt · ( 1 + 10 ( - S / N ) / 10 ) - 1 - - - ( 5 ) .
7. method of testing according to claim 6, it is characterized in that adopting acoustic software to measure room impulse response, and application software calculates the MTI of 250~4000Hz octave band kValue, then calculate the STI value according to (4) formula.
CN2011100825142A 2011-04-01 2011-04-01 Method for testing intelligibility of speech transmission index Expired - Fee Related CN102148033B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2011100825142A CN102148033B (en) 2011-04-01 2011-04-01 Method for testing intelligibility of speech transmission index

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2011100825142A CN102148033B (en) 2011-04-01 2011-04-01 Method for testing intelligibility of speech transmission index

Publications (2)

Publication Number Publication Date
CN102148033A CN102148033A (en) 2011-08-10
CN102148033B true CN102148033B (en) 2013-11-27

Family

ID=44422262

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2011100825142A Expired - Fee Related CN102148033B (en) 2011-04-01 2011-04-01 Method for testing intelligibility of speech transmission index

Country Status (1)

Country Link
CN (1) CN102148033B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106816158B (en) * 2015-11-30 2020-08-07 华为技术有限公司 Voice quality assessment method, device and equipment
CN105869656B (en) * 2016-06-01 2019-12-31 南方科技大学 Method and device for determining definition of voice signal
CN106531183A (en) * 2016-11-17 2017-03-22 中国传媒大学 Chinese speech articulation evaluation algorithm based on transmission system acoustic parameters
CN106504771A (en) * 2016-11-17 2017-03-15 中国传媒大学 A kind of Chinese summary evaluation and test algorithm suitable for noise circumstance
CN106558319A (en) * 2016-11-17 2017-04-05 中国传媒大学 A kind of Chinese summary evaluation and test algorithm suitable for limited bandwidth transmission conditions
CN108682430B (en) * 2018-03-09 2020-06-19 华南理工大学 Method for objectively evaluating indoor language definition
CN110956979B8 (en) * 2019-10-22 2024-06-07 合众新能源汽车股份有限公司 MATLAB-based automatic calculation method for in-vehicle language definition
CN111328008B (en) * 2020-02-24 2021-11-05 广州市迪士普音响科技有限公司 Sound pressure level intelligent control method based on sound amplification system
CN111757235A (en) * 2020-06-15 2020-10-09 杭州艾力特数字科技有限公司 Sound expansion system with classroom language definition measuring function
CN113873399B (en) * 2021-09-13 2022-05-27 中山大学 Method for improving speech definition of audio system
CN115424636B (en) * 2022-11-04 2023-02-03 广州声博士声学技术有限公司 Voice quality measurement method and device based on voice transmission index

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4975956A (en) * 1989-07-26 1990-12-04 Itt Corporation Low-bit-rate speech coder using LPC data reduction processing
CN101356854A (en) * 2006-03-03 2009-01-28 唯听助听器公司 Method and system of noise reduction in a hearing aid

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7433821B2 (en) * 2003-12-18 2008-10-07 Honeywell International, Inc. Methods and systems for intelligibility measurement of audio announcement systems
US7702112B2 (en) * 2003-12-18 2010-04-20 Honeywell International Inc. Intelligibility measurement of audio announcement systems
EP1994791B1 (en) * 2006-03-03 2015-04-15 GN Resound A/S Automatic switching between omnidirectional and directional microphone modes in a hearing aid

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4975956A (en) * 1989-07-26 1990-12-04 Itt Corporation Low-bit-rate speech coder using LPC data reduction processing
CN101356854A (en) * 2006-03-03 2009-01-28 唯听助听器公司 Method and system of noise reduction in a hearing aid

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
彭健新 et al.调制转移函数在厅堂语言清晰度预测中的应用.《电声技术》.2001,(第2期),15-19.
调制转移函数在厅堂语言清晰度预测中的应用;彭健新 et al;《电声技术》;20010228(第2期);15-19 *

Also Published As

Publication number Publication date
CN102148033A (en) 2011-08-10

Similar Documents

Publication Publication Date Title
CN102148033B (en) Method for testing intelligibility of speech transmission index
CN102157147B (en) Test method for objectively evaluating voice quality of pickup system
Gao et al. On the selection of acoustic/vibration sensors for leak detection in plastic water pipes
CN102664017B (en) Three-dimensional (3D) audio quality objective evaluation method
CN102301216B (en) leak detector
CN101430882B (en) Method and apparatus for restraining wind noise
Pal et al. Detecting & locating leaks in water distribution polyethylene pipes
CN101762806B (en) Sound source locating method and apparatus thereof
CN103440869A (en) Audio-reverberation inhibiting device and inhibiting method thereof
EP3166239A1 (en) Method and system for scoring human sound voice quality
Anderson et al. English verification of the STI method for estimating speech intelligibility of a communications channel
CN103050128B (en) Vibration distortion-based voice frequency objective quality evaluating method and system
CN105716707A (en) Ultra-low-frequency abnormal infrasound signal judging method
US20200378825A1 (en) Transducer transfer impedance calibration device based on spatial frequency domai smoothing technology
CN110534121A (en) A kind of monitoring method and system of the audio content consistency based on frequency domain character
Paliwal Estimation of noise variance from the noisy AR signal and its application in speech enhancement
Stelmachowicz et al. Complex and pure-tone signals in the evaluation of hearing-aid characteristics
Yang et al. Characterization and modeling of underwater acoustic communications channels for frequency-shift-keying signals
CN105510904B (en) Distance-finding method based on the summation of multifrequency audible wave amplitude
KR101369043B1 (en) Method of tracing the sound source and apparatus thereof
CN106710602B (en) Acoustic reverberation time estimation method and device
CN104730384A (en) Power disturbance identification and localization method based on incomplete S transformation
US20160080095A1 (en) Method and system for user speed estimation in wireless networks
CN113048404B (en) Urban gas pipeline tiny leakage diagnosis method
CN110796036B (en) Method for improving identification precision of structural modal parameters

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20131127

Termination date: 20190401

CF01 Termination of patent right due to non-payment of annual fee