sound-quality(聲品質)教學講解課件

sound-quality(聲品質)教學講解課件ObjectiveClassificationofSubjectivePercievedSoundQuality主觀音質評價的客觀分析方法朗德科技有限公司e-mail:landtp@ObjectiveClassificationofSuFhG德國夫瑯和佛研究所FhGIBP下屬建筑物理研究所朗德科技有限公司執(zhí)行DatakustikGmbH專利聲學材料和構件生產廠家中國客戶北京公司香港公司THALESHEIMSystemsGmbHNorsonicASHEADacousticGmbHFhG德國夫瑯和佛研究所FhGIBP下屬建筑物理研究所HEADacousticsHEADquarterGermany德國總部founded1986成立于1986年100employees現(xiàn)有雇員100名HEADacousticsHEADquarterGerHEADacousticsInc.HEADacoustics美國有限公司founded1988成立于1988年13employees現(xiàn)有雇員13名

HEADacousticsInc.HEADHEADconsultHEAD咨詢部Foundedin1992

成立于1992年3acousticchambersandtestrigs

3間聲學實驗室和3套測試臺10experiencedconsultingengineers

10名經驗豐富的工程師Technicians,mechanic,workshop

提供技師,機械師,車間Nationalandinternationalprojects

承接國內和國際合作項目Nearly50internationalcustomers

約50多個國際客戶Severalindustrialbranches

若干工業(yè)分支機構HEADconsultHEAD咨詢部FoundVehicleTestChamber機動車測試室Absorbentlining

ofceilingandwalls室內頂層和墻壁都有吸聲體覆蓋Cut-offfrequency

ofapprox.125Hz下限頻率約為125赫茲

Externalelectric

drive37kW外部電驅動功率37千瓦

Eddy-current

brake200kW渦流制動器功率200千瓦

Computercontrol全計算機控制VehicleTestChamber機動車測試室AbEngineTestRig發(fā)動機測試臺Externaldrive(37kW)andeddy-currentbrake(200kW)positioned

outsidetestchamber外驅動器（功率37千瓦）和置于測試室外的渦流制動器（功率200千瓦）

Speedrangefrom

700-7000rpm加速范圍從700~7000rpm

Externalcooling-system外冷卻系統(tǒng)Computercontrol全計算機控制EngineTestRig發(fā)動機測試臺ExternaAnechoicChamber全消聲室Absorbentliningonfloor,wallsandceiling室內頂層、地面和四面墻壁都有吸聲材料Cut-offfrequency180Hz下限頻率180赫茲Rotarytablefor

"Round-About-Measurements”為“Round-About-測量”所設置的旋轉臺AnechoicChamber全消聲室AbsorbenIntroduction導言UseofBinauralTechnologyinAcousticEngineeringis“StateoftheArt”在工程聲學領域運用雙耳模擬技術是目前最先進的方法SoundQualityrequiresconsiderationofhumanhearing’scharacteristics聲質量評估要求考慮人耳的聽覺特性WorkofNVHengineersisalwaysorientedtocustomers’needsNVH工程師的工作總是以用戶需求為中心NVHtasksofthefuturerequireintegrationofbinauraltechnologyandotherNVHmeasurementtechnologyNVH的未來任務是要把雙耳模擬技術和其它NVH測量技術結合在一起使用Introduction導言UseofBinauraHowdoesNoisetalktous?我們是如何感知噪聲的？Noiseisinformative噪聲是一種信息quality,functionality,danger,environment,…

質量性,功能性,危險性,環(huán)境,…Noiseimpliesacertainimage噪聲意味著某種特性luxury,sportive,cheap,…

豪華型,運動型,便宜型,…

Noisemayidentify

噪聲是可以被識別的similartoopticalimpression類似于視覺感受corporatesound,productsound

總體的聲音，個體的聲音HowdoesNoisetalktous?我們

SameA-level,SameThirdOctaveSpectrum

同樣的A聲級，同樣的1/3倍頻程頻響曲線74,2dB75,1dB74,7dBSameA-level,SameThirdOctaProblem:objectivedescriptionofsubjectiveevaluation

問題：主觀評價的客觀描述Problem:objectivedescriptionsound-quality(聲品質)教學講解課件sound-quality(聲品質)教學講解課件sound-quality(聲品質)教學講解課件Conventional/Binaural傳統(tǒng)手段/雙耳模擬Comparison:Conventional-binauralmeasurementtechnique對照：傳統(tǒng)測量技術----雙耳模擬測量技術Theouterearisadirectionalfilterwhichweightsthesoundpressureinarangeof+15/-30dB,dependingonfrequencyanddirectionofsound

外耳是有方向性的濾波器，根據(jù)聲音頻率和方向的不同，其感受聲音的聲壓級大小在+15/-30分貝范圍內有一個加權。Humanhearinghastwoinputs:人耳聽覺輸入包括兩個部分：

binauralsignalprocessing雙耳輸入信號處理

directionhearing,selectivity,noisesuppression

聽覺的方向性，信號的選擇性，對噪聲信號的抑制Psychoacousticofhearingdeterminesthesoundimpression:

loudness,sharpness,roughness,tonality

人耳聽覺對聲音的印象取決于心理聲學的參數(shù)：響度、尖銳度、粗糙度等。Highresolutionofhearingintheamplitude,frequency-andtimedomain:dynamics>120dB

BxT<0.3,analysisbandwidthxanalysisinterval

人耳聽覺在幅值、頻率、時間三維空間里的高分辨率Conventional/Binaural傳統(tǒng)手段/Loudness:響度Considerationofdistributionofcriticalbandsandmaskingpropertiesinthehearing,考慮到特征頻帶的分布和人耳

DIN45631,ISO532B掩蔽效應，符合DIN45631和Sharpness:尖銳度ISO532B的標準Weightedfirstmomentofdistributionofcriticalbandratesofspecificloudness,proportionofhigh-frequencyspectralcomponentstolow-frequencyones

反映首次特征頻帶分布和指定響度級的關系，把高頻段成分和低頻段成分相比。

Roughness/FluctuationStrength:粗糙度/起伏程度Timestructureofthesoundsignal,modulationfactorandleveldifferencedetermineroughnessamplitude-andfrequencymodulation聲音信號的時域結構，調制因子，和聲壓級的不同決定了粗糙度對幅值和頻率的調制PsychoacousticalParameter心理聲學參數(shù)Loudness:響度PsychoacousticaphysicspsychoacousticspsychologyHumanHearingphysicspsychoacousticspsycholo77dB(A)76.7dB(A)76.6dB(A)36,7sone14.0sone38,8soneLoudnessversusA-weightedSPL

響度級:A計權聲壓級77dB(A)76.7dB(A)76.6dB(A)36SimultaneousMasking同時掩蔽

SimultaneousMasking同時掩蔽Sharpness尖銳度Version1:AuresVersion2:vonBismarckSharpness尖銳度Version1:AurComparison,SharpnessvonBismark/Aureswei?esRauschenwei?esRauschen,8kHzverst?rktvonBismarkAuresvonBismarkAuresComparison,SharpnessvonBismSinustone1kHzSinustone1kHzand1001HzInterferenzSinustone1kHzand1004HzFluctuationSinustone1kHzand1020HzR-Roughness(Rumbling)Sinustone1kHzand1070HzRoughnessSinustone1kHzand1414HztwotonecomplexTwoTones:DifferentPerceptions兩種音調：不同感覺

Sinustone1kHzSinustone1kHzParametersofRoughness粗糙度的參數(shù)ParametersofRoughness粗糙度ApplicationExample:“Rumbling”Noise應用舉例：“隆隆聲”噪音specialkindofroughness

特殊種類的粗糙度psychoacousticroughnesscalculationfails

心理聲學粗糙度級的計算不成功psychoacousticfluctuationstrengthanalysisfails心理聲學波形起伏程度分析的計算不成功socalled?r“roughness

所謂的?r“粗糙度modulationfrequencyrangearound15to30Hz

調制頻率范圍15~30赫茲relevantfrequencyrangefrom150to450Hz

相關頻率范圍150~450赫茲causedbyinteractionofdifferent0.5orders

由0.5階互動分析的差異所引起ApplicationExample:“RumblingaverageFFTspectrum?rumbling“隆隆聲的平均FFT頻譜5cylinderengineat2420rpm5缸發(fā)動機在2420轉時averageFFTspectrum?rumblingbandpassfilteredtimesignal經過帶通濾波器處理的時域信號bandpassfilteredtimesignal5thorder4.5thorder5.5thorder5thorder4.5thorder5.5thomodulationspectrumof250Hzoctave250赫茲調制頻譜modulationspectrumof250HzMultidimensionalAspectsofSound聲音的多維性designaspect設計方面physicalaspect物理方面psycho-acousticalaspect心理聲學方面cognitiveaspect認知度方面MultidimensionalAspectsofSoComparisonLoudness/dB(A)比較響度/分貝（A計權）ComparisonLoudness/dB(A)ImportanceoftheA-weighted

SoundPressureLevel

A計權聲壓級的重要性85dB(A)LSPL65dB(A)Physicaldamageofhearingpossiblycausedbysound聲音可能會導致聽力的物理損傷MeasurementbysufficientA-weightedsoundpressurelevel足夠滿足要求的A計權聲壓級測量Physiologicalimpairmentcausedbysound由于聲音導致的生理損傷A-weightedsoundpressurelevelmeasurementsnecessarybutnotsufficientA計權聲壓級的測量是必要的，但并不足以滿足要求Annoyancebynoisewithvegetativeeffectspossible日常生活中令人煩惱的噪聲A-weightedsoundpressurelevelisnotsuitedfordescribingthesoundqualityA計權聲壓級不足以用來描述聲音的質量ImportanceoftheA-weighted

ParametersofSoundQuality聲質量的參數(shù)Thesoundqualitysurelydoesn’tdependonlyupontheA-weightedsoundpressurelevel.很顯然,聲質量不僅僅取決于A-計權聲壓級Thesoundqualityofasoundeventdependsnotonlyonabsoluteparameterslikeloudness,sharpness,roughness,fluctuationstrengthandsoon

聲音的聲質量不僅僅取決于響度、尖銳度、粗糙度、起伏程度等絕對參數(shù)butitdependsonspectrum‘sstructureandthepatternintimedomain,

還取決于聲音的頻譜構成結構和在時域的聽覺模板組成情況

thatmeans:這就是說Changesoftheshorttimespectrumwithrespecttothefrequencyandtimedomainhaveaninfluenceontheclassificationofsoundquality

短時間頻譜或時域組成方面的變化都會對聲音的質量產生影響ParametersofSoundQuality聲RelativeApproachofHumanHearing評估人耳聽覺的相關方法RelativeApproachofHumanHea"good"3rdoctavespectrumspec.loudnessspec.roughness4.Application:Tirenoiseinthevehicleinterior應用4：輪胎產生的車內噪聲"good"3rdoctavespectrumspec."noisy"-3dB"good"3rdoctavespectrumspec.loudnessspec.roughnessExample:Tirenoiseinthevehicleinterior舉例：輪胎產生的車內噪聲"noisy"-3dB"good"3rdoctave"good"“annoying"Modulationspectrumversusfrequency15Hz+harmonics->TirerotationstrongmodulationinthehigherfrequentareaExample:Tirenoiseinthevehicleinterior舉例：輪胎產生的車內噪聲"good"“annoying"Modulationspe"good"“annoying"frequencyselectedlevelversustime,2msintegrationtimeconstantnarrowbandlevelat3200Hz

higherlevelstrongtemporalstructureca.20dBleveldifferenceca.10dBlevelfluctuationExample:Tirenoiseinthevehicleinterior舉例：輪胎產生的車內噪聲"good"“annoying"frequencyselesound-quality(聲品質)教學講解課件Workingina

recordingstudio

在實驗室工作Workingina

recordingstudioConclusion結論Soundengineeringbecomesveryimportant聲音工程學非常重要Soundqualitycannotbemeasuredwithconventionalmeasurementtechnique傳統(tǒng)測量手段不能對聲質量進行分析ArtificialHead,psychoacoustics,multi-channelanalysisarehelpfultoolsforsoundengineering人工頭、心理聲學、多通道分析都是對聲音工程學有幫助的工具but:但是Forsounddesignweneedthehumanfactor,theexperttocreateagoodsound由于聲音設計有人的因素，所以需要專家來幫助我們創(chuàng)造出好聽的聲音Conclusion結論Soundengineerinsound-quality(聲品質)教學講解課件ObjectiveClassificationofSubjectivePercievedSoundQuality主觀音質評價的客觀分析方法朗德科技有限公司e-mail:landtp@ObjectiveClassificationofSuFhG德國夫瑯和佛研究所FhGIBP下屬建筑物理研究所朗德科技有限公司執(zhí)行DatakustikGmbH專利聲學材料和構件生產廠家中國客戶北京公司香港公司THALESHEIMSystemsGmbHNorsonicASHEADacousticGmbHFhG德國夫瑯和佛研究所FhGIBP下屬建筑物理研究所HEADacousticsHEADquarterGermany德國總部founded1986成立于1986年100employees現(xiàn)有雇員100名HEADacousticsHEADquarterGerHEADacousticsInc.HEADacoustics美國有限公司founded1988成立于1988年13employees現(xiàn)有雇員13名

HEADacousticsInc.HEADHEADconsultHEAD咨詢部Foundedin1992

成立于1992年3acousticchambersandtestrigs

3間聲學實驗室和3套測試臺10experiencedconsultingengineers

10名經驗豐富的工程師Technicians,mechanic,workshop

提供技師,機械師,車間Nationalandinternationalprojects

承接國內和國際合作項目Nearly50internationalcustomers

約50多個國際客戶Severalindustrialbranches

若干工業(yè)分支機構HEADconsultHEAD咨詢部FoundVehicleTestChamber機動車測試室Absorbentlining

ofceilingandwalls室內頂層和墻壁都有吸聲體覆蓋Cut-offfrequency

ofapprox.125Hz下限頻率約為125赫茲

Externalelectric

drive37kW外部電驅動功率37千瓦

Eddy-current

brake200kW渦流制動器功率200千瓦

Computercontrol全計算機控制VehicleTestChamber機動車測試室AbEngineTestRig發(fā)動機測試臺Externaldrive(37kW)andeddy-currentbrake(200kW)positioned

outsidetestchamber外驅動器（功率37千瓦）和置于測試室外的渦流制動器（功率200千瓦）

Speedrangefrom

700-7000rpm加速范圍從700~7000rpm

Externalcooling-system外冷卻系統(tǒng)Computercontrol全計算機控制EngineTestRig發(fā)動機測試臺ExternaAnechoicChamber全消聲室Absorbentliningonfloor,wallsandceiling室內頂層、地面和四面墻壁都有吸聲材料Cut-offfrequency180Hz下限頻率180赫茲Rotarytablefor

"Round-About-Measurements”為“Round-About-測量”所設置的旋轉臺AnechoicChamber全消聲室AbsorbenIntroduction導言UseofBinauralTechnologyinAcousticEngineeringis“StateoftheArt”在工程聲學領域運用雙耳模擬技術是目前最先進的方法SoundQualityrequiresconsiderationofhumanhearing’scharacteristics聲質量評估要求考慮人耳的聽覺特性WorkofNVHengineersisalwaysorientedtocustomers’needsNVH工程師的工作總是以用戶需求為中心NVHtasksofthefuturerequireintegrationofbinauraltechnologyandotherNVHmeasurementtechnologyNVH的未來任務是要把雙耳模擬技術和其它NVH測量技術結合在一起使用Introduction導言UseofBinauraHowdoesNoisetalktous?我們是如何感知噪聲的？Noiseisinformative噪聲是一種信息quality,functionality,danger,environment,…

質量性,功能性,危險性,環(huán)境,…Noiseimpliesacertainimage噪聲意味著某種特性luxury,sportive,cheap,…

豪華型,運動型,便宜型,…

Noisemayidentify

噪聲是可以被識別的similartoopticalimpression類似于視覺感受corporatesound,productsound

總體的聲音，個體的聲音HowdoesNoisetalktous?我們

SameA-level,SameThirdOctaveSpectrum

同樣的A聲級，同樣的1/3倍頻程頻響曲線74,2dB75,1dB74,7dBSameA-level,SameThirdOctaProblem:objectivedescriptionofsubjectiveevaluation

問題：主觀評價的客觀描述Problem:objectivedescriptionsound-quality(聲品質)教學講解課件sound-quality(聲品質)教學講解課件sound-quality(聲品質)教學講解課件Conventional/Binaural傳統(tǒng)手段/雙耳模擬Comparison:Conventional-binauralmeasurementtechnique對照：傳統(tǒng)測量技術----雙耳模擬測量技術Theouterearisadirectionalfilterwhichweightsthesoundpressureinarangeof+15/-30dB,dependingonfrequencyanddirectionofsound

外耳是有方向性的濾波器，根據(jù)聲音頻率和方向的不同，其感受聲音的聲壓級大小在+15/-30分貝范圍內有一個加權。Humanhearinghastwoinputs:人耳聽覺輸入包括兩個部分：

binauralsignalprocessing雙耳輸入信號處理

directionhearing,selectivity,noisesuppression

聽覺的方向性，信號的選擇性，對噪聲信號的抑制Psychoacousticofhearingdeterminesthesoundimpression:

loudness,sharpness,roughness,tonality

人耳聽覺對聲音的印象取決于心理聲學的參數(shù)：響度、尖銳度、粗糙度等。Highresolutionofhearingintheamplitude,frequency-andtimedomain:dynamics>120dB

BxT<0.3,analysisbandwidthxanalysisinterval

人耳聽覺在幅值、頻率、時間三維空間里的高分辨率Conventional/Binaural傳統(tǒng)手段/Loudness:響度Considerationofdistributionofcriticalbandsandmaskingpropertiesinthehearing,考慮到特征頻帶的分布和人耳

DIN45631,ISO532B掩蔽效應，符合DIN45631和Sharpness:尖銳度ISO532B的標準Weightedfirstmomentofdistributionofcriticalbandratesofspecificloudness,proportionofhigh-frequencyspectralcomponentstolow-frequencyones

反映首次特征頻帶分布和指定響度級的關系，把高頻段成分和低頻段成分相比。

Roughness/FluctuationStrength:粗糙度/起伏程度Timestructureofthesoundsignal,modulationfactorandleveldifferencedetermineroughnessamplitude-andfrequencymodulation聲音信號的時域結構，調制因子，和聲壓級的不同決定了粗糙度對幅值和頻率的調制PsychoacousticalParameter心理聲學參數(shù)Loudness:響度PsychoacousticaphysicspsychoacousticspsychologyHumanHearingphysicspsychoacousticspsycholo77dB(A)76.7dB(A)76.6dB(A)36,7sone14.0sone38,8soneLoudnessversusA-weightedSPL

響度級:A計權聲壓級77dB(A)76.7dB(A)76.6dB(A)36SimultaneousMasking同時掩蔽

SimultaneousMasking同時掩蔽Sharpness尖銳度Version1:AuresVersion2:vonBismarckSharpness尖銳度Version1:AurComparison,SharpnessvonBismark/Aureswei?esRauschenwei?esRauschen,8kHzverst?rktvonBismarkAuresvonBismarkAuresComparison,SharpnessvonBismSinustone1kHzSinustone1kHzand1001HzInterferenzSinustone1kHzand1004HzFluctuationSinustone1kHzand1020HzR-Roughness(Rumbling)Sinustone1kHzand1070HzRoughnessSinustone1kHzand1414HztwotonecomplexTwoTones:DifferentPerceptions兩種音調：不同感覺

Sinustone1kHzSinustone1kHzParametersofRoughness粗糙度的參數(shù)ParametersofRoughness粗糙度ApplicationExample:“Rumbling”Noise應用舉例：“隆隆聲”噪音specialkindofroughness

特殊種類的粗糙度psychoacousticroughnesscalculationfails

心理聲學粗糙度級的計算不成功psychoacousticfluctuationstrengthanalysisfails心理聲學波形起伏程度分析的計算不成功socalled?r“roughness

所謂的?r“粗糙度modulationfrequencyrangearound15to30Hz

調制頻率范圍15~30赫茲relevantfrequencyrangefrom150to450Hz

相關頻率范圍150~450赫茲causedbyinteractionofdifferent0.5orders

由0.5階互動分析的差異所引起ApplicationExample:“RumblingaverageFFTspectrum?rumbling“隆隆聲的平均FFT頻譜5cylinderengineat2420rpm5缸發(fā)動機在2420轉時averageFFTspectrum?rumblingbandpassfilteredtimesignal經過帶通濾波器處理的時域信號bandpassfilteredtimesignal5thorder4.5thorder5.5thorder5thorder4.5thorder5.5thomodulationspectrumof250Hzoctave250赫茲調制頻譜modulationspectrumof250HzMultidimensionalAspectsofSound聲音的多維性designaspect設計方面physicalaspect物理方面psycho-acousticalaspect心理聲學方面cognitiveaspect認知度方面MultidimensionalAspectsofSoComparisonLoudness/dB(A)比較響度/分貝（A計權）ComparisonLoudness/dB(A)ImportanceoftheA-weighted

SoundPressureLevel

A計權聲壓級的重要性85dB(A)LSPL65dB(A)Physicaldamageofhearingpossiblycausedbysound聲音可能會導致聽力的物理損傷MeasurementbysufficientA-weightedsoundpressurelevel足夠滿足要求的A計權聲壓級測量Physiologicalimpairmentcausedbysound由于聲音導致的生理損傷A-weightedsoundpressurelevelmeasurementsnecessarybutnotsufficientA計權聲壓級的測量是必要的，但并不足以滿足要求Annoyancebynoisewithvegetativeeffectspossible日常生活中令人煩惱的噪聲A-weightedsoundpressurelevelisnotsuitedfordescribingthesoundqualityA計權聲壓級不足以用來描述聲音的質量ImportanceoftheA-weighted

ParametersofSoundQuality聲質量的參數(shù)Thesoundqualitysurelydoesn’tdependonlyupontheA-weightedsoundpressurelevel.很顯然,聲質量不僅僅取決于A-計權聲壓級Thesoundqualityofasoundeventdependsnotonlyonabsoluteparameterslikeloudness,sharpness,roughness,fluctuationstrengthandsoon

聲音的聲質量不僅僅取決于響度、尖銳度、粗糙