200920160 九、發明說明： 【發明所屬之技術領域】 本發明涉及-種音頻測試裝置及方法，特別涉及一種 對發聲設備音頻介面進行測試的裝置及方法。 【先前技術】 二前’對有音頻介面如耳機介面的發聲設備左右聲道 的測忒一般有以下兩種方法，一 桩入n德L 裡疋人工松測，主要依靠 機’由作業員耳絲從發聲設備音頻 聲音來判別發聲設備的品質 〜㈣出的 業員主㈣塑士 &、 間早，但受作 質不高引起的聲音音質下:兄…響以及耳機品 同時作鞏η 果不準確，檢測品質低， 吋作業貝的聽力也會受到損宝。一 測量儀器測量如Audi。Preci二八 rr頻分析儀進㈣試，該音_簡^=== 確’但測試儀器的價格昂貴。 ‘果很準 此外，一般音頻介面如耳機介 測試從音頻介面輸出聲音的始貝好壞是通過 頻率回應以及滿格度失直：::波失真、信噪比、串音、 頻介面輸出音頻文件進m的，而現在測試時從音 卿的音頻試的方法是-般輸出-段 的日頻4藏求出總譜波失直权 輪出-段掃頻信號即混合多個頻率的，，然 等參數’這樣導致測試時間比較長。a來求出頻率回 【發明内容】 有鑒於此，提供 種音頻測試裝置和方法， 該方法可 200920160 在分成不同時段的-短暫時間内輸出一系列各 :編號，同時採集各時段的音頻信號，根據採年： 頻^可對音頻介面的多個參數同時進行測量。不曰 貴的儀器’也不需要針對不同的參數播放不同的=叩 件’該裳置和方法可大大減小成本並縮短測試時間。、文 —-種音頻測試1置，包括—聲音採集裝置、—己 ^顯不m中央處㈣。其中，該記憶體儲、 =频文件，該測試音频文件的音頻數據包括有多:: 」數據部分，料音紐據部分分概表現為播放時.二 -、第二聲道均為靜音狀態的音頻數據部分；第 =轉ΓΤ:第二聲道保持為靜音狀態的音： ㈣刀，第-、第二聲道均輸出多個 =頻數據部分;第一聲道為靜音狀態，第二 頻率的音頻信號的音頻數據部分。 ， 的數裝置的數據介面通過—數據線與發聲設備 二同時’該音頻測試裝置的音頻介面通過 時，曰投備的音頻介面連接，在進行音頻測試 聲部= 送到發聲設備，發 隹穿置文件通過音頻介面輸出，同時音頻採 二頻介面採集該測試音频文件各部分的音頻數 存儲的 =在記憶體中，最後由中央處理器根據記憶體. 器輸出^行各項參數的測試，並將結果通過顯示 種音頻測試的方法，包括以下步驟：播放一音频文 9 200920160 :出放隼過第程V广時_ ^ *絲枝抓集弟一時間段内左右聲道的噪立作 :左聲道輸後儲存於記憶體中;在第二時： 態;採集第二時間道保持為靜音狀 道的串擾彳耳 率音齡號以及右聲 第三時間。左儲存於記憶體中；在 音頻信號， ===信號以及左聲道的串擾信號，=: 要的==置::法可在-段短暫的時間裏輸出測試需 每個時間騎:出;的 保二度對:=::個參數同時進行測量，在確 【實施方式】 的同時，減少了測量時間。 請參閱圖1，為音頻測試裝 奋 20、—記情體m β 一中央處理器 錯存有_、5 及—顯示設備3〇。其中，記憶體4〇中 發聲設備件’ #對發聲設備1進㈣試時，該 、數據介面（圖+未示）如USB介面或ieee 1394 200920160 ··=通過-數據、_中未示）與音制試 . 峨，該音频文件被傳輪至發聲設備…缺 音頻介面_輸出。本實施方式中:、該發ί; 備1為有第-、第二聲道的雙聲道音 =又 、^立 曰效卡。該發聲設備1的音頻介面102通 =1曰，線12連接至聲音採集裝置H)，則由發聲設備i ^頻^面1〇1的第一聲道與第二聲道輸出的音頻信號經 40,巾處裔20將該數位音頻信號儲存於記怜體 i〇卜然後中央處理器對儲存於記憶體40中的第ϋ體 的數:音頻信號進行處理，以對發聲設備音頻介 油借波失直、ί—聲運的各項參數如信噪比、聲道串音、 〜咕波失真、滿格度失真及頻率 結果通過顯示設備30輸出。應進订料，並將測試 該測試音频文件為一特殊音频文件，其專用于本發明 數試，該測試音频文件的音頻數據包括有多個音頻 時㈣所述音頻數據部分分別能表現為播放時：第一 分.曰第又^内第一、第二聲道均為靜音狀態的音頻數據部 二聲内第一聲道輸出單頻率的音頻信號，第 内签持為靜音狀態的音頻數據部分；第三時間段T3 數據_弟Γ聲道均輸出多個頻率混合的音頻信號的音頻 道輸in四時間段T4内第—聲道為靜音狀態，第二聲 ，出早頻率的音頻信號的音頻數據部分。所述第一、第 11 200920160 二聲道分別為左、右聲道中的—種’在本實施 =:方[聲道定義為左聲道，第二聲道定2 右}道虽然該弟―、第二聲道的定義可以互換。 在不同實施方式中，第時間段T1、第二時間段了丄 時間段T3及第四時間段T4對應的音頻數 : 可以變換的。 刀幻增序疋 〇月多閱圖2 ’為發聲設備音頻介面測試方法中聲立播 放過程的示意圖。在第一時間段η内，第―、第二曰， 右聲道)均為靜音狀態，在第二時間段τ2内左n 頻率的音頻信號，右聲道保持為靜音狀態，在第三:= ^内左右聲道均輪出多個頻率混合的音頻信號，在第四^ Η & Τ4内左聲道為靜音狀態’右聲道輸$單解的音頻Γ 遽。在本實施方式巾，該單鮮錢g1KHZ的音頻 :參閱：3,為發聲設備音頻介面的音頻測試二 文件傳輸至發聲設備,，該發聲設備“:二 ==件⑽。);在第一段時間τι，音= ^逼輸出為靜音狀態⑽υ;聲音採集裝置1G採隹今 二噪音信號，並將該噪音信號轉換餘位 .1 、处理益20將其儲存到記憶體40中（S312); —&時間T2 ’音頻介面逝左聲道輪出單頻率立頻广 $=聲道保持靜音狀態⑽3);聲音採集裝 該&時間左聲道單解音頻信號以及 = 將其轉換成數位音頻信號後，由中央處理器2=二 12 200920160 記憶體4〇中（S314);在第三 道輪出的均為多_率的二15)曰介面左右聲 10按隹兮队士 肩1口谈（b315)，聲音採集裝置 1〇知木該段時右聲道 平表置 成數位音頻信號後，由中 f H，並分別轉換 40中（S3l6Vy_ 處理益20將其錯存到記憶體 H ^ 段_T4,右聲道輸㈣頻率音頻 “虎’左聲道輪出靜音狀態（s3i7);聲音採 集該段時間裏的右聲道單頻率信號以及左γ 木 並分別轉換成數位音頻 串曰‘唬， 情體40中W1S、 由中央處理器2〇儲存到記 H );婦咖⑽中存㈣各時間段内的 2位:頻錢對發聲設備音頻介面ΠΠ左右聲道的各項參 ^订·（S319)。各參數的具體職方法請參閱圖*〜 圃8。 u閱圖4，為發聲設備音頻介面的信噪比(Signal to 二雜，SNR)測試流程圖。首先，中央處理器如根據記憶 八的在第—相段了1所採#的左右聲料音信號 刀別求出左右聲道的嚼音電平NL、NR(s術）；中央處理 器2〇根據記憶體40儲存的在第二段時間T2採集的左聲道 早頻率音頻信號求出左聲道信號電平％ (讀）；中央處 理器20根據記憶體4〇儲存的在第四時間段丁4採集的右聲 道早頻率音頻信號求出右聲道信號電平sr(s4g3);中央 處理器2〇基於左聲道信號電平SL以及左聲道噪音電平NL 求出左聲道信噪比為SNR=朋帆⑹，基於右聲道信號電 平SR以及右聲道噪音電平腿求出右聲道信噪比為 遞=2〇lg(V〜（S404 );將測試結果輸出至顯示設備30 200920160 (S405) 0 請參閱圖5,為發聲設備音頻介面的左右聲道串音測 試流程圖。首先，+央處理器2〇根據記憶體4〇存儲的在 第二時間段T2職的左聲料辭音齡號以及右聲道 的串音求出左聲道信號電平SL以及右聲道串音信號 電平CR(S5〇1);中央處理器20根據記憶體40儲存的在° 第四時間段T4採集的右聲道單頻率音頻信號以及左聲道 串音信號’未出右聲道信號電平狄以及左聲道串音信號電 平CL (S5f，中央處理器2〇基於左聲道串音信號電平 CL以及右荦遏信號電平SR求出左聲道串音， 基於右聲道串音信號電平CR以及左聲道信號電平％求出 右聲道串音湖3);將測試結果通過顯示 30 輸出（S504)。 明參閱圖6’為發聲設備音頻介面的總譜波失真（丁〇如 Harmonic Distortion ’ THD)測試流程圖。首先，中央處理 器20根據記憶體4〇儲存的在第二時間段τ2採集的左聲道 單頻率音頻信號以及在第四時間段W採集的右聲道單頻 率音頻信號分別做傅立葉變換得到該左聲道單頻率音頻信 號的頻域信號以及右聲道單頻率音頻信號的頻域信號 (S601)’巾央處理n加根據該左聲道單解音頻信號的 頻域信號求出左聲道音頻信號各諧波成分的幅值 凡根據該右聲道單頻率音頻信號的頻域信號求 出右聲道音頻信號各諳波成分的幅值W ( ); 基於左聲道音齡號各敎成分幅值，求丨左聲道總譜波 200920160 失真為：THD = /(//L22+HL32+...+HLN2) /(HLI2 +Hl22+Hl32 +...+HLN2) 士 ，&於右聲 逼音頻信號各諧波成分幅值，求出右聲道總諧波失真為. THD ~ Vc^2~+^i?32+... +HRNZ)/(HR\2 +ΗR22+HΛ32+... +Hrn2) ( C/ςπ^ ^ ^ , (S603)，將測試結 果輪出至顯示設備30 (S604)。 請參閱圖7,為發聲設備音頻介面的滿格度失真測試 机耘圖。在測試時，中央處理器2〇執行如下步驟：根據記 憶體40儲存的在第二時間段T2和第四時間段Τ4採集的 左、右聲道單頻率音頻信號，求出左聲道信號電平和右聲 號％平（S701 ) ’將得到的左、右聲道信號電平與—預 設』規格電平作比較⑽2);若左聲道或右聲道信號電平與 預設規格電平差值超ά—預設範圍，職明左聲道或右聲 ，存在滿格度失真，否則不存在滿格度失真（S7G3);將測 試結果輪出至顯示設備3〇 (S7〇4)。 ’、 請參閱圖8’為發聲設備音頻介面的頻率響應測气^ 測試時，中央處理請執行如下步驟：對記= 儲存的在第三時間段T3採集的左、右聲道多頻率 段相同長度的音頻信號⑽1);對該左 到加窗處理後分別進行快速傅立葉變換得 多個頻域信號取平均值求出—左聲道的平的 頻域信=屮! 3);根據該左、右聲道的平均 s 左右聲道的頻率回應值（S804);將社果幹出 至顯示設備30 (S805)。 果輪出 200920160BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio testing device and method, and more particularly to an apparatus and method for testing an audio interface of a sounding device. [Prior technology] Two fronts have the following two methods for measuring the left and right channels of the sounding device with audio interface such as the earphone interface. One is into the n-L 疋 疋 artificial tampering, mainly relying on the machine 'by the operator's ear Silk from the sound of the sound equipment to determine the quality of the sound equipment ~ (four) out of the owner (four) plastics & early, but the quality of the sound caused by low quality: brother ... ring and earphones at the same time If the result is inaccurate, the quality of the test is low, and the hearing of the homework will also be damaged. A measuring instrument measures such as Audi. The Preci 28 rr frequency analyzer enters (4) and the sound _ 简 ^=== YES 'But the test instrument is expensive. 'The fruit is very accurate. In addition, the general audio interface such as the earphone test test the sound output from the audio interface is through the frequency response and the full scale of the straight line::: wave distortion, signal to noise ratio, crosstalk, frequency interface output audio The file is entered into m, and now the test method from the voice of the voice is the general output - the daily frequency of the segment is 4 to find the total spectrum wave loss right-round-segment sweep signal that mixes multiple frequencies, And then the parameter 'this results in a longer test time. a to find the frequency back [invention] In view of this, an audio testing device and method are provided, which can output a series of numbers in a short time divided into different time periods in 200920160, and simultaneously collect audio signals of each time period, According to the picking date: Frequency can be measured simultaneously on multiple parameters of the audio interface. Not expensive instruments don't need to play different 叩 针对 针对 ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ 。 。 。 , text - a kind of audio test 1 set, including - sound collection device, - has not been the center of the m (four). Wherein, the memory storage, the frequency file, the audio data of the test audio file includes a plurality of:: ” data part, the part of the material sound data is represented as playing time. The second channel and the second channel are all muted. The audio data part; the first = turn: the second channel keeps the mute state: (4) the knife, the first and second channels both output multiple = frequency data parts; the first channel is muted, the second The audio data portion of the frequency audio signal. The data interface of the digital device passes through the data line and the sounding device 2. When the audio interface of the audio test device passes, the audio interface of the audio test device is connected, and the audio test part is sent to the sounding device, and the sound is worn. The file is output through the audio interface, and the audio frequency is collected by the second frequency interface to collect the audio number of each part of the test audio file. In the memory, finally, the central processor outputs the test according to the memory. And the result is displayed by the method of displaying the audio test, including the following steps: playing an audio file 9 200920160: when the first pass V wide is released _ ^ * silk branch captures the noise of the left and right channels in a period of time : The left channel is stored in the memory after the input; in the second time: the state; the crosstalk rate of the second time track is kept silent, and the third time of the right sound. Left stored in memory; in audio signal, === signal and left channel crosstalk signal, =: desired == set:: method can be output in a short period of time - test needs to ride every time: out The second degree of protection: =:: The parameters are measured at the same time, and the measurement time is reduced while the [implementation] is confirmed. Please refer to Figure 1. Competing for the audio test. 20—The syllabus m β A central processor has _, 5 and — display devices 3 错. Among them, the memory device 4 in the memory device '# to the sound device 1 into (four) test, the data interface (Figure + not shown) such as USB interface or ieee 1394 200920160 · · = pass - data, _ not shown With the sound system test. Hey, the audio file is transferred to the sounding device... lack of audio interface_output. In the present embodiment: the device 1 is a two-channel tone with a first-and second channel, and a valid card. The audio interface 102 of the sounding device 1 is connected to the sound collecting device H), and the audio signal output by the first channel and the second channel of the sounding device i ^ frequency surface 1〇1 is 40, the towel person 20 stores the digital audio signal in the memory and then the central processor processes the number of the first body: audio signal stored in the memory 40 to borrow from the sounding device audio oil The parameters of the wave loss, ί-sound, such as signal to noise ratio, channel crosstalk, ~ chopping distortion, full-scale distortion, and frequency results are output through the display device 30. The test audio file should be tested as a special audio file, which is dedicated to the digital test of the present invention. When the audio data of the test audio file includes multiple audios, (4) the audio data portion can be represented as playing. Time: the first point. The first and second channels in the first and second channels are in the mute state of the audio data portion. The first channel outputs the single frequency audio signal, and the first is the mute state audio data. Part; the third time period T3 data _ Γ Γ channel output audio signal of a plurality of frequency mixed audio channel input in the four time period T4, the first channel is mute state, the second sound, the early frequency audio signal The audio data portion. The first and the eleventh 200920160 two channels are respectively in the left and right channels - in this implementation =: square [channel defined as left channel, second channel fixed 2 right] although the brother ―, the definition of the second channel can be interchanged. In different embodiments, the number of audios corresponding to the time period T1, the second time period, the time period T3, and the fourth time period T4: can be transformed. Knife illusion ascending 疋 〇月多多图2 ’ is a schematic diagram of the sound playback process in the audio interface audio test method. In the first time period η, the first, second, and right channels are all in a mute state, and in the second time period τ2, the left n frequency audio signal, the right channel remains muted, in the third: = ^ The left and right channels are all rounded out with multiple frequency mixed audio signals. In the fourth ^ Η & Τ4, the left channel is muted and the right channel is input with a single solution of audio Γ 遽. In the embodiment of the towel, the single fresh money g1KHZ audio: see: 3, for the audio device audio interface audio test two files are transmitted to the sounding device, the sounding device ": two == piece (10).); at the first The segment time τι, tone = ^ forced output to the mute state (10) υ; the sound collecting device 1G picks up the second noise signal, and converts the noise signal into the remaining position. 1 , the processing benefit 20 stores it in the memory 40 (S312) ; - & time T2 'audio interface dead left channel round single frequency vertical frequency wide $= channel keep mute state (10) 3); sound collection installed the &time; left channel single solution audio signal and = convert it into After the digital audio signal, by the central processor 2 = two 12 200920160 memory 4 〇 (S314); in the third round is more than _ rate of two 15) 曰 interface left and right sound 10 button team shoulder 1 talk (b315), the sound collection device 1 knows the wood when the right channel is set to a digital audio signal, by the middle f H, and converted 40 (S3l6Vy_ processing benefits 20 to save it to memory Body H ^ segment _T4, right channel input (four) frequency audio "Tiger" left channel round out mute state (s3i7); sound The right channel single frequency signal and the left γ wood in the period are collected and converted into a digital audio string 曰 '唬, the W1S in the genre 40 is stored by the central processing unit 2〇 to the record H); the woman coffee (10) is stored (4) 2 digits in each time period: frequency to the audio interface of the sounding device, and the parameters of the left and right channels (S319). Please refer to Figure *~ 圃8 for the specific method of each parameter. u. Figure 4 is a flow chart of the signal to noise ratio (Signal to 2, SNR) test of the audio interface of the sounding device. First, the central processing unit obtains the chewing sound levels NL and NR of the left and right channels according to the left and right sound signal signals of the first phase segment of the first phase segment according to the memory eight; the central processing unit 2求出 Calculating the left channel signal level % (read) according to the left channel early frequency audio signal collected at the second time T2 stored in the memory 40; the central processor 20 stores the memory according to the memory 4 在 at the fourth time The right channel early frequency audio signal collected by Duan Ding 4 finds the right channel signal level sr (s4g3); the central processor 2 求出 finds the left sound based on the left channel signal level SL and the left channel noise level NL The signal-to-noise ratio is SNR=朋帆(6). Based on the right channel signal level SR and the right channel noise level leg, the right channel signal-to-noise ratio is determined as ==2〇lg (V~(S404); The result is output to the display device 30 200920160 (S405) 0 See Figure 5 for the left and right channel crosstalk test flow chart of the audio interface of the sounding device. First, the + central processor 2 is stored according to the memory 4〇 at the second time. The left sound signal of the segment T2 and the crosstalk of the right channel determine the left channel signal level SL and the right channel string signal. Level CR (S5〇1); the central processor 20 stores the right channel single frequency audio signal and the left channel crosstalk signal 'the right channel signal' that is stored in the fourth time period T4 stored in the memory 40. Pingdi and left channel crosstalk signal level CL (S5f, central processor 2 求出 based on left channel crosstalk signal level CL and right stun signal level SR to find left channel crosstalk, based on right channel The crosstalk signal level CR and the left channel signal level % determine the right channel crosstalk lake 3); the test result is output through the display 30 (S504). See Fig. 6' for the total spectrum of the audio interface of the sounding device. Distortion (Harmonic Distortion 'THD) test flow chart. First, the central processor 20 collects the left channel single frequency audio signal collected in the second time period τ2 according to the memory 4〇 and collects in the fourth time period W. The right channel single frequency audio signal is separately subjected to Fourier transform to obtain the frequency domain signal of the left channel single frequency audio signal and the frequency domain signal of the right channel single frequency audio signal (S601). The frequency domain signal of the channel single solution audio signal is found to be left The amplitude of each harmonic component of the channel audio signal is determined according to the frequency domain signal of the right channel single frequency audio signal, and the amplitude W ( ) of each chopping component of the right channel audio signal is obtained;敎 component amplitude, 丨 left channel total spectrum wave 200920160 distortion is: THD = / (/ / L22 + HL32 + ... + HLN2) / (HLI2 + Hl22 + Hl32 + ... + HLN2), & The right harmonic is forced to the amplitude of each harmonic component of the audio signal, and the total harmonic distortion of the right channel is obtained. THD ~ Vc^2~+^i?32+... +HRNZ)/(HR\2 +ΗR22+ HΛ32+... +Hrn2) (C/ςπ^ ^ ^ , (S603), the test result is rotated to the display device 30 (S604). Please refer to Figure 7 for a full-scale distortion tester diagram of the audio interface of the sounding device. During the test, the central processing unit 2 performs the following steps: determining the left channel signal power according to the left and right channel single frequency audio signals stored in the second time period T2 and the fourth time period Τ4 stored in the memory 40. Ping and right sound number % flat (S701) 'Compare the left and right channel signal levels with the -preset' specification level (10) 2); if the left or right channel signal level and the preset specification level The difference is more than the default range, the left or right voice of the service, there is full-scale distortion, otherwise there is no full-scale distortion (S7G3); the test result is rounded to the display device 3〇 (S7〇4) . ', Please refer to Figure 8' for the frequency response of the sound interface of the sounding device. When testing, the central processing should be as follows: the left and right channels of the third time period T3 stored in the record = stored are the same. The length of the audio signal (10) 1); after the left-to-window processing, respectively, performing fast Fourier transform to obtain a plurality of frequency domain signals to obtain an average value of the left channel - the flat frequency domain signal of the left channel = 屮! 3); according to the left The average s of the right channel is the frequency response value of the left and right channels (S804); the result is dried out to the display device 30 (S805). Fruit round out 200920160
【圖式簡單說明】 圖1為音頻測試裝置第一實施方式的方塊圖。 圖2為發聲設備音頻介面測試方法中聲音播放過程的 示意圖。 圖3為發聲設備音頻介面的音頻測試方法的流程圖。 圖4為發聲設備音頻介面的信噪比測試流程圖。 圖5為發聲設備音頻介面的聲道串音测試流程圖。 圖6為發聲設備音頻介面的總諧波失真測試流程圖。 圖7為發聲設備音頻介面的滿格度失真測試流程圖。 圖8為發聲設備音頻介面的頻率響應測試流程圖。 【主要元件符號說明】 1 2 10 12 20 30 40 101 發聲設備 音頻測試裝置 聲音採集裝置 音頻線 中央處理器 顯示設備 記憶體 音頻介面 音頻介面 16 102BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a first embodiment of an audio test apparatus. 2 is a schematic diagram of a sound playing process in a sound interface device audio interface test method. 3 is a flow chart of an audio test method for an audio interface of a sounding device. Figure 4 is a flow chart of the signal to noise ratio test of the audio interface of the sounding device. Figure 5 is a flow chart of the channel crosstalk test of the audio interface of the sounding device. Figure 6 is a flow chart of the total harmonic distortion test of the audio interface of the sounding device. Figure 7 is a flow chart of the full-scale distortion test of the audio interface of the sounding device. Figure 8 is a flow chart of the frequency response test of the audio interface of the sounding device. [Main component symbol description] 1 2 10 12 20 30 40 101 Sounding device Audio test device Sound collection device Audio cable Central processing unit Display device Memory Audio interface Audio interface 16 102