TW202147300A

TW202147300A - Head-mounted apparatus and stereo effect controlling method thereof

Info

Publication number: TW202147300A
Application number: TW109119461A
Authority: TW
Inventors: 杜博仁; 張嘉仁; 曾凱盟
Original assignee: 宏碁股份有限公司
Priority date: 2020-06-10
Filing date: 2020-06-10
Publication date: 2021-12-16
Also published as: TWI746001B

Abstract

A head-mounted apparatus and a stereo effect controlling method thereof are provided. In the method, an expect received signal for receiving the output of the first speaker is determined, a signal difference between the expect received signal and an actual received signal is determined, an interference of the output signal of the second speaker on the actual received signal is estimated according to the signal difference, and the output signal of the first speaker is generated according to the interference of the output signal of the second speaker. The expected received signal is related to the received signal on the first side which is received by merely receiving the output signal of the first speaker. The actual received signal is the received signal on the first side which is received by actually receiving the output signals of the first and second speakers. The signal difference is minimized. Accordingly, the impact of the head for the output signal may be considered, and the signal quality may be further improved.

Description

Head mounted device and its stereo sound control method

本發明是有關於一種音訊處理技術，且特別是有關於一種頭戴式裝置及其立體音效控制方法。The present invention relates to an audio processing technology, and more particularly, to a head-mounted device and a method for controlling stereo sound effects.

市場上已經出現許多類型的頭戴式裝置(例如，數位眼鏡、頭戴式顯示器或虛擬實境(Virtual Reality，VR)頭戴套件等)。這些頭戴式裝置可能會配備左、右雙聲道的揚聲器。設於機體兩側的兩揚聲器各自播放一側聲道的聲音訊號。此雙聲道的聲音訊號原本會依據自由空間進行調整。例如，筆記型電腦或手機的揚聲器的訊號處理，或是兩個隔開耳機的訊號處理。一般而言，這些雙聲道的聲音訊號並沒有考慮人體及其頭部對聲音訊號造成的影響。然而，不同人的頭型和髮型不同。此外，頭部會造成部分的遮蔽效應，其中高頻訊號部分會被頭部阻隔，但低頻訊號部分可以穿越頭部，從而造成不同時間延遲，進而影響聲音品質。Many types of head-mounted devices (eg, digital glasses, head-mounted displays, or virtual reality (VR) headsets, etc.) have appeared on the market. These headsets may be equipped with left and right speakers. The two speakers arranged on both sides of the body play the sound signal of one channel respectively. This two-channel sound signal would have been scaled according to free space. For example, the signal processing of the speaker of a laptop or mobile phone, or the signal processing of two separate headphones. Generally speaking, these two-channel sound signals do not take into account the influence of the human body and its head on the sound signal. However, different people have different head shapes and hairstyles. In addition, the head will cause a partial shading effect, in which the high-frequency signal part will be blocked by the head, but the low-frequency signal part can pass through the head, resulting in different time delays, thereby affecting the sound quality.

有鑑於此，本發明提供一種頭戴式裝置及其立體音效控制方法，預先去除另一側揚聲器的輸出對一側收音的干擾，進而提升訊號品質。In view of this, the present invention provides a head-mounted device and a stereo sound effect control method thereof, which can preliminarily remove the interference of the output of the speaker on the other side to the sound reception on one side, thereby improving the signal quality.

本發明實施例的立體音效控制方法適用於包括第一揚聲器及第二揚聲器的頭戴式裝置，第一揚聲器設於頭戴式裝置的第一側，第二揚聲器設於頭戴式裝置的第二側。立體音效控制方法包括(但不僅限於)下列步驟：決定對第一揚聲器收音所得的期望接收訊號，決定期望接收訊號與實際接收訊號之間的訊號差異，依據訊號差異估測第二揚聲器的輸出訊號對實際接收訊號的干擾，並依據第二揚聲器的輸出訊號的干擾產生該第一揚聲器的輸出訊號。期望接收訊號相關於僅對第一揚聲器在第一側收音所得的接收訊號。實際接收訊號是實際對第一揚聲器及第二揚聲器在第一側收音所得。對訊號差異最小化。The stereo sound control method of the embodiment of the present invention is applicable to a head-mounted device including a first speaker and a second speaker, the first speaker is arranged on the first side of the head-mounted device, and the second speaker is arranged on the first side of the head-mounted device. two sides. The stereo sound control method includes (but is not limited to) the following steps: determining the expected received signal obtained by the first speaker, determining the signal difference between the expected received signal and the actual received signal, and estimating the output signal of the second speaker according to the signal difference The interference to the actual received signal generates the output signal of the first loudspeaker according to the interference of the output signal of the second loudspeaker. The desired received signal is related to the received signal obtained by only listening to the first loudspeaker on the first side. The actual received signal is obtained by actually receiving sound from the first speaker and the second speaker on the first side. Minimize signal variance.

本發明實施例的頭戴式裝置包括(但不僅限於)第一揚聲器、第二揚聲器、第一收音器、第二收音器及處理器。第一揚聲器設於頭戴式裝置的第一側，且第二揚聲器設於頭戴式裝置的第二側。第一收音器設於第一側，且第二收音器設於第二側。處理器耦接第一揚聲器、第二揚聲器、第一收音器及第二收音器。處理器載入並執行數個模組。這些模組包括差異決定模組、干擾估測模組及訊號合成模組。差異決定模組決定對第一揚聲器收音所得的期望接收訊號，並決定期望接收訊號與實際接收訊號之間的訊號差異。期望接收訊號相關於透過第一收音器僅對第一揚聲器在第一側收音所得的接收訊號，且實際接收訊號是實際透過第一收音器對第一揚聲器及第二揚聲器在第一側收音所得。干擾估測模組依據訊號差異估測第二揚聲器的輸出訊號對實際接收訊號的干擾，且對訊號差異最小化。訊號合成模組依據第二揚聲器的輸出訊號的干擾產生第一揚聲器的輸出訊號。The head-mounted device of the embodiment of the present invention includes (but is not limited to) a first speaker, a second speaker, a first sound receiver, a second sound receiver, and a processor. The first speaker is provided on the first side of the head-mounted device, and the second speaker is provided on the second side of the head-mounted device. The first receiver is arranged on the first side, and the second receiver is arranged on the second side. The processor is coupled to the first speaker, the second speaker, the first receiver and the second receiver. The processor loads and executes several modules. These modules include difference determination module, interference estimation module and signal synthesis module. The difference determining module determines the expected received signal obtained by the first speaker, and determines the signal difference between the expected received signal and the actual received signal. The expected received signal is related to the received signal obtained through the first receiver only to the first speaker on the first side, and the actual received signal is actually obtained through the first receiver to the first speaker and the second speaker on the first side. . The interference estimation module estimates the interference of the output signal of the second speaker to the actual received signal according to the signal difference, and minimizes the signal difference. The signal synthesis module generates the output signal of the first speaker according to the interference of the output signal of the second speaker.

基於上述，本發明實施例的頭戴式裝置及其立體音效控制方法，其對某一側的期望接收訊號及實際接收訊號之間的訊號差異最小化，以估測位於另一側的揚聲器的輸出對此側收音的干擾。接著，可依據此干擾調整此側揚聲器的輸出。藉此，可了解頭部對揚聲器的發聲所造成的影響，並能自適應地調整揚聲器的輸出，讓使用者體驗較為理想的聲音品質。Based on the above, the head-mounted device and the stereo sound control method thereof according to the embodiments of the present invention minimize the signal difference between the expected received signal and the actual received signal on one side, so as to estimate the sound effect of the speaker on the other side. Output interference to the radio on this side. Then, the output of the speaker on this side can be adjusted according to the interference. In this way, the influence of the head on the sound of the speaker can be understood, and the output of the speaker can be adjusted adaptively, so that the user can experience a more ideal sound quality.

為讓本發明的上述特徵和優點能更明顯易懂，下文特舉實施例，並配合所附圖式作詳細說明如下。In order to make the above-mentioned features and advantages of the present invention more obvious and easy to understand, the following embodiments are given and described in detail with the accompanying drawings as follows.

圖1是依據本發明一實施例的頭戴式裝置100的方塊圖。請參照圖1，頭戴式裝置100可以是數位眼鏡(或稱智慧眼鏡)、頭戴式顯示器(Head-Mounted Display，HMD)或其他供人類頭部配戴的電子裝置。頭戴式裝置100包括但不僅限於揚聲器110, 120、儲存器130、收音器140, 145及處理器150。FIG. 1 is a block diagram of a head-mounted device 100 according to an embodiment of the present invention. Referring to FIG. 1 , the head-mounted device 100 may be digital glasses (or smart glasses), a head-mounted display (HMD), or other electronic devices for human head wear. Head mounted device 100 includes, but is not limited to, speakers 110 , 120 , storage 130 , radios 140 , 145 and processor 150 .

揚聲器110, 120可以是喇叭或擴音器。在一實施例中，揚聲器110, 120分別設於頭戴式裝置100的第一側及第二側(例如，左、右兩側)，並分別對應到左、右兩聲道，以形成雙聲道揚聲器。Speakers 110, 120 may be speakers or loudspeakers. In one embodiment, the speakers 110 and 120 are respectively disposed on the first side and the second side (eg, left and right sides) of the head-mounted device 100, and correspond to the left and right channels respectively, so as to form a dual channel speakers.

儲存器130可以是任何型態的固定或可移動隨機存取記憶體(Radom Access Memory，RAM)、唯讀記憶體(Read Only Memory，ROM)、快閃記憶體(flash memory)、傳統硬碟(Hard Disk Drive，HDD)、固態硬碟(Solid-State Drive，SSD)或類似元件。在一實施例中，儲存器130用以記錄程式碼、軟體模組(例如，差異決定模組131、干擾估測模組133及訊號合成模組135等)、聲音訊號、訊號差異、權重、對聲音訊號的影響、誤差最小化函數及其他資料或檔案，其詳細內容待後續實施例詳述。The storage 130 may be any type of fixed or removable random access memory (RAM), read only memory (ROM), flash memory, conventional hard disks (Hard Disk Drive, HDD), Solid-State Drive (Solid-State Drive, SSD) or similar components. In one embodiment, the storage 130 is used to record code, software modules (eg, the difference determination module 131, the interference estimation module 133, the signal synthesis module 135, etc.), sound signals, signal differences, weights, The influence on the sound signal, the error minimization function and other data or files will be described in detail in the following embodiments.

收音器140, 145可以是動圈式(dynamic)、電容式(Condenser)、或駐極體電容(Electret Condenser)等類型的麥克風，收音器140, 145也可以是其他可接收聲波(例如，人聲、環境聲、機器運作聲等)而轉換為聲音訊號的電子元件、類比至數位轉換器、濾波器、及音訊處理器之組合。在一實施例中，收音器140, 145分別設於頭戴式裝置100的第一側及第二側(例如，左、右兩側)，使得揚聲器110與收音器140設於同側，且揚聲器120與收音器145設於同側。The microphones 140, 145 may be dynamic, condenser, or electret condenser (Electret Condenser) type microphones, and the microphones 140, 145 may also be other sound waves that can be received (eg, human voices). , ambient sound, machine operation sound, etc.), a combination of electronic components, analog-to-digital converters, filters, and audio processors that are converted into sound signals. In one embodiment, the microphones 140, 145 are respectively disposed on the first side and the second side (eg, left and right sides) of the head-mounted device 100, so that the speaker 110 and the microphone 140 are disposed on the same side, and The speaker 120 and the receiver 145 are arranged on the same side.

處理器150耦接揚聲器110, 120、儲存器130及收音器140, 145，處理器150並可以是中央處理單元(Central Processing Unit，CPU)，或是其他可程式化之一般用途或特殊用途的微處理器(Microprocessor)、數位信號處理器(Digital Signal Processor，DSP)、可程式化控制器、特殊應用積體電路(Application-Specific Integrated Circuit，ASIC)或其他類似元件或上述元件的組合。在一實施例中，處理器150用以執行頭戴式裝置100的所有或部份作業，且可載入並執行儲存器130所記錄的各軟體模組、檔案及資料。在一些實施例中，儲存器130所記錄的那些軟體模組也可能是實體電路所實現。The processor 150 is coupled to the speakers 110, 120, the storage 130 and the radios 140, 145, and the processor 150 can be a central processing unit (Central Processing Unit, CPU), or other programmable general-purpose or special-purpose Microprocessor (Microprocessor), digital signal processor (Digital Signal Processor, DSP), programmable controller, application-specific integrated circuit (Application-Specific Integrated Circuit, ASIC) or other similar components or a combination of the above components. In one embodiment, the processor 150 is used to execute all or part of the operations of the head-mounted device 100 , and can load and execute various software modules, files and data recorded in the storage 130 . In some embodiments, those software modules recorded in the storage 130 may also be implemented by physical circuits.

圖2是依據本發明一實施例的數位眼鏡200的示意圖。請參照圖2，數位眼鏡200的一側(例如，右側)設有揚聲器110與收音器140，且其另一側(例如，左側)設有揚聲器120與收音器145。使用者配戴頭戴式裝置100後，揚聲器110, 120及收音器140, 145鄰近於使用者的左、右耳。FIG. 2 is a schematic diagram of digital glasses 200 according to an embodiment of the present invention. Referring to FIG. 2 , one side (eg, the right side) of the digital glasses 200 is provided with the speaker 110 and the microphone 140 , and the other side (eg, the left side) is provided with the speaker 120 and the microphone 145 . After the user wears the head-mounted device 100, the speakers 110, 120 and the microphones 140, 145 are adjacent to the user's left and right ears.

須說明的是，圖2的處理器150設於數位眼鏡200的主體。然而，在一些實施例中，處理器150可能不與揚聲器110, 120及收音器140, 145設於相同的裝置，且元件之間的訊號傳輸可經由有線或無線通訊(例如，Wi-Fi、藍芽、或USB等)達成。It should be noted that the processor 150 shown in FIG. 2 is disposed on the main body of the digital glasses 200 . However, in some embodiments, the processor 150 may not be located on the same device as the speakers 110, 120 and the radios 140, 145, and the signal transmission between the components may be via wired or wireless communication (eg, Wi-Fi, Bluetooth, or USB, etc.).

下文中，將搭配頭戴式裝置100中的各項裝置、元件及模組說明本發明實施例所述之方法。本方法的各個流程可依照實施情形而隨之調整，且並不僅限於此。另須說明的是，為了方便說明，下文將以揚聲器110與收音器140設於頭戴式裝置100的一側(下文稱第一側)且揚聲器120與收音器145設於頭戴式裝置100的另一側(下文稱第二側)為例。Hereinafter, the method according to the embodiment of the present invention will be described in conjunction with various devices, components and modules in the head-mounted device 100 . Each process of the method can be adjusted according to the implementation situation, and is not limited to this. It should be noted that, for the convenience of description, the speaker 110 and the receiver 140 will be provided on one side of the head-mounted device 100 (hereinafter referred to as the first side), and the speaker 120 and the receiver 145 will be provided on the head-mounted device 100. the other side (hereinafter referred to as the second side) as an example.

圖3是依據本發明一實施例的立體音效控制方法的流程圖。請參照圖3，差異決定模組131決定對揚聲器110收音所得的期望接收訊號(步驟S310)。具體而言，期望接收訊號相關於透過收音器140僅對揚聲器110在第一側(即，同側)收音所得的接收訊號。本發明實施例的收音器140, 145是用於估測使用者聽到揚聲器110, 120所發出的聲音。期望接收訊號代表期望使用者的某一耳朵僅聽到同側揚聲器的聲音。FIG. 3 is a flowchart of a method for controlling stereo sound according to an embodiment of the present invention. Referring to FIG. 3 , the difference determination module 131 determines the expected received signal obtained by the speaker 110 (step S310 ). Specifically, the desired received signal is related to the received signal obtained by only picking up the speaker 110 on the first side (ie, the same side) through the microphone 140 . The microphones 140 and 145 of the embodiment of the present invention are used to estimate the sound that the user hears from the speakers 110 and 120 . The desired reception signal means that one ear of the user is expected to hear only the sound from the same side speaker.

無可避免地，揚聲器110的輸出訊號可能會經頻率響應增益及/或時間延遲等變化/影響因素而到達收音器140(對應到使用者的一耳)。於此，將對揚聲器110的輸出訊號透過收音器140在第一側收音所遭受的變化統稱為第一傳播介質(medium)(包括空氣、及/或人體等介質)的影響結果，且收音器140的期望接收訊號是揚聲器110的輸出訊號經過第一傳播介質所形成。相似地，揚聲器120的輸出訊號也可能會經頻率響應增益及/或時間延遲等變化/影響因素而到達收音器145(對應到使用者的另一耳)。於此，將對揚聲器120的輸出訊號透過收音器145在第二側(即，同側)收音所遭受的變化統稱為第二傳播介質的影響結果，且收音器145的期望接收訊號是揚聲器120的輸出訊號經過第二傳播介質所形成。Inevitably, the output signal of the speaker 110 may reach the receiver 140 (corresponding to one ear of the user) through changes/influence factors such as frequency response gain and/or time delay. Herein, the changes suffered by the output signal of the speaker 110 through the receiver 140 on the first side of the sound collection are collectively referred to as the effect of the first medium (including air, and/or the human body, etc.), and the receiver The desired received signal of 140 is formed by the output signal of the speaker 110 through the first propagation medium. Similarly, the output signal of the speaker 120 may also reach the receiver 145 (corresponding to the user's other ear) through varying/influencing factors such as frequency response gain and/or time delay. Herein, the changes suffered by the output signal of the speaker 120 to be received on the second side (ie, the same side) through the receiver 145 are collectively referred to as the effect of the second propagation medium, and the expected signal received by the receiver 145 is the speaker 120 The output signal is formed through the second propagation medium.

以圖2為例，

為揚聲器110對應於收音器140的第一傳播介質的頻率響應，

為揚聲器120對應於收音器145的第二傳播介質的頻率響應。在本實施例中，

、

是以頻域上的響應來分別代表第一及第二傳播介質，且其值可事先設計，並可依據設計者的需求而變化。而收音器145的期望接收訊號

的數學表示式如下：

…(1)

為揚聲器120的原始輸出訊號。收音器140的期望接收訊號可依此類推，於此不再贅述。Take Figure 2 as an example,

is the frequency response of the speaker 110 corresponding to the first propagation medium of the receiver 140,

is the frequency response of the second propagation medium for the speaker 120 corresponding to the receiver 145 . In this embodiment,

,

The first and second propagation media are respectively represented by the responses in the frequency domain, and their values can be designed in advance and can be changed according to the designer's needs. And the desired reception signal of the radio 145

The mathematical expression is as follows:

…(1)

is the original output signal of the speaker 120 . The expected received signal of the microphone 140 can be deduced by analogy, and details are not repeated here.

差異決定模組131可透過收音器140對揚聲器110, 120在第一側收音所得的實際接收訊號，並透過收音器145對揚聲器110, 120在第二側收音所得的實際接收訊號。接著，差異決定模組131分別決定兩側的期望接收訊號與對應側的實際接收訊號之間的訊號差異(步驟S330)。具體而言，無可避免地，收音器140可能會接收到來自揚聲器120所發出的聲音，且收音器145也可能會接收到來自揚聲器110所發出的聲音。然而，另一側所發出的聲音都不是希望被聽到的聲音。由於期望接收訊號是僅有同側的輸出訊號被接收，因此期望接收訊號與對應側的實際接收訊號之間的訊號差異越小越符合期望。The difference determination module 131 can collect the actual received signal obtained by the receiver 140 from the speakers 110 and 120 on the first side, and can use the receiver 145 to receive the actual received signal from the speakers 110 and 120 on the second side. Next, the difference determination module 131 respectively determines the signal difference between the expected received signal on the two sides and the actual received signal on the corresponding side (step S330 ). Specifically, it is inevitable that the receiver 140 may receive the sound from the speaker 120 , and the receiver 145 may also receive the sound from the speaker 110 . However, none of the sounds coming from the other side were the ones that were expected to be heard. Since the expected received signal is only received on the same side of the output signal, the smaller the signal difference between the expected received signal and the actual received signal on the corresponding side, the better the expectation.

相似地，揚聲器120的輸出訊號可能會經頻率響應增益及/或時間延遲等變化/影響因素而到達收音器140。於此，將對揚聲器120的輸出訊號透過收音器140在第一側收音所遭受的變化統稱為第三傳播介質的影響結果，且揚聲器120的輸出訊號經過此第三傳播介質將形成收音器140的干擾(即，不是希望被聽到的聲音)。此外，揚聲器110的輸出訊號也可能會經頻率響應增益及/或時間延遲等變化/影響因素而到達收音器145(對應到使用者的另一耳)。於此，將對揚聲器110的輸出訊號透過收音器145在第二側收音所遭受的變化統稱為第四傳播介質的影響結果，且揚聲器110的輸出訊號經過此第四傳播介質將形成收音器145的干擾。Similarly, the output signal of the loudspeaker 120 may reach the receiver 140 through varying/affecting factors such as frequency response gain and/or time delay. Here, the changes suffered by the output signal of the speaker 120 through the receiver 140 on the first side are collectively referred to as the effect of the third propagation medium, and the output signal of the speaker 120 passes through the third propagation medium to form the receiver 140 interference (that is, sounds that are not intended to be heard). In addition, the output signal of the speaker 110 may also reach the receiver 145 (corresponding to the other ear of the user) through changes/influence factors such as frequency response gain and/or time delay. Herein, the changes suffered by the output signal of the speaker 110 through the receiver 145 on the second side are collectively referred to as the effect of the fourth propagation medium, and the output signal of the loudspeaker 110 passes through the fourth propagation medium to form the receiver 145 interference.

以圖2為例，

為揚聲器120對應於收音器140的第三傳播介質的頻率響應，

為揚聲器110對應於收音器145的第四傳播介質的頻率響應。在本實施例中，

、

是以頻域上的響應來分別代表第三及第四傳播介質。與第一及第二傳播介質不同之處在於，不同使用者的頭部可能形成不同的第三及第四傳播介質的頻率響應，因此第三及第四傳播介質的頻率響應初始視為未知。而第二側的訊號差異

的數學表示式如下：

…(2)Take Figure 2 as an example,

is the frequency response of the speaker 120 corresponding to the third propagation medium of the receiver 140,

is the frequency response of the fourth propagation medium for the speaker 110 corresponding to the receiver 145 . In this embodiment,

,

are the responses in the frequency domain to represent the third and fourth propagation media, respectively. The difference from the first and second propagation media is that different users' heads may form different frequency responses of the third and fourth propagation media, so the frequency responses of the third and fourth propagation media are initially regarded as unknown. And the signal difference on the second side

The mathematical expression is as follows:

…(2)

為收音器145的實際接收訊號。第一側的訊號差異可依此類推，於此不再贅述。

is the actual received signal of the radio 145 . The signal difference of the first side can be deduced by analogy, and details are not repeated here.

干擾估測模組133可依據訊號差異估測揚聲器110的輸出訊號對第二側的實際接收訊號的干擾，並可依據另一訊號差異估測揚聲器120的輸出訊號對第一側的實際接收訊號的干擾(步驟S350)。具體而言，為了將訊號差異最小化，干擾估測模組133可先估測第三及第四傳播介質(如圖2的

、

)。在一實施例中，干擾估測模組133可採用權重更新演算法來近似第三及第四傳播介質，以對左側消除右聲道輸出的相關訊號或對右側消除左聲道輸出的相關訊號。The interference estimation module 133 can estimate the interference of the output signal of the speaker 110 on the actual received signal on the second side according to the signal difference, and can estimate the actual received signal on the first side from the output signal of the speaker 120 according to another signal difference interference (step S350). Specifically, in order to minimize the signal difference, the interference estimation module 133 may first estimate the third and fourth propagation media (as shown in FIG. 2 ).

,

). In one embodiment, the interference estimation module 133 may use a weight update algorithm to approximate the third and fourth propagation media to cancel the correlation signal output by the right channel for the left side or cancel the correlation signal output by the left channel for the right side. .

干擾估測模組133可依據揚聲器110, 120的輸出訊號及對應側的訊號誤差分別決定權重更新演算法中的權重，此權重分別相關於前述第三及第四傳播介質的頻率響應，且最終的權重對應到最小化的訊號誤差(即，權重更新演算法是用於最小化訊號誤差)。The interference estimation module 133 can respectively determine the weights in the weight update algorithm according to the output signals of the speakers 110 and 120 and the signal errors of the corresponding sides. The weights are respectively related to the frequency responses of the third and fourth propagation media, and finally The weights correspond to the minimized signal error (ie, the weight update algorithm is used to minimize the signal error).

權重更新演算法有很多種。以最小均方(Least Mean Square，LMS)為例，其時域的表示為：

…(3) ，其中

代表了更新的間格大小(其值為常數，並代表更新步階大小，或稱步長)，

影響了收斂(例如，訊號差異趨近於特定數值)的準確度和速度。此外，

及

分別代表不同取樣時間點對應的權重(例如，n +1是n 的下一個取樣時間點，並對應到頻域的權重

)，且

是揚聲器110的輸出訊號。方程式(3)代表第二側對應的權重更新演算法，而第一側對應的權重更新演算法可依此類推，於此不再贅述。There are many kinds of weight update algorithms. Taking Least Mean Square (LMS) as an example, its time domain representation is:

…(3) , where

Represents the update interval size (its value is constant and represents the update step size, or step size),

Affects the accuracy and speed of convergence (eg, signal differences approaching a specific value). also,

and

respectively represent the weights corresponding to different sampling time points (for example, n + 1 is the next sampling time point of n and corresponds to the weight in the frequency domain

),and

is the output signal of the speaker 110 . Equation (3) represents the weight update algorithm corresponding to the second side, and the weight update algorithm corresponding to the first side can be deduced in the same way, which will not be repeated here.

圖4是依據本發明一實施例的訊號關係的示意圖。請參照圖4，第二側的訊號差異

及揚聲器110的輸出訊號

輸入到權重更新演算法LE_L 即可得出權重

。相似地，第一側的訊號差異

及揚聲器120的輸出訊號

輸入到權重更新演算法LE_R 即可得出權重

。4 is a schematic diagram of a signal relationship according to an embodiment of the present invention. Please refer to Figure 4, the signal difference of the second side

and the output signal of the speaker 110

Input to the weight update algorithm LE _L to get the weight

. Similarly, the signal difference on the first side

and the output signal of the speaker 120

Input to the weight update algorithm LE _R to get the weight

.

透過不停更新對於欲消除訊的權重

，且使用訊號差異

輔助更新演算法，當訊號差異收斂時，會使得揚聲器110的輸出訊號經過此權重

後會接近揚聲器110對應於收音器145的第四傳播介質的頻率響應：

…(4) 相似地，第一側對應的權重更新演算法所估測的權重

也可用於逼近揚聲器120對應於收音器140的第三傳播介質的頻率響應。揚聲器110的輸出訊號

對第二側的實際接收訊號

的干擾即是

，且揚聲器120的輸出訊號

對第一側的實際接收訊號

的干擾即是

。而第一及第二傳播介質的頻率響應(對應到

、

)及權重

、

此時皆已知，對應干擾也可據以得出。By constantly updating the weight of the message to be eliminated

, and use the signal difference

The auxiliary update algorithm, when the signal difference converges, will make the output signal of the speaker 110 pass through this weight

The frequency response of the fourth propagation medium of the speaker 110 corresponding to the receiver 145 will then be approached:

...(4) Similarly, the corresponding weights on the first side update the weights estimated by the algorithm

It can also be used to approximate the frequency response of the loudspeaker 120 corresponding to the third propagation medium of the receiver 140 . The output signal of the speaker 110

The actual received signal to the second side

interference is

, and the output signal of the speaker 120

The actual received signal on the first side

interference is

. while the frequency responses of the first and second propagation media (corresponding to

,

) and weight

,

All are known at this time, and the corresponding interference can also be derived accordingly.

須說明的是，在其他實施例中，權重更新演算法也可能是基於訊號差異(作為誤差)的誤差最佳化(最小化)演算法。例如，最小均方(Least Square，LS)、或最小均方誤差估測演算法(Minimum Mean Square Error，MMSE)等演算法。It should be noted that, in other embodiments, the weight update algorithm may also be an error optimization (minimization) algorithm based on signal differences (as errors). For example, algorithms such as Least Square (LS) or Minimum Mean Square Error (Minimum Mean Square Error, MMSE) are used.

請回到圖3，訊號合成模組135可依據揚聲器120的輸出訊號的干擾產生揚聲器110的輸出訊號，或依據揚聲器110的輸出訊號的干擾產生揚聲器120的輸出訊號(步驟S370)。具體而言，本發明實施例是預先對輸出訊號去除估測的干擾(基於訊號合成處理)，使最後實際接收訊號即便受到另一側輸出訊號的干擾也能接近或等同於期望接收訊號。3, the signal synthesis module 135 can generate the output signal of the speaker 110 according to the interference of the output signal of the speaker 120, or generate the output signal of the speaker 120 according to the interference of the output signal of the speaker 110 (step S370). Specifically, the embodiment of the present invention removes the estimated interference from the output signal in advance (based on signal synthesis processing), so that the final actual received signal can be close to or equal to the expected received signal even if it is interfered by the output signal on the other side.

在一實施例中，訊號合成模組135可依據兩權重及揚聲器110, 120的輸出訊號分別決定對應的預先消除訊號。請參照圖4，揚聲器110對應的預先消除訊號即是揚聲器120的輸出訊號與對應權重在頻域上的乘積(即，

)，且揚聲器120對應的預先消除訊號即是揚聲器110的輸出訊號與對應權重在頻域上的乘積(即，

)。In one embodiment, the signal synthesis module 135 can determine the corresponding pre-elimination signal according to the two weights and the output signals of the

speakers

110 and 120, respectively. Referring to FIG. 4 , the pre-elimination signal corresponding to the speaker 110 is the product of the output signal of the speaker 120 and the corresponding weight in the frequency domain (ie,

), and the pre-elimination signal corresponding to the speaker 120 is the product of the output signal of the speaker 110 and the corresponding weight in the frequency domain (ie,

).

訊號合成模組135可對揚聲器110, 120的輸出訊號分別移除對應的預先消除訊號，以產生揚聲器110, 120的最終輸出訊號。請參照圖4，揚聲器110, 120的最終輸出訊號

、

的數學表示式如下：

…(5)

…(6) 。即，訊號合成模組135依據揚聲器120在先前時間點的輸出訊號

決定當前時間點的原始輸出訊號

對應的預先消除訊號

，從而得出當前時間的最終輸出訊號

，其中先前時間早於當前時間點(例如是相差一個取樣點數)。另一方面，訊號合成模組135依據揚聲器110在先前時間點的輸出訊號

決定當前時間點的原始輸出訊號

對應的預先消除訊號

，從而得出當前時間的最終輸出訊號

。The signal synthesis module 135 can remove the corresponding pre-elimination signals from the output signals of the

speakers

110 and 120 respectively, so as to generate the final output signals of the

speakers

110 and 120 . Please refer to FIG. 4, the final output signal of the

speakers

110, 120

,

The mathematical expression is as follows:

…(5)

…(6). That is, the signal synthesis module 135 is based on the output signal of the speaker 120 at the previous time point

Determine the original output signal at the current time point

Corresponding pre-cancellation signal

, so as to obtain the final output signal of the current time

, where the previous time is earlier than the current time point (eg, by one sampling point). On the other hand, the signal synthesis module 135 is based on the output signal of the speaker 110 at the previous time point

Determine the original output signal at the current time point

Corresponding pre-cancellation signal

, so as to obtain the final output signal of the current time

.

此時，收音器140的實際接收訊號即是：

…(7) 而收音器145的實際接收訊號即是：

…(8) 。藉此，實際接收訊號將趨近或等同於預設的期望接收訊號，從而消除另一側輸出的干擾。At this time, the actual received signal of the radio 140 is:

...(7) The actual received signal of the radio 145 is:

…(8) . In this way, the actual received signal will approach or be equal to the preset expected received signal, thereby eliminating the interference output from the other side.

在實際應用上，此立體音效控制方法可在使用者的頭部戴上頭戴式裝置100後執行(例如，反應於使用者的觸發行為或額外感測器的偵測結果)，從而得出當前使用者頭部對揚聲器110, 120的輸出訊號所造成的影響，進而得出客製化的訊號調整。In practical applications, the stereo sound control method can be executed after the user wears the head-mounted device 100 (eg, in response to the user's triggering behavior or the detection result of an additional sensor), thereby obtaining The current impact of the user's head on the output signal of the speakers 110, 120, and then the customized signal adjustment can be obtained.

綜上所述，在本發明實施例的頭戴式裝置及其立體音效控制方法中，基於將期望接收訊號及實際接收訊號之間的訊號差異視為誤差，並基於最小化誤差的目的來估測來自另一側輸出對應的權重。權重與另一側的輸出訊號即可用於估測來自另一側的干擾。此外，對所欲輸出的訊號提前去除干擾，使實際接收訊號能趨近或等於期望接收訊號。藉此，可排除頭部對頭戴式裝置的聲音輸出所造成的影響，並使頭戴式裝置自動調整而讓使用者獲得較為理想的聲音品質。To sum up, in the head-mounted device and the stereo sound control method thereof according to the embodiments of the present invention, the signal difference between the expected received signal and the actual received signal is regarded as an error, and the estimation is based on the purpose of minimizing the error. Measure the weight corresponding to the output from the other side. The weights and the output signal of the other side can be used to estimate the interference from the other side. In addition, the interference is removed in advance for the signal to be output, so that the actual received signal can approach or be equal to the expected received signal. In this way, the influence of the head on the sound output of the head-mounted device can be excluded, and the head-mounted device can be automatically adjusted so that the user can obtain relatively ideal sound quality.

雖然本發明已以實施例揭露如上，然其並非用以限定本發明，任何所屬技術領域中具有通常知識者，在不脫離本發明的精神和範圍內，當可作些許的更動與潤飾，故本發明的保護範圍當視後附的申請專利範圍所界定者為準。Although the present invention has been disclosed above by the embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the technical field can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, The protection scope of the present invention shall be determined by the scope of the appended patent application.

100:頭戴式裝置 110、120:揚聲器 130:儲存器 131:差異決定模組 133:干擾估測模組 135:訊號合成模組 140、145:收音器 150:處理器 200:數位眼鏡

:第一傳播介質的頻率響應

:第二傳播介質的頻率響應

:第三傳播介質的頻率響應

:第四傳播介質的頻率響應 S310~S370:步驟

、

:實際接收訊號

、

:原始輸出訊號

:期望接收訊號

、

:權重 LE_R 、LE_L :權重更新演算法

、

:輸出訊號100: Head mounted device 110, 120: Speaker 130: Storage 131: Difference determination module 133: Interference estimation module 135: Signal synthesis module 140, 145: Radio 150: Processor 200: Digital glasses

: Frequency response of the first propagation medium

: Frequency response of the second propagation medium

: Frequency response of the third propagation medium

: Frequency response of the fourth propagation medium S310~S370: Steps

,

: Actual received signal

,

: original output signal

: Expect to receive a signal

,

: Weight LE _R , LE _L : Weight update algorithm

,

: output signal

圖1是依據本發明一實施例的頭戴式裝置的方塊圖。圖2是依據本發明一實施例的頭戴式裝置的示意圖。圖3是依據本發明一實施例的立體音效控制方法的流程圖。圖4是依據本發明一實施例的訊號關係的示意圖。FIG. 1 is a block diagram of a head-mounted device according to an embodiment of the present invention. FIG. 2 is a schematic diagram of a head-mounted device according to an embodiment of the present invention. FIG. 3 is a flowchart of a method for controlling stereo sound according to an embodiment of the present invention. 4 is a schematic diagram of a signal relationship according to an embodiment of the present invention.

S310~S370:步驟S310~S370: Steps

Claims

A stereo sound control method is suitable for a head-mounted device including a first speaker and a second speaker, the first speaker is arranged on a first side of the head-mounted device, and the second speaker is arranged on the head A second side of the wearable device, and the stereo sound control method includes: determining a desired received signal from the first speaker, wherein the desired received signal is related to a received signal from only the first loudspeaker on the first side; determining a signal difference between the expected received signal and an actual received signal, wherein the actual received signal is actually obtained from the first speaker and the second speaker on the first side; estimating the interference of the output signal of the second speaker to the actual received signal according to the signal difference, wherein the signal difference is minimized; and The output signal of the first speaker is generated according to the interference of the output signal of the second speaker.

The stereo sound control method according to claim 1, wherein the step of estimating the interference of the output signal of the second speaker to the actual received signal according to the signal difference comprises: A weight is determined according to the output signal of the second speaker and the signal error, wherein the weight is related to a first propagation medium through which the output signal of the second speaker is received on the first side, and the output of the second speaker The interference is formed by a signal passing through the first propagation medium, and the weight corresponds to the minimized error of the signal.

The stereo sound control method according to claim 2, wherein the desired received signal is formed by the output signal of the first speaker passing through a second propagation medium, and the second propagation medium is related to the output signal of the first speaker in the The step of generating the output signal of the first loudspeaker according to the disturbance of the output signal of the second loudspeaker according to the change suffered by the first side of the sound reception includes: determining a pre-cancellation signal according to the weight and the output signal of the second speaker, wherein the frequency response of the first propagation medium is determined according to the weight and the frequency response of the second propagation medium; and The pre-cancellation signal is removed from the output signal of the first speaker to generate a final output signal of the first speaker, wherein the final output signal is played through the first speaker.

The stereo sound control method as claimed in claim 3, wherein the step of determining the pre-elimination signal comprises: The pre-elimination signal at a current time point is determined according to the output signal of the second speaker at a previous time point, wherein the previous time is earlier than the current time point.

The stereo sound effect control method according to claim 1, further comprising: The output signal of the second speaker is generated according to the second disturbance of the output signal of the first speaker, wherein the second disturbance is related to the expected reception signal of the second speaker on the second side and the actual signal at the second speaker The difference between the received signal from the side radio.

A head-mounted device comprising: a first speaker disposed on a first side of the head-mounted device; a second speaker disposed on a second side of the head-mounted device; a first microphone, located on the first side; a second microphone disposed on the second side; and A processor, coupled to the first speaker, the second speaker, the first receiver and the second receiver, loads and executes a plurality of modules, the modules include: A difference determination module determines an expected received signal obtained from the first speaker, and determines a signal difference between the expected received signal and an actual received signal, wherein the expected received signal is related to the first received signal the received signal obtained by the receiver only from the first speaker on the first side, and the actual received signal is actually obtained from the first speaker and the second speaker on the first side through the first receiver; an interference estimation module for estimating the interference of the output signal of the second speaker to the actual received signal according to the signal difference, wherein the signal difference is minimized; and A signal synthesis module generates the output signal of the first speaker according to the interference of the output signal of the second speaker.

The head-mounted device of claim 6, wherein the interference estimation module determines a weight according to the output signal of the second speaker and the signal error, wherein the weight is related to the second speaker through the first receiver A first propagation medium through which the output signal of the loudspeaker is collected on the first side, and the output signal of the second loudspeaker passes through the first propagation medium to form the interference, and the weight corresponds to the minimized signal error.

The head-mounted device of claim 7, wherein the desired received signal is formed by the output signal of the first speaker passing through a second propagation medium, and the second propagation medium is related to the output signal of the first speaker through the The change suffered by the first receiver on the first side of the sound collection, and the signal synthesis module determines a pre-elimination signal according to the weight and the output signal of the second speaker, and removes the signal from the output signal of the first speaker. cancel the signal in advance to generate the final output signal of the first speaker, wherein the interference estimation module determines the frequency response of the first propagation medium according to the weight and the frequency response of the second propagation medium, and the signal synthesis module The final output signal is played through the first speaker.

The head-mounted device of claim 7, wherein the signal synthesis module determines the pre-elimination signal at a current time point according to the output signal of the second speaker at a previous time point, wherein the previous time is earlier than the current time point in time.

The head-mounted device of claim 6, wherein the signal synthesis module generates the output signal of the second speaker according to a second disturbance of the output signal of the first speaker, wherein the second disturbance is related to passing through the second speaker The difference between an expected received signal of the second speaker on the second side and the received signal actually obtained by the second speaker on the second side.