Sound quality evaluation methods and sound quality evaluation systems using them

Abstract

This application provides a sound quality evaluation method and a sound quality evaluation system using the same. The sound quality evaluation system records a test audio file played by multiple playback devices to generate multiple audio data, and then segments the audio data into multiple frequency bands. The sound quality evaluation system calculates the frequency bands to obtain multiple evaluation scores for the playback devices. The sound quality evaluation system retrieves sound quality ranking information for the corresponding playback devices from a reference source, and adjusts the evaluation scores based on the sound quality ranking information to obtain a reference model. The sound quality evaluation system adjusts the evaluation scores of the multiple playback devices under test according to the reference model to obtain sound quality ranking information for the playback devices under test.

Description

Technical Field

[0001] This application relates to a method and system for evaluating the sound quality of a playback device. Background Technology

[0002] When purchasing media players, consumers typically listen to the music they hear to determine their preferred product. Most online product analyses rely on the subjective feelings of the analysts to evaluate media players. In other words, there is currently no objective and accurate method for analyzing the performance of media players. Furthermore, due to the influence of subjective feelings, everyone's evaluation of the same product can vary greatly. This situation makes it difficult for consumers to choose a suitable product from subjective media player rankings. Summary of the Invention

[0003] The technical problem to be solved by this application is to provide a sound quality evaluation method for providing sound quality ranking information of multiple playback devices, comprising: defining these playback devices as a first group and a second group; recording at least one test audio file played by the first group and the second group respectively to generate multiple first audio data and multiple second audio data; segmenting each first audio data and each second audio data respectively to generate multiple first group frequency bands and multiple second group frequency bands; calculating and processing the first group frequency bands and the second group frequency bands respectively to obtain multiple first evaluation scores of the first group and multiple second evaluation scores of the second group; extracting a first sound quality ranking information corresponding to the first group from a reference source; adjusting the first evaluation score accordingly based on the first sound quality ranking information to obtain a first reference model; and adjusting the second evaluation score accordingly based on the first reference model to obtain a second sound quality ranking information of the second group.

[0004] This application also discloses a sound quality evaluation system, comprising an audio recording module, a calculation module, a communication module, and a processing module. The audio recording module defines multiple playback devices as a first group and a second group, and records at least one test audio file played by each of the first and second groups to generate multiple first audio data and multiple second audio data. The calculation module segments each of the first and second audio data to generate multiple first group frequency bands and multiple second group frequency bands, and calculates and processes each of the first and second group frequency bands to obtain multiple first evaluation scores for the first group and multiple second evaluation scores for the second group. The communication module retrieves a first sound quality ranking information corresponding to the first group from a reference source. The processing module adjusts the first evaluation scores accordingly based on the first sound quality ranking information to obtain a first reference model, and adjusts the second evaluation scores accordingly based on the first reference model to obtain a second sound quality ranking information for the second group.

[0005] The sound quality evaluation method and system of this application train an objective sound quality evaluation algorithm model by referring to the sound quality ranking information of audio devices published in one or more publicly available online databases, without requiring the intervention of acoustic experts during the training process. As training data accumulates, the evaluation scores calculated by the sound quality evaluation model of this application not only closely approximate the judgments of acoustic experts, but also completely avoid the evaluation biases caused by occasional changes in the physiological and psychological conditions of acoustic experts during evaluation. Therefore, it can evaluate the sound quality of various playback devices more objectively and consistently than acoustic experts, thus providing an objective and accurate evaluation method and conveniently allowing users to understand the performance of playback devices.

[0006] The other effects and embodiments of this application are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic diagram of the implementation environment of the sound quality evaluation method in one embodiment of this application;

[0009] Figure 2 This is a schematic diagram of a sound quality evaluation system performing a sound quality evaluation method in one embodiment of this application;

[0010] Figure 3 This is an example flowchart of the sound quality evaluation method of this application. Detailed Implementation

[0011] In the embodiments described below, the positional relationships include: up, down, left, and right. Unless otherwise specified, they are all based on the direction shown by the components in the diagram.

[0012] like Figure 1 As shown, in some embodiments, the sound quality evaluation method is implemented in a listening room 10, which is a space for evaluating electroacoustic products and loudspeakers as defined by the European Telecommunications Standards Institute (ETSI) and the International Electron Technology Association (IEC). The listening room 10 includes a playback device under test 110, an artificial head device 120, and a computer host 130. The computer host 130 is located next to the artificial head device 120 and is electrically connected to the artificial head device 120.

[0013] Reference Figure 2 and Figure 3 , Figure 2 This is a schematic diagram of a sound quality evaluation method performed by the sound quality evaluation system 210 in one embodiment of this application. Figure 3 This is an example flowchart of the sound quality evaluation method of this application. The sound quality evaluation system 210 can be used to execute the sound quality evaluation method and includes an audio recording module 211, a calculation module 212, a communication module 213, and a processing module 214, wherein the audio recording module 211 is electrically connected to the calculation module 212, the calculation module 212 is electrically connected to the processing module 214, and the processing module 214 is electrically connected to the communication module 213.

[0014] like Figure 3 As shown, after defining the multiple playback devices 200 as the first group 201 and the second group 202 (step S10), the audio recording module 211 of the sound quality evaluation system 210 will record at least one test audio file played by the playback devices of the first group 201 and the playback devices of the second group 202 respectively, so as to generate multiple first audio data and multiple second audio data (step S20).

[0015] In one embodiment, the sound quality evaluation system 210 may be a mobile phone, a tablet computer, or a personal computer.

[0016] In one embodiment, the audio recording module 211 is an artificial head device 120, wherein the artificial head device 120 is a microphone that simulates the structure of a human ear, and its function is to simulate the human ear to receive audio data in order to analyze the impact of the structure of various parts of the human body on human hearing.

[0017] In one embodiment, the playback device 200 can be any type of speaker, audio system, mobile phone, tablet, or personal computer.

[0018] In one embodiment, the test file can be an audio file of any audio file format, such as an MP3 file, WAV file, AAC file, or FLAC file. The audio recording module 211 records the test file as audio data in a fixed audio format.

[0019] The calculation module 212 of the sound quality evaluation system 210 segments each first audio data and each second audio data to generate multiple first group frequency bands and multiple second group frequency bands (step S30). The frequencies of these frequency bands are in the range of 100Hz to 22kHz, which is the frequency range of sound that is generally audible to the human ear. Segmenting each audio data into multiple frequency bands is used to extract the sound frequencies that are audible to the human ear and to filter out those sound frequencies that are not audible to the human ear.

[0020] In one embodiment, the computing module 212 may be a central processing unit (CPU), a graphics processing unit (GPU), or a computing unit with computing capabilities.

[0021] In one embodiment, the calculation module 212 divides each first audio data and each second audio data into multiple frequency bands, such as, but not limited to, 26 frequency bands.

[0022] After the calculation module 212 has divided each first audio data and each second audio data into multiple first group frequency bands and multiple second group frequency bands, it will continue to calculate and process these first group frequency bands and these second group frequency bands respectively to obtain multiple first evaluation scores of the first group 201 and multiple second evaluation scores of the second group 202 (step S40).

[0023] The calculation module 212 uses a machine learning algorithm and a sound quality evaluation algorithm model to calculate the first group frequency bands and the second group frequency bands to obtain the first evaluation score and the second evaluation score. The first evaluation score is the sound quality performance of the playback device in the first group 201, and the second evaluation score is the sound quality performance of the playback device in the second group 202. The higher the evaluation score, the better the sound quality performance of the playback device.

[0024] In one embodiment, the machine learning algorithm is a gradient descent method, wherein the formula for gradient descent is x. t+1 =x t -γ×Δf(x t f(x) is a sound quality evaluation function (i.e., a sound quality evaluation algorithm model), x is the energy of each frequency band of the first audio data, γ is a learning rate, Δf is a target score, and t is the number of updates. The initial model of the sound quality evaluation algorithm model adopts a random initial reference model. Those skilled in the art to which this application pertains are familiar with the initial reference model, so it will not be described in detail.

[0025] The learning rate refers to the update increment during each update, and its value needs to be adjusted gradually during the update process. In this embodiment, the learning rate is between 0.001 and 0.002, and the adjustment increment of the learning rate is between 0.00001 and 0.0001.

[0026] The communication module 213 of the sound quality evaluation system 210 retrieves the first sound quality ranking information 221 of the playback device corresponding to the first group 201 from a reference source 220 (step S50), wherein the communication module 213 can connect to the reference source 220 via a wired network or a wireless network.

[0027] In one embodiment, reference source 220 is a publicly available online database containing audio quality ranking information for multiple playback devices 200 of various models. For example, the communication module 213 of the audio quality evaluation system 210 can retrieve audio quality ranking information for multiple playback devices 200 of various models from a mobile phone review website.

[0028] After the first sound quality ranking information 221 is extracted, the processing module 214 of the sound quality evaluation system 210 will refer to the first sound quality ranking information 221 to adjust the first evaluation score accordingly, thereby obtaining a first reference model (step S60).

[0029] The processing module 214 adjusts one parameter in the initial reference model to a first parameter to obtain the first reference model, so that the order of these first evaluation scores, after being calculated by the machine learning algorithm and the first reference model, will match the first sound quality ranking information 221, that is, make the order of these first evaluation scores the same as the order of the playback devices in the first group 201 in the first sound quality ranking information 221.

[0030] In one embodiment, the processing module 214 may be a central processing unit (CPU), a graphics processing unit (GPU), or a computing unit with computing capabilities.

[0031] The processing module 214 of the sound quality evaluation system 210 adjusts these second evaluation scores according to the first reference model, thereby obtaining a second sound quality ranking information for the second group 202 (step S70). At this time, the sound quality evaluation algorithm model f(x) has been trained and can objectively evaluate the sound quality performance of one or more playback devices 200. Therefore, after calculating the second audio data using a machine learning algorithm and the first reference model, objective second sound quality ranking information and the sound quality performance of the playback devices in the second group 202 can be obtained.

[0032] In one embodiment, the calculation module 212 of the audio quality evaluation system 210 further calculates these second audio data using a spatial algorithm to obtain multiple spatial scores for the second group 202. A higher spatial score indicates better spatial performance of the playback device for the second group 202 when playing audio. The spatial algorithm includes a head-related transformation function and a minimum variation distortionless response algorithm. Those skilled in the art are familiar with the head-related transformation function and the minimum variation distortionless response algorithm, and therefore will not elaborate further.

[0033] In one embodiment, the calculation module 212 of the sound quality evaluation system 210 further calculates these second audio data using a dynamic algorithm to obtain multiple dynamic scores for the second group 202. A higher dynamic score indicates better dynamic performance of the playback device of the second group 202 when playing audio. The dynamic algorithm includes a spectrum analysis method, a linear regression method, and a Gini coefficient method. Those skilled in the art are familiar with spectrum analysis, linear regression, and Gini coefficient methods, and therefore will not elaborate further.

[0034] In one embodiment, the calculation module 212 of the sound quality evaluation system 210 further calculates these second audio data using a volume algorithm to obtain multiple volume scores for the second group 202. A higher volume score indicates better volume performance of the playback device in the second group 202 when playing audio. The volume algorithm is a dynamic range suppression method, which is well-known to those skilled in the art and will not be described in detail here.

[0035] In one embodiment, the calculation module 212 of the sound quality evaluation system 210 further calculates these second audio data using a distortion algorithm to obtain multiple distortion scores for the second group 202. A higher distortion score indicates worse distortion performance of the playback device of the second group 202 when playing audio. The distortion algorithm includes a dynamic intermodulation distortion method and a sharpness spectrum analysis method (also known as sibilance spectrum analysis). Those skilled in the art are familiar with the dynamic intermodulation distortion method and the sibilance spectrum analysis method, and therefore will not elaborate further.

[0036] The sound quality evaluation method and system of this application train an objective sound quality evaluation algorithm model by referring to the sound quality ranking information of audio devices published in one or more publicly available online databases, without requiring the intervention of acoustic experts during the training process. As training data accumulates, the evaluation scores calculated by the sound quality evaluation model of this application not only closely approximate the judgments of acoustic experts, but also completely avoid the evaluation biases caused by occasional changes in the physiological and psychological conditions of acoustic experts during evaluation. Therefore, it can evaluate the sound quality of various playback devices more objectively and consistently than acoustic experts, thus providing an objective and accurate evaluation method and conveniently allowing users to understand the performance of playback devices.

[0037] The embodiments and / or implementation methods described above are merely preferred embodiments and / or implementation methods for implementing the technology of this application, and are not intended to limit the implementation methods of the technology of this application in any way. Any person skilled in the art may make some modifications or alterations to other equivalent embodiments without departing from the scope of the technical means disclosed in this application, but these should still be regarded as the technology or embodiments that are substantially the same as those of this application.

Claims

1. A sound quality evaluation method for providing sound quality ranking information for multiple playback devices, characterized in that, Include: These playback devices are defined as a first group and a second group; Record at least one test audio file played in the first group and the second group respectively to generate multiple first audio data and multiple second audio data; Each of the first audio data and each of the second audio data are segmented to generate multiple first group frequency bands and multiple second group frequency bands; The frequency bands of the first group and the frequency bands of the second group are calculated and processed respectively to obtain multiple first evaluation scores of the first group and multiple second evaluation scores of the second group; Extract the first audio quality ranking information corresponding to the first group from the reference source; The first reference model is obtained by adjusting the first evaluation scores accordingly based on the first sound quality ranking information. as well as The second evaluation scores are adjusted according to the first reference model to obtain the second sound quality ranking information of the second group.

2. The sound quality evaluation method according to claim 1, characterized in that, Also includes: These second audio data are calculated using a spatial algorithm to obtain multiple spatial scores for the second group; These second audio data are calculated using a dynamic algorithm to obtain multiple dynamic scores for the second group; The second audio data is calculated using a volume algorithm to obtain multiple volume scores for the second group; as well as The second audio data is calculated using a distortion algorithm to obtain multiple distortion scores for the second group.

3. The sound quality evaluation method according to claim 1, characterized in that, The first audio data and the second audio data are generated by recording at least one test audio file played by the first group and the second group using an artificial head device.

4. The sound quality evaluation method according to claim 1, characterized in that, The reference sources are publicly available online databases.

5. A sound quality evaluation system, characterized in that, Include: An audio recording module is used to define multiple playback devices as a first group and a second group, and to record at least one test audio file played by the first group and the second group respectively, so as to generate multiple first audio data and multiple second audio data; The calculation module is used to segment each of the first audio data and each of the second audio data to generate multiple first group frequency bands and multiple second group frequency bands, and to calculate and process the first group frequency bands and the second group frequency bands respectively to obtain multiple first evaluation scores of the first group and multiple second evaluation scores of the second group. The communication module is used to retrieve the first audio quality ranking information corresponding to the first group from the reference source; as well as The processing module is used to adjust the first evaluation scores by referring to the first sound quality ranking information, thereby obtaining a first reference model, and to adjust the second evaluation scores according to the first reference model, thereby obtaining the second sound quality ranking information of the second group.

6. The sound quality evaluation system according to claim 5, characterized in that, The calculation module is also used to calculate the second audio data using a spatial algorithm to obtain multiple spatial scores for the second group, to calculate the second audio data using a dynamic algorithm to obtain multiple dynamic scores for the second group, to calculate the second audio data using a volume algorithm to obtain multiple volume scores for the second group, and to calculate the second audio data using a distortion algorithm to obtain multiple distortion scores for the second group.

7. The sound quality evaluation system according to claim 5, characterized in that, The audio recording module is an artificial head device.

8. The sound quality evaluation system according to claim 5, characterized in that, The reference sources are publicly available online databases.

Images