Research on audio quality test methods for smart video doorbell products

1. Overview

1.1 Test Purpose

This solution aims to provide a set of audio test solutions for customers to refer to for the production and development of smart video doorbell products. It uses scientific and rigorous measurement methods to test the audio performance indicators of products, such as frequency response, THD+N, noise, signal-to-noise ratio, audio spectrum, etc. It provides visual data guidance for the production and development of smart video doorbell products and improves product quality control.

1.2 Test indicators

1.2.1 Frequency Response

Frequency response refers to the output signal level of the DUT when it is stimulated by excitation signals of known levels and different frequencies. The most common method is to sweep the sine wave signal from the lowest frequency to the highest frequency within the frequency range of the DUT and plot the results.

When doing a frequency sweep test, the first thing to do is to determine the sweep level. You can sweep at a low level, but noise or other spurious signals may appear in the response; you can also sweep at a high level, but high distortion may occur.

1.2.2 THD+N

The full name of THD+N is Total Harmonic Distortion plus Noise. Harmonic distortion refers to the unnecessary tones added to the original audio signal. It is the harmonic related tone of the original signal. When the signal is a sine wave with frequency f1, the harmonics are f2, f3, etc., which are integer multiples of the original tone. Total harmonic distortion is the sum of all harmonics in the bandwidth of the device under test.

THD and noise have always been tested together. Some people ask why harmonics and noise are not tested separately. This is because when testing THD and noise, the Fast Fourier Transform (FFT) test is used, and it is difficult to separate harmonics and noise. However, testing together is relatively simple. Of course, the ABTEC audio analyzer can view the total harmonics, noise, and the distribution of harmonic signals of each order from f2 to f10 in real time as needed.

1.2.3 Noise and Signal-to-Noise Ratio

Definition: The term "noise" refers to any unwanted signal, including AC power hum, stray magnetic field interference from circuit components, etc. Noise measurement must use a filter. General noise measurement only refers to random noise with energy distributed over a wide frequency band, as well as unwanted coherent signals such as hum and interference.

Unit: decibel (dB) is generally used to measure the intensity of noise, and signal-to-noise ratio (S/N) is used to measure the impact of noise on useful signals.

How much noise is too much? That depends on how loud the signal is.

The signal-to-noise ratio (SNR) is the measurement of this difference. The signal-to-noise ratio refers to the ratio of the signal to the noise in an electronic device or electronic system. In the process of testing the signal-to-noise ratio, two steps were required in the past. First, an excitation level was input to the product under test.

The instrument obtains the size of the signal, then turns off the signal, obtains the noise signal, and then compares and calculates the signal-to-noise ratio. Today's instruments all automatically control signals and automatically calculate data.

1.3 Test system principle

Generally, the audio system of smart video doorbell products can be divided into microphone audio channel and speaker audio channel. The microphone audio channel collects sound signals and converts them into analog audio electrical signals. The subsequent system receives the analog audio electrical signals from the microphone and converts them into digital audio signals, which are transmitted to the mobile terminal through the wireless network. The mobile terminal converts them into .wav format audio files through the corresponding APP application. The speaker audio channel converts the .wav files stored in the SD card or the local memory of the smart video doorbell products into analog audio electrical signals and transmits them to the speaker. The speaker receives the analog audio electrical signals and converts them into sound signals and sends them out.

2. Microphone Path Test

2.1 Whole machine test

For the whole machine test, the test instrument is a professional audio analyzer plus an artificial mouth, which can provide analog and digital audio test signals and sound wave signals. During the test, the artificial mouth and the system (product) under test should be placed in a professional soundproof box to prevent the ambient sound from affecting the test system. The professional audio analyzer sends the test sound wave signal to the system (product) under test, and the system under test outputs a .wav format audio file to the professional audio analyzer, forming a closed-loop test system.

2.1.1 THD+N test

Test steps:

(1) Calibrate the artificial mouth using a standard microphone;

(2) Use an audio analyzer to generate an analog test audio signal with a frequency of 1 kHz and an amplitude of 1 V;

(Note: The amplitude is temporarily set at 1V. The actual test should be adjusted according to the volume of the human voice in the application environment.)

(3) Use the corresponding APP application to record the test audio signal into an audio file in .wav format;

(4) Use an audio analyzer to analyze the .wav audio file, observe and record the product THD+N indicator (can be combined with the audio spectrum for analysis);

(5) Repeat the above steps several times to avoid test contingencies.

2.1.2 Spectrum Analysis

Test steps:

(1) Use a standard microphone to calibrate the artificial mouth (if it has been calibrated in the previous test, it does not need to be calibrated again in this test);

(2) Use an audio analyzer to generate an analog test audio signal with a frequency of 1 kHz and an amplitude of 1 V;

(Note: The amplitude is temporarily set at 1V. The actual test should be adjusted according to the volume of the human voice in the application environment.)

(3) Use the corresponding APP application to record the test audio signal into an audio file in .wav format;

(4) Use an audio analyzer to analyze the .wav audio file, save it, and observe and analyze the audio spectrum.

(5) Repeat the above steps several times to avoid test contingencies.

2.1.3 Signal-to-noise ratio test
Test steps:

(1) Use a standard microphone to calibrate the artificial mouth (if it has been calibrated in the previous test, it does not need to be calibrated again in this test);

(2) Use an audio analyzer to generate a signal-to-noise ratio test audio signal (the first half has a frequency of 1 kHz and an amplitude of 1 V, and the second half does not generate an audio signal);

(Note: The amplitude is temporarily set at 1V. The actual test should be adjusted according to the volume of the human voice in the application environment.)

(3) Use the corresponding APP application to record the test audio signal into an audio file in .wav format;

(4) Use an audio analyzer to analyze the .wav audio file and observe and record the product signal-to-noise ratio.

(5) Repeat the above steps several times to avoid test contingencies.

2.1.4 Frequency response test
Test steps:

(1) Use a standard microphone to calibrate the artificial mouth (if it has been calibrated in the previous test, it does not need to be calibrated again in this test);

(2) Use an audio analyzer to generate a step-sweep test audio (frequency range 20 Hz-20 kHz, linear sweep, 30 sweep points, amplitude 1 V);

(Note: The frequency range and the number of sweep points are determined according to the actual audio processing range of the system under test. The amplitude is temporarily set at 1V. The actual test is adjusted according to the volume of the human voice in the application environment.)

(3) Use the corresponding APP application to record the test audio signal into an audio file in .wav format;

(4) Use an audio analyzer to analyze the .wav audio file and observe and record the product frequency response indicators.

(5) Repeat the above steps several times to avoid test contingencies.

2.2 Audio Capture Coding System Test

The door-mounted part of the smart video doorbell will collect and encode the signal input by the microphone. This part also needs to be tested. The test instrument is a professional audio analyzer, which can provide analog and digital audio test signals. The professional audio analyzer sends the analog test audio electrical signal to the subsequent system (skipping the microphone), and the subsequent system outputs the .wav format audio file to the professional audio analyzer, forming a closed-loop test system.

2.2.1 THD+N test

Test steps:

(1) Use an audio analyzer to generate an analog test audio signal with a frequency of 1KHz and an amplitude of 1V;

(2) Use the corresponding APP application to record the test audio signal into an audio file in .wav format;

(3) Use an audio analyzer to analyze the .wav audio file, observe and record the THD+N indicator of the product (can be combined with the audio spectrum for analysis).

(4) Repeat the above steps several times to avoid test contingencies.

2.2.2 Spectrum Analysis

Test steps:

(1) Use an audio analyzer to generate an analog test audio signal with a frequency of 1KHz and an amplitude of 1V;

(Note: The amplitude is temporarily set at 1V. The actual test should be adjusted according to the volume of the human voice in the application environment.)

(2) Use the corresponding APP application to record the test audio signal into an audio file in .wav format;

(3) Use an audio analyzer to analyze the .wav audio file, save it, and observe and analyze the audio spectrum.

(4) Repeat the above steps several times to avoid test contingencies.

2.2.3 Signal-to-noise ratio test

Test steps: