Sound quality evaluation

1. Listening Room
The main listening place for audio is indoors. Home audio is the family home. Professional audio requires a studio or recording room with certain acoustic requirements. When listening and evaluating, a special listening room should be used, and the actual application environment should be simulated in the area used, such as the listening environment of a typical family or a specific application place. Usually, the acoustic characteristics of a family home are limited by decoration and economic factors, and the acoustic characteristics of the room cannot be in the best state. From the perspective of listening evaluation, in order to facilitate the exchange of listening evaluation results, it is necessary to establish a special listening room, focusing on the following 5 aspects.
1. Room size. The fundamental element of standing waves in a rectangular room is the ratio of length, width and height. If the size ratio is not appropriate, peaks and valleys will form at some frequency points in certain positions in the room, causing audible sound coloration. The recommended listening room volume is: 90-120m3 for mono programs and 120-150m3 for stereo programs; the best ratio of length, width and height is 1.0:1.4:1.9; 1:1.26:1.59 can also be used. If other ratios are used, the indoor resonant frequency should be calculated to avoid isolated and grouped simple harmonic vibrations and standing waves.
2. Reverberation time. The reflected sound and reverberation sound in the listening room have a great influence on the quality of the entire sound. The amount, type and placement of the sound-absorbing material determine the indoor reverberation time (T), the uniformity of the reverberation sound field, and the degree of diffusion. The reverberation time should be measured using a pink noise source and a frequency point of 1/3 octave in an empty field without a listener. According to international standards, the reverberation time T and the uniformity of the frequency are: between 125Hz and 4kHz, the average value of T is 0.25 to 0.45 seconds, the frequency characteristics should be flat or slightly longer at low frequencies and slightly shorter at high frequencies, and the difference between the single measurement value and the average value at each point within the frequency should not exceed 25%; 0.35±0.05 seconds is taken within 125 to 165Hz; 0.3±0.05 seconds is taken from 200Hz to 4kHz, and T is allowed to deviate from the average value of the intermediate frequency by more than 25%. According to practical experience, if the listening room is mainly used for listening evaluation, the reverberation time is better at 0.3 to 0.4 seconds, and it should not be too short, otherwise the sense of openness will be worse; if the reverberation time is too long, the sound image positioning will not be clear enough, but the sense of openness will be better.
3. Transmission frequency response. After selecting the sound excitation point, the sound source position and the receiving point, that is, the microphone or the listening ear position, the unevenness of the amplitude-frequency response should be within the specified range. The measurement uses a sine signal, and a pink noise signal can also be used. When evaluating stereo products, the imbalance of the sound field (the symmetrical value of the left and right channels) should be higher than that of the mono channel.
4. Environmental conditions. When there is no listener in the empty field, the background noise in the listening area should be lower than 25dB when the lights and air conditioner are turned on; it should not exceed 35dB when there are staff. There should be no low-frequency hum, high-frequency hiss or flutter echo in the room. When measuring the indoor environment with a sound level instrument of 95dB or above in the frequency range of 100Hz to 5kHz at a distance of 1m from the sound source, there should not be any abnormal resonance and flutter echo. Since the reliability of listening evaluation is greatly affected by the extreme values ​​of environmental conditions, the ambient temperature is 15-30℃, preferably 20℃; the relative humidity is 25-75%; the air pressure is 360-1060Hg, and the indoor lighting should be soft and not directly shine into the eyes of the listener.
5. Listening feeling. Use vocal and language program tapes corresponding to the purpose of the room, and a group of trained evaluators will evaluate at the designated listening position. The listening evaluation results should be determined and analyzed, and defects caused by the acoustic conditions of the room or the electroacoustic system should be eliminated. If necessary, the original voice of the actor can be used to evaluate specific parameters.
2. Programs
There are two methods for selecting programs for listening evaluation: one is to select according to the sound quality characteristics; the other is to select according to the main types of audio products. Regardless of which method is used, the following points should be considered.
1. Real and natural. The program sound source should be real and natural, not too exaggerated, and the principle is to be able to evaluate the sound quality
of electro-acoustic products, and not excessively pursue a certain artistic effect. 2. Distinctive features. The sound quality should have relatively distinct features, and the features of each program segment are different from each other, so that there are many differences in the evaluation. The whole program should be able to basically assess the sound quality level of the electro-acoustic product, that is, the program materials cannot be piled up at will, making the evaluation time very long but with little information; it is also necessary to prevent incomplete materials, so that some sound quality cannot be expressed.
3. Neutral emotions. Avoid choosing programs with too strong "emotional colors" to prevent affecting the psychological state of the sound quality evaluators.
4. High popularity. Try to choose familiar works and relatively familiar actors and announcers. The audition program includes various types of language, symphonic orchestral music, classical music, chamber music and songs, various light music, including classical light music, folk music, dance music with or without vocal music and jazz music, as well as a small number of instrumental combinations, such as piano, violin and cello, wooden orchestral instruments, steel instruments, percussion instruments, solo and instrumental accompaniment.
5. Typical programs.
(1) Voice. The program time should not be too long. The voice can use male and female voice commentary. The female voice should have a natural and friendly tone, good resonance of the vowels, and sound solid and full; the male voice should have a broad and rich voice.
(2) Piano solo. The third movement of Beethoven's "Moonlight" Sonata can be selected, which lasts about 37 seconds. It includes 5 octaves of the piano, with a smooth melody and large fluctuations in the playing intensity, basically showing the characteristics of the piano. The sound is solid and elastic, harmonious and clear, and the climbing notes from low to high add color to the program.
(3) String ensemble. You can choose the first movement of Mozart's Serenade for Strings, with a duration of 48 seconds. The performance technique is sophisticated, clean and unified. The sound of each part of the violin is solid, distinctive, naturally balanced, and has a strong sense of integrity and harmonious beauty.
(4) Woodwind Quartet. You can choose the first movement of Jean Françoise's Woodwind Quartet, with a program time of 37 seconds. This is a typical chamber music work. The sounds of the flute, clarinet, oboe and bassoon are natural, real and balanced. Even the keystrokes and breathing sounds are clearly identifiable, which adds to the chamber music characteristics.
(5) Brass. You can choose the first movement of Tchaikovsky's Symphony No. 4, with a program time of 38 seconds. The performance is grand and sounds high-pitched and rich. The French horn at the beginning is far away and not empty, the trombone that comes in later is majestic and powerful, and the trumpet is crisp and soft.
(6) Orchestra. You can choose Pictures at an Exhibition composed by Mussorgsky and orchestrated by Ravel, with a program time of 1 minute and 6 seconds. Performed by a large symphony orchestra, it is magnificent and has a strong sense of hearing. Each voice is fully and naturally played, especially the solid and powerful sound of the timpani and the deep and powerful sound of the drum, which is earth-shaking.
(7) A cappella singing. You can choose the song "The More You Sing, the Deeper the Feelings", and the program time is 40 seconds. Each part is bright, beautiful and mellow.
(8) Female solo. Choose the song "In the Field of Hope", and the program time is 1 minute and 2 seconds. The voice is crisp and sweet, and the pronunciation is natural and clear. The folk singing method is used to make the song more intimate.
(9) Male solo. Choose the song "Chai Tou Feng", and the program time is 53 seconds. The voice is relatively relaxed and natural, and the accompaniment and singing are more classical in style.
(10) Folk music ensemble. Choose the song and dance "Silk Road Flower Rain", and the program time is 47 seconds. The national style is very strong, and the various musical instruments are realistic and pleasant. The drum sounds are relaxed and soft, the flute sounds crisp and cold, and the stringed and plucked instruments are also distinctive and full of vitality.
(11) Percussion. Duration 40 seconds. The percussion methods of various instruments are realistic, the inherent characteristics of the sound are sufficient, and the sound and image are real.
(12) Electronic synthesis segment. Duration 54 seconds. The use of typical electronic music, the sound is loud and lively, and a moving sound and image are interspersed in the middle, which enhances the vivid interest.
(13) Peking opera version. Select "Fan Xipi Xiao Kaimen", the duration is 1 minute and 8 seconds. The layers are clear and clean, the sound of the jinghu is solid but not rough, the sound of the percussion instruments is round but not broken, the sound of the plucked instruments is clear and real, and the sound and image positioning is stable and clear.
(14) Peking opera dialogue. Select "Shisanmei", the program duration is 51 seconds.
(15) Peking opera singing. The song "Li Qingzhao" was selected. The duration is 1 minute and 6 seconds. The female role's voice is clear and melodious, and it is clear and gentle.
(16) A steam locomotive passes through a fork in the road. The duration is 51 seconds. It shows the sound of a loaded freight train passing through a fork in the road. It comes from far away and approaches at a high speed, giving people a sense of being there. Due to the full load, the sound of the wheels hitting the rail joint is particularly prominent.
(17) The sound of a bell. The duration is 30 seconds. The impact is distinctive, the sound is solid and powerful, the reverberation decays slowly and evenly, and there are faint human voices and bird and insect sounds, which makes the environment natural and real.