To establish the subjective evaluation of noise on a scientific basis, we must first measure the perceptual characteristics of human ears to the loudness of various frequencies. It has long been known that the sensitivity of human ear to low-frequency sound and high-frequency sound is very different, but the quantitative measurement of this characteristic began in the 1930s. H. Fletcher and W. A. Manson measured many subjects with pure tones of various frequencies at 1933, and drew a set of isoacoustic curves reflecting the subjective feelings of human ears. The isoacoustic curve measured by D.W. Robinson and R.S. Dadson at 1956 was accepted by the International Organization for Standardization (ISO/R226- 196 1). In addition, some people use narrow-band noise and octave-band noise to measure the isosound curve, and get the relationship between pure tone and noise.

According to the noise spectrum, S.S. Stevens put forward a method to calculate the noise loudness in 1956, and later E. Cvik put forward a method to calculate the loudness by calculating the area with the spectrum curve of 1/3 octave band. Both methods are accepted by the International Organization for Standardization (ISO/R532- 1966).

No matter what method is used to find loudness, complicated calculation is needed. In order to directly read out the evaluation quantity reflecting people's subjective loudness feeling with instruments, R.W. Yang and others proposed to simulate the frequency characteristics of human ears at different sound intensities with electronic networks, so as to directly measure the subjective evaluation quantity of noise with instruments (sound level meters), and the measured result is called weighted sound level or sound level for short. Common sound levels are A, B and C (see noise meter). When measuring noise, people gradually find that the sound level measured by A-weighted network is closer to the overall evaluation of noise by human ears. Grade A has been adopted by the International Organization for Standardization and most countries as the main index of subjective evaluation of noise.

With the appearance of jet aircraft, the problem of environmental noise has become increasingly prominent. In order to evaluate aircraft noise and traffic noise correctly, many evaluation methods have been put forward. One of the most important is K.D. Crete's method of evaluating the noise near the airport with the perceived noise level (). This method has been adopted by the International Organization for Standardization (ISO/R507- 1970). Many evaluation methods developed in the future can be roughly divided into two categories. One is based on sound level A, such as equivalent sound level, day and night equivalent sound level, statistical sound level, noise pollution level and evaluation noise level. The other is based on perceived noise level, such as effective perceived noise level, effective continuous perceived noise level and D sound level.

In addition, as for the evaluation of the influence of noise on language intelligibility, there is the level of language interference proposed by L.L. Berenak in 1947; There are noise standard curves and priority standard curves proposed by Brenak in 1957 and 197 1 respectively. The International Organization for Standardization (ISO) adopts noise evaluation number to comprehensively evaluate the hearing loss, language interference and annoyance caused by noise. The latter proves that the noise evaluation number can generally be converted into decibels.