Sound and Hearing Theory--The Masking Effect of Human Ear

The masking effect of the human ear

The phenomenon in which the auditory perception of a weaker sound (masked sound) is affected by another stronger sound (masked sound) is called the "masking effect" of the human ear. The decibel value of the hearing threshold when the masked sound exists alone, or the smallest value of the pure tone that can be heard by the human ear in a quiet environment, is called the hearing threshold. Experiments show that the threshold of 3kHz-5kHz is small, that is, the human ear is sensitive to its weak sound; and the threshold is much larger in the low and high frequency regions. In the range of 800Hz-1500Hz, the hearing threshold does not change significantly with frequency, that is, the language can be stored in this range. In the case of masking, increasing the intensity of the masked weak tone so that the human ear can hear the threshold is called the masking threshold (or masking threshold), and the decibel value increased by the masked weak tone is called the masking amount (or threshold shift) .

1. Masking effect

Existing experiments have shown that the masking effect of pure tones and noise on pure tones is as follows:

A. Masking between pure tones

① The effective masking of pure tones at medium intensity occurs near its frequency.

② The low-frequency pure tones can effectively mask the high-frequency pure tones, but the reverse is very small.

B. Noise masks pure tones. Noise is composed of a variety of pure tones and has an infinitely wide spectrum.

If the masking sound is broadband noise and the masked sound is pure tones, the masking threshold generated by it is generally 17dB higher than the noise power spectral density in the low frequency range, and is relatively flat; when it exceeds 500Hz, it will increase by 10dB per decade. If the masked sound is narrow-band noise and the masked sound is pure tones, the situation is more complicated. Among them, the narrow-band noise composed of pure tone components located near the masked sound, that is, the critical band has an obvious masking effect. The so-called critical frequency band means that when a pure tone is masked by continuous noise with a certain bandwidth with it as its center frequency, if the pure tone can just be heard, the power is equal to the power of the noise in this frequency band. The bandwidth is called the critical bandwidth. The unit of the critical frequency band is called Bark, and 1Bark = a critical frequency band width. When the frequency is less than 500Hz, 1Bark is approximately equal to freq/100; when the frequency is greater than 500Hz, 1Bark is approximately equal to 9+41 og (freq/1000), which is about 20% of the center frequency of a certain pure tone. It is generally believed that there are 24 sub-critical frequency bands in the range of 20Hz-16kHz. When a certain pure tone is outside the critical frequency band of the masking sound, the masking effect still exists.

2. Masking type

(1) Frequency domain masking

The so-called frequency domain masking refers to the masking effect that occurs when the masking sound and the masked sound act simultaneously, which is also called simultaneous masking. At this time, the masking sound has been active during the occurrence of the masking effect, which is a strong masking effect. Generally, a strong tone in the frequency domain will mask the nearby weak tone that is sounded simultaneously with it. The closer the weak tone is to the strong tone, the easier it is generally to be masked; conversely, the weak tone that is farther away from the strong tone is not easy to be masked. For example, if a 1000Hz tone is 18dB higher than another 900Hz tone, then the 900Hz tone will be masked by the 1000Hz tone. And if the 1000Hz tone is 18dB higher than the 1800Hz pitch farther from it, the two tones will be heard by the human ear at the same time. To make the 1800Hz sound inaudible, the 1000Hz sound is 45dB higher than the 1800Hz sound. Generally speaking, low-frequency sounds are easy to mask high-frequency sounds; at a distance from strong sounds, the threshold is higher than the masking threshold caused by the strong sounds. At this time, the masking threshold of noise should be the threshold.

(2) Time domain masking

The so-called temporal masking means that the masking effect occurs when the masking sound and the masked sound do not occur at the same time, which is also called different time masking. Simultaneous masking is divided into leading masking and lagging masking. If the masking effect occurs within a period of time before the appearance of the masking sound, it is called leading masking; otherwise, it is called lagging masking. The main reason for temporal masking is that it takes a certain amount of time for the human brain to process information, and different temporal masking will quickly decay over time, which is a weak masking effect. Under normal circumstances, the leading mask is only 3ms-20ms, while the lagging mask can last for 50ms-100ms.