dB (Decibel) ratings

Sound pressure

Sound is defined as a variation in air pressure that can be heard by the human ear. The average audible frequency range for humans extends from 20 Hz to 20 kHz. In terms of sound pressure level (SPL), audible sound ranges from the threshold of hearing at 0 dB to the threshold of pain, which can exceed 130 dB.

An increase of 3 dB represents a doubling of sound pressure. However, a sound must increase by approximately 10 dB before it is perceived by the human ear as being twice as loud. Under normal conditions, the smallest change in sound level that most people can detect is about 3 dB.

Perceived loudness and frequency response

Perceived (subjective) loudness is influenced by several factors, one of the most significant being frequency. The human ear does not respond equally to all frequencies. Sensitivity is greatest in the 2 kHz to 5 kHz range and decreases at both lower and higher frequencies.

This variation in sensitivity is more pronounced at lower sound pressure levels. For example, a 50 Hz tone must be approximately 15 dB higher than a 1 kHz tone at a 70 dB level in order to be perceived as equally loud.

Impulse and short-duration sounds

Sounds with a duration of less than one second are classified as impulse or impulsive sounds. Due to their short duration, the human ear is less sensitive to their perceived loudness. In general, sounds shorter than 70 milliseconds are perceived as less loud than continuous sounds with the same measured SPL.

Sound level weighting (A, B, and C)

Sound level meters use frequency weighting networks to better reflect how humans perceive sound:

• A-weighting – dB(A): Approximates human hearing sensitivity at low sound pressure levels
• B-weighting – dB(B): Corresponds to medium sound pressure levels
• C-weighting – dB(C): Corresponds to high sound pressure levels

A-weighting (dB(A)) is the most commonly used standard, as B and C weightings do not correlate well with subjective hearing tests. One reason for this limitation is that equal-loudness contours were originally developed using pure tones, whereas most real-world sounds consist of complex, broadband signals.