Audio noise measurement tests the quality of sound equipment. It does this by measuring the amount of noise present in any audio recording. Any distortion and noise are then removed through a system of sound weighting. Such measurements are performed on radio and television broadcast studios, recording studios, home recording, and audio equipment.
Noise in a recording is the presence of any unwanted sounds. Microphones, amplifiers, speakers and other recording equipment can cause foreground noise. Background noise can also be created through sounds as distant as traffic or in-studio sounds such as shuffling papers and general noises associated with an audience. Not all noises can be caught prior to or during recording. Audio noise measurement detects these noises when a recording is played back through the right equipment.
The human ear is attuned to pick up medium frequencies because these are the frequencies of the human voice. As a result of this, the human ear is not good at picking up very low and very high frequencies. This means that audio noise measurement is compared to a standard sound pressure level, or SPL. An SPL of 0 is seen as the threshold for sound reaching the human ear. An SPL of 10 to 20 is the sound of rustling leaves and whispers, whereas a level of 220 is akin to a person placing his or her head in front of a canon as it fires.
Condenser microphones tend to be used during audio noise measurement. The condenser microphone has a broad range of frequency responses. It also has a polarized diaphragm. There are three basic types of condenser microphones used for capturing noise: the free field microphone, the pressure microphone and the random incidence microphone.
There are a number of ways of measuring the presence of noise in recordings. One basic method is to take a recording with the source of the sound and then a recording of just the background noise. Signal-to-noise ratio is an audio noise measurement technique that uses decibels to measure the root mean square (RMS) of the energy of all sounds excluding harmonics. Signal-to-noise plus distortion ratio includes harmonics, while the dynamic range measurement compares the ratio of the greatest magnitude against the quietest signal.
Sound weighting takes an audio sound measurement and seeks to remove noises that are audible to the human ear. There are two main weighting methods: A-weighting and ITU-R 468 weighting. The A-weighting curve is created from an equal-loudness contour designed to show which sounds the human ear is sensitive to. After criticism of the a-weighting curve’s accuracy, the British Broadcasting Corporation (BBC) created the ITU-R 468 curve, which includes 11 decibels of noise reduction.
