Historical development of the use of A-weighting

The original writers of the specifications for American sound level meters wished to provide a standard instrument that would indicate the loudness of sounds as perceived by the normal human ear. As loudness varies with the sound pressure level and frequency of a sound, three frequency weighting networks were specified to mimic the response of the ear at the loudness levels of 40, 70, and approximately 100 phons. The first of these was designated the A-weighting and the last was called the C-weighting, which was a flat response. When the rules given for using the weighting networks were followed, the sound level meter indicated very accurately the loudness levels of pure tones but not broadband noises. For measuring noises more precisely, special rules were developed by sound level meter users for combining the readings obtained using two or more of the weighting networks, particularly the A- and C-weighting networks. Long experience with estimating human responses to noise from sound level meter readings has led to the general use of the A-weighting network alone for many purposes.