weighting filters« BACK TO DICTIONARY INDEX
(n.) In audio, special filters used in measuring loudness levels and in noise measurements of equipment. The filter weights certain frequency bands higher than others. The objective is to obtain measurements that correlate closely with the subjective perception of noise. Weighting filters are band-limiting filters that compensate for the fact that the ear’s loudness vs. frequency response is not flat. Four weighting filter designs are described below:
- A-weighting The A-curve is a wide bandpass filter centered at 2.5 kHz, with ~20 dB attenuation at 100 Hz and ~10 dB attenuation at 20 kHz. It heavily attenuates the low end, with a more modest effect on high frequencies. It is the inverse of the 30-phon (30 dB-SPL) equal-loudness curve of Fletcher Munson. This is the least revealing scale.
- C-weighting The C-curve is flat but has very limited bandwidth, with -3 dB corners at 31.5 Hz and 8 kHz, respectively.
- ITU-R 468-weighting This filter was designed to maximize its response to the types of impulsive noise coupled into audio cables as they pass through telephone weighting filters switching facilities. It also correlates well with noise perception since modern research has shown that frequencies between 1 kHz and 9 kHz are more annoying than indicated by A-weighting curve testing. The ITU-R 468-curve peaks at 6.3 kHz, where it has 12 dB of gain (relative to 1 kHz). From that point, it gently rolls off low frequencies at a 6 dB per octave rate, but it quickly attenuates high frequencies at ~30 dB per octave (it is down -22.5 dB at 20 kHz, relative to +12 dB at 6.3 kHz).
- ITU-R (CCIR) 2 kHz-weighting This curve is based on the ITU-R 468 curve. Dolby Laboratories proposed using an average-response meter with the ITU-R 468-curve instead of the true quasi-peak meters used by the Europeans in specifying their equipment. They also shifted the 0 dB reference point up from 1 kHz to 2 kHz, sliding the curve down 6 dB. This became known as the ITU-R ARM (average response meter).