bit rate and dither bit depth
- Thread starter banjo71
- Start date
bouldersoundguy
Well-known member
Word length (bit depth) is how big a number can be used to encode the level of a wave over time. The bigger the number the bigger the range from 0dBFS (the absolute maximum level) down to the noise floor. 24 bit audio has a very low noise floor, 8 bit has a much higher noise floor.
Recording in 24 bit is good, but most consumer audio is 16 bit so you have to truncate, chop off the lower bits, for the end user. This adds a certain kind of noise that degrades the sound. Dithering is the process of adding another specific kind of noise to the signal before truncating to lessen the effects of the truncation noise. When converting from 24 to 16 bit you'll want the very last process before truncating to be dithering for 16 bit.
Mo Facta
Farts of Nature
Well explained, bouldersoundguy.
Although, I would add that dither doesn't "lessen" the effects of truncation, but rather remedies it altogether with a slight increase in noise. With properly implemented dither, a signal will attenuate smoothly and without artifacts into silence. The only trade off is about 3dB of added noise to the noise floor. The ear can distinguish signals below the noise floor which is why it is possible to gain an extra couple of bits in a standard CD.
Cheers
Although, I would add that dither doesn't "lessen" the effects of truncation, but rather remedies it altogether with a slight increase in noise. With properly implemented dither, a signal will attenuate smoothly and without artifacts into silence. The only trade off is about 3dB of added noise to the noise floor. The ear can distinguish signals below the noise floor which is why it is possible to gain an extra couple of bits in a standard CD.
Cheers