[net.audio] Compact Discs and DAC size

evans (03/31/83)

	A discussion of the 14 bit (vs 16 bit) playback compact disc
players can be found in the May 1982 issue of Electronic Components
and Applications.
	I gather the main reason this is done is to use a 14 bit DAC
which, presumably, is cheaper than a 16 bit part. They improve the SNR
from 84 dB to 90 dB by oversampling. The data streams are increased
in frequency by a factor of four which distributes the noise across
a 88 kHz bandwidth rather than 22 kHz. Now they simply filter out the
high frequency component. The next trick is to use a noise shaper,
which buys them another 7 dB.
	It would be interesting to see if these tricks really work
as claimed. They start with the theoretical SNR in the paper (84 dB
for 14 bits) and add their improvements from there. If memory serves
me the 16 bit players are around 90 dB a number < the 98 dB
theoretical value. Comparisons between 14 and 16 bit players would
be interesting -- e.g. are such differences audible?

				Steve Crandall
				mhuxt!evans