User talk:Kdarcangelis870

I've had a quick look at your sources, but I don't accept their conclusions. If digital systems added lots of distortion on conversion, then this would be shown in objective tests, such as dynamic range or signal-to-noise ratio tests, intermodulation distortion tests, harmonic distortion tests, and null tests.

A null test is where the original signal (analog) can be subtracted from the digitized signal, thus (after normalization) leaving the distortion products. On a good digital converter, the residuals should be small and would, in normal listening, most likely be masked by the music signal itself.

Malcolm Hawksford has written on this subject, and proposed a null test that uses ADC's:

C30 TOWARDS A DEFINITIVE ANALYSIS OF AUDIO SYSTEM ERRORS, Dunn, C., and Hawksford, M.O.J., 91st Convention of the Audio Engineering Society, New York, October 1991, preprint 3137 (P-1)

http://www.essex.ac.uk/ese/research/audio_lab/malcolmspubdocs/C30%20Analysis%20of%20audio%20system%20errors.pdf

The paper cites earlier null tests done by Louis Fielder of Dolby labs which examine digital audio converter performance.

I'd also point to subjective tests, where in double-blind conditions, digital converters are shown to act transparently:

http://www.bostonaudiosociety.org/bas_speaker/abx_testing2.htm

One other paper which examines the performance of digital systems is by Robert Stuart of Meridian Audio:

http://www.meridian-audio.com/ara/coding2.pdf

He points to the work of Stanley Lipshitz which shows that use of dither (of a triangular probability distribution form) allows a digital signal to theoretically have an infinite dynamic range. Enescot 15:07, 30 December 2006 (UTC)