Post on Tom Cosm's Facebook - page. Just want to share with the community.

"Why analog audio within the limits of human hearing (20hz - 20 kHz) can be reproduced with PERFECT fidelity using a 44.1 kHz 16 bit DIGITAL signal." - Tom Cosm

 

Just only one person read this ?? lol.

 

Very informative.
Thanks a lot!
I enjoyed that 1/2 hour

 

HP-3625 Signal Generator & HP-3585 Spectrum Analyzer. Cool gear. High dollar stuff once. Now owned by hobbyists & Ham Radio home-brewers.
Sure do wish I had them on my bench.

Monty Montgomery presents well. I think I'll seek more of his presentations.

But... all he really said was that the D/A conversion RESULTED in a fairly pure waveform, not that the exact same information was transferred. For most of us the information transfer is what counts. The information that would have been contained in the spaces between each sample point is essentially lost, and in no way recoverable on down the signal chain. [because it is just not there] He says that digital bit-depth does not matter. I disagree.

Possibly it is because his perspective is from the mind of an engineer, concerned mostly with the physics of a given scrutiny, rather than that of an artist.

All that said, why not sample a signal multiple times simultaneously [obviously the more the better] but shift time in micro-second increments in the A/D converter so as many gaps as feasible are covered. This would require a companion D/A converter that would recombine the segments in a similar fashion to give a close approximation of the original signal. Somewhat like spread-spectrum but in a time framework rather than a relative position framework. This would be somewhat like what 24/96 does compared to 16/44.1 but it could do it in an infinitely precise manner. I predict that this tech will be in production very soon. It was always limited by hardware's capabilities.
[imagine nitrogen-cooled A/D & D/A converters ]

Disclaimer:
I have no excuse for this post as I am NOT drunk nor superbly herbed.