In Reply to: dcs Dacs and what makes 'em tick (hint: wavelets) :) posted by Pinkus on May 4, 2000 at 20:40:19:
There's precious little information in that document,
as as such footnote 5 is nothing less than ridiculous.With all respect to Mike Story, I feel that even his
'upsampling FAQ' has been inspired more by their marketing
people than by technical precision.He defines oversampling as something needed by low-bit DACs
(or ADCs) to perform acceptably, and upsampling as
something fairly unique to dCS. Well ...1) According to his own definitions, even the Ring DAC falls
in the apparently despised group of low-bit DACs.
2) Synchronous oversampling, or upsampling, as he defines it,
has been with us from day 1, from the very first Philips
CD-player, and since then in every multi-bit DAC (Burr
Brown, Analog Devices, TDA1541, ...), and exactly for
the reasons Mike claims dCS are using it now.
3) Oversampling is not a trick to make low-bit stuff perform.
It is an unseparable part of the architecture of
such DACs and ADCs, where the time domain is traded
versus the amplitude domain. There is nothing wrong with
that. Even bolder: if you stick to traditional architectures,
there is not even a chance to ever get beyond 21 bits
resolution. With time domain tricks, you can get to
23-24 bits, as dCS themselves prove.
So back to the original document: it is a marketing blurb.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Re: dcs Dacs and what makes 'em tick (hint: wavelets) :) - Werner 05/5/0001:08:52 05/5/00 (4)
- Re: dcs Dacs and what makes 'em tick (hint: wavelets) :) - PaulN 10:48:46 05/5/00 (1)
- above post meant for Pinkus : ) - PaulN 10:50:40 05/5/00 (0)
- 24 bit resolution or not.... - Pinkus 09:08:48 05/5/00 (1)
- Re: 24 bit resolution or not.... - Werner 23:28:02 05/10/00 (0)