[linux-audio-user] [OT] DSD +/- [was:44.1, 48 or 96?]

Steve Harris S.W.Harris at ecs.soton.ac.uk
Mon Sep 12 13:34:08 EDT 2005


On Mon, Sep 12, 2005 at 12:16:46 -0400, Greg Wilder wrote:
> > there is a still quite a lot of controversy over whether or not Sony's 1
> > bit/MHz clock design actually *is* better than more "traditional"
> > approaches. i know there is a lot of good marketing to explain why its
> > so cool, but that doesn't seem to be backed up by any substantive double
> > blind testing. even the math analysis is heavy on marketing bullshit.
> 
> Hmm - I didn't know that.  I'm no DSD expert, but I understand that
> there are some clear drawbacks.  For example, the DSD process introduces
> high levels of ultrasonic noise which can cause problems with certain
> playback equipment, and it is very troublesome to attempt DSP with pure
> DSD signals.  But if you're trying to accurately capture a "real-world"
> performance - it certainly is a strong contender.

Hum. I'm certainly no expert on this, but my impression is that the DSD
format is just an encoding of the intermediate signal produced by decent
ADCs before conversion into PCM and used in "digital amplifiers". The
advantage being that you save a conversion step in both directions - if
your source and final output are both analogue.

If you have to convert to PCM in either direction (say, conventional
digital processing in ardour or for output over SPDIF) then the advantge
would be wiped out in at least that direction.

It's not really new for what its worth, underneath its the same modulation
system that was used to encode laserdiscs.

- Steve 



More information about the Linux-audio-user mailing list