[Harmon]

I have been reading lately that CD players that do not employ oversampling or upsampling in there design sound better than players that do?  What do you think?

[Sa-dono]

I would say you are proceeding on shaky ground, and advise people to continue cautiously.

As with everything, there will be good and bad implementations of either design. It is best to go out and listen to as many versions of both, as you can, and decide for yourself.

[_scotty_]

To state the obvious,whatever technology sounded closest to the live mic
feed during playback could be considered the most accurate. This kind of comparision has been done with SACD and DVD-A but I cannot remember seeing any test of a playback system without a digital filter in the chain.

[Carlman]

This thread reads like a troll but, I'll assume it's not... and will say this has been discussed quite a bit.  

My experience is that it's not better or worse overall, it's better in some areas and just a different kind of sound.  I've only heard one person say they can identify this difference sonically.  He will probably chime in on this thread.

If you were to change the settings on a DAC... like if there were a switch to go from filtering to non-filtering; you could probably hear a difference but would you be able to identify oversampling as the change?  Could you label it and determine this for future reference?  I doubt I could and I hear very well.

I think Scott Nixon provides this feature now on his DAC's.  However, his DAC's have their own personality that you have to identify first.  

So, isolating oversampling as a feature to disect seems like really fine hair-splitting to me.  Which is partly why you'll get a lot of 'you have to listen' responses.  Because it's true.

-C

[mb]

Hear, hear! An excellent, balanced response from Carlman.

A bunch of local DIYers got together and listened to a number of non-OS dacs -- DIY as well as a fully spec'd SN TubeDac+. The long and short of that session is that implementation seems like a much larger determinant to "good sound" than the chip (1543 or 1541) or filtering (compared to the host's nice Sony cdp).

I've also been tuning a number of dacs + cdps: old Philips bitstream, recent NAD 521i, Channel Islands VDA-1, classic Proceed dac. Invariably, the age of the design and technology (16 bit vs. latest 24 bit, bitstream vs. sigma-delta vs. multibit) seems to be much less of a factor vs. quality of power supply, attention to grounding, and component in critical sections.

Right now, my preference for the various units is:
1) VDA-1
2) My non-OS parallel TDA1543A dac (battery powered)
2) Philips CD750 (budget SAA7350 bitstream dac)
4) NAD 521i (PCM1710 budget 20-bit dac)
5) Proceed PDP-2 (PCM58)

All units have been modded. In the end, implementation is very important (the VDA-1 was always the best sounding, even before mods), but modest budget technology like the old Philips bitstream has very surprising potential.

[Alex Peychev]

Yeah Channel Islands VDA-1 sounded good. If I am not mistaking it uses Burr-Brown Delta/Sigma (bitstream) DAC. I also think that its interpolation filter runs at 8X OS.

PCM needs to be "smoothen out" by converting it to Bitstream... It works!...

Of course, most of the current Delta/Sigma (bitstream) DACs use from 8x to 128x OS. My all times favorite Crystal DACs run CD (16/44.1) at 128x.


Regards,