Well Wadia know?

[tony]

Posted my thoughts on last night over on the PC thread before I discovered this thread but agree with all above, and it was a very interesting and enjoyable night.
Thanks again Tony.

Thank you Pearse for bringing over the zuma yet again. That is the problem with having kit everybody wants to hear you have to bring it with you all the time!

[tony]

Listening with xstream now on dual laptop setup and would love to have the wadia for comparision. Simon just pick it up and drop it over whenever! Last night I switched from my setup to Pearse's and introduced the ciunas and as Pearse advised it needed fiddling to get working in ultrastream. Regardless the wadia sounded fantastic but when/if the rematch occurs against the meitner I would be interested to try briefly the ciunas with a settled dual laptop setup.

[Steve]

Yes, definitely agree that it was a very enjoyable evening! Many thanks Tony for generously hosting us and very much appreciated the insight into computer audio given by yourself and Pearse. It certainly left plenty of food for thought!
The Wadia did indeed appear to acquit itself very well. A stronger grip on the baseline with smooth but well textured mids that really invited the listener into the mix on a number of different tracks. The computer audio setup certainly gave a notably different presentation, particularly at the top end, where speed and decay of transients, on some tracks, appeared very fleet of foot.
Very interesting that all involved had tube amps along with 3 of 4 punters having Quad panels. So perhaps a certain blend of the older technologies coupled with the latest source options...
The discussion afterwards was a further highlight - always great to compare experiences.. It certainly left me with a lot of thoughts around converting to computer audio solution..
On that last point, has anyone had any experiences with, for example, Weiss 202, Berkeley Audio Alpha, Eximus dacs?
The only reason I haven't included the Meitner dac is that I'm hoping I'll be able to hear Pearse's Meitner along with the zuma in the near future and so could experience it myself!!

[jkeny]

Just back from a night out & saw this thread moved from the PC build discussion.

Great to hear about the differences - it fuels my interest & motivates me further.

Looked into the Wadia to see what might be at the bottom of the differences. It uses PCM1702 DAc chips, the last multi-bit chips manufactured & no longer made. This & the PCM1704 are recognised as some of the best multibit chips ever made along with Burr Brown PCM56 PCM61 PCM 63 PCM1702 and the TDA1541(S) Dac chip. Mulbiit chips are a different way of handling D to A conversion & represent a much more direct way of doing this conversion, using precision laser trimmed on-chip resistors to do the conversion. As the bit depth increased from 16 to 24 bits, the cost of manufacturing these chips became more & more expensive requiring more laser trimmed on-chip resistors & was the reason that the industry turned towards sigma-delta DAC chips. Sigma-Delta DACs use a variety of techniques to try to achieve the same performance with fewer precision resistors - they started with 1 bit devices but have moved to 4-5 bit devices. Through various techniques these 4-5 bits are manipulated to provide a similar performance level (on paper) to the 24 bit multibit systems they replaced.

The point of all this is that the manufacturers opted for sigma-delta & lower costs rather than sound quality. They can "prove" that the S-D DAc chips are as good as the M-B chips based on the specs :) But most people that hear a well implemented M-B DAC can hear the difference when compared to a well-implemented S-D DAC & this is what you guys have noticed, I reckon.

The typical differences noted are pretty much what you have identified - better grip on the bass & a more textured, more dynamic & textured mid-range.

Here's an interesting discussion about those differences for anybody wishing to follow this further http://www.stereo.net.au/forums/index.p ... ld-vs-new/

[tony]

Steve: Re Berkeley there is a lad with one on the forum but he lives in Hawaii I believe.
The Metrum Hex dac seems very interesting as well. It is priced similarly to the eximus without the preamp/headphone. It is one I would like to hear as it straddles the ground between the Berkeley/Meitner price category. He uses some industrial chips that nobody else does or is aware of. John will probably elaborate here.

John: That is very interesting, fair play to you for chasing it up and nice to know we were not just collectively imagining it. We were all very surprised on the night.

Looks like Simon\Steve have a very valuable scarce brick on their hands and should investigate the optional spdif card(I believe) Used with a JKCiunas SPDIF or similar maybe that is going to better the meitner.

[Diapason]

Very interesting, John. Is there anyone left pushing the state of the art in cost-no-object DAC chip manufacture? I know nothing about this stuff!

Steve, the boys in the Listening Suite (beside the Peppercanister Church) have Eximus DACs and I think another forum member was thinking of buying one? Anyway, it should certainly be possible to hear it.

[Fran]

DCS use a unique discrete solution as well I think. I have a pass d1 clone dac that uses pcm63k chips. While I say it's a clone, it's one built with nelson pass' approval/input.


Fran

[jkeny]

Steve: Re Berkeley there is a lad with one on the forum but he lives in Hawaii I believe.
The Metrum Hex dac seems very interesting as well. It is priced similarly to the eximus without the preamp/headphone. It is one I would like to hear as it straddles the ground between the Berkeley/Meitner price category. He uses some industrial chips that nobody else does or is aware of. John will probably elaborate here.

Yes, the Metrum uses some non-audio DAC chips which are multibit but it uses them in non-oversampling mode (NOS). NOS is an interesting but flawed concept, I believe. In it's purest realisation it entails no changing of the input sample rate (S-D DACs use 4X or 8X oversampling) along with avoiding a digital output filter. The avoidance of OS results in a roll off of HF starting down at 15KHz & the avoidance of output filter results in no suppression of the images that are always created in the frequency range 22-44KHz, 44-66KHz, etc. This is outside the audible spectrum but it can have an audible effect depending on how your downstream equipment handles such ultrasonic frequencies - some handle them elegantly, some do not. hence the Metrum is agreed by all to sound better once the 16/44 files have been upsampled to 16/88, 16/96 or whatever prior to input to the Metrum. There's a lot more to it than this but that's as brief a summary of some of the issues.

John: That is very interesting, fair play to you for chasing it up and nice to know we were not just collectively imagining it. We were all very surprised on the night.

Yep, many are surprised by the sound of this old technology. It begs the question about where the DAC chip manufacturers are leading us & why they use specs to cover up the issues when a lot of people who trust their ears & listen report that multibit DACs sound more dynamic than S-D DACs. There was an interesting video I posted a while back from ESS about this very issue & how ESS have attempted to address this. Successfully? I don't think so.

Looks like Simon\Steve have a very valuable scarce brick on their hands and should investigate the optional spdif card(I believe) Used with a JKCiunas SPDIF or similar maybe that is going to better the meitner.

[jkeny]

Very interesting, John. Is there anyone left pushing the state of the art in cost-no-object DAC chip manufacture? I know nothing about this stuff!

Nope, everyone has fallen for the specs & the ubiquitous DAC is now S-D. The only area where multi-bit DACs are still used is in industrial applications (for precision) but these tend to be 10, 12 14 bit devices with a few 16 bit ones & no 20 or 24bit ones that know of.