DSD DACs that do, and don’t

[One and a half]

Many threads at CA talk about clean power for computers or streamers, utilising linear power supplies (a most undeserved title since the current they draw from the mains is hardly linear). In an effort to reduce the noise floor from the computer to the DAC, reducing the amount of noise emanated by the computer is still a problem, however the PSU could shunt noise back to itself rather than down the USB pipe. Since there are countless other power supplies in a computer, there may not be much point.

 

 

There are as many threads concerning themselves with the effects of using different USB cables and how they affect the sound, most notably in the criteria for removing noise from the computer to play files, again with a reduced noise floor, enhancing low level details. There are those that would argue, they don’t hear a difference with USB cables with their DAC at all, despite attempts to try different cables. Perception and technical details of the cables aside for now, not going to fill another thread, but there’s something in common between using linear PSU and cabling.

 

 

Added to this complication is the method of transmission from the computer to the DAC. I don’t know all the ins and outs of the protocol, however let’s assume it’s a waveform of sorts with a given rise of voltage and time. Both voltage and time can vary, and that’s the problem, the CPU controls the cycles of data or whatever controls the bit(stream), with async USB at least the bits can be re-sent, but for DSD, how is this going to work effectively?

 

 

An interesting interview is available with the top engineers and directors at Accuphase conducted at a dealer’s opening in Belgium. Under discussion was why Accuphase has not produced a streaming product. Their reply was very interesting.

 

 

“Beneath the surface, our players have worked with a high-resolution internal data-signal for many years now, so at first it looked like a relatively small step to implement external data as well. But this turned out to be much harder to do than I expected. For internal data, we can use our own transports that generate a very accurate data stream. But when you use external data sources, that are usually much less precise, you run into a mountain of trouble. For some time it even looked like we would be unable to clean up the external data enough to get real Accuphase quality at the end of the process. That would have been a legitimate reason not to provide this service, but fortunately we pulled it off just in time. So there you have it: a small glance into the possible future of Accuphase Streaming Audio.”

 

 

Accuphase describe a difficulty to control external data, since they cannot control the data flow as accurately as they can from their own transport. With a transport, you can achieve 100% data extraction, but more importantly in the correct time, since the data flow can be controlled with precision internal to your own closed loop system, yes, it’s called feedback, and in this case it works. The way around the problem is to burn a DSD (DVD) disc to the Sony DSD_DISC protocol (not that complex), and use SACD player to play back native DSD. To burn a DVD is clunky, involves optical media, but their reasoning has merit.

 

With a computer, there’s no such tight tolerance at all and no feedback either, it’s an hourglass pouring in sand, pebbles, and rocks, mixed with water..! Notable DAC manufacturers such as Exasound, Meitner, Chord and Playback Designs have chosen an approach that comes close to control of data of what Accuphase identify, enough that it’s listenable, but more importantly, the noise/jitter/crud that’s transported along the USB is buffered, cleaned, isolated. Techniques vary of course between different manufacturers but would include some if not all these elements.

 

 

It is therefore immaterial to be concerned with linear PSU, USB Cables, USB filters (ifiUSB et al), provided the DAC that does clean up the front end. For DACs that don’t, they will all sound different, since the amount of noise is dependent on the computer, PSU, build, and how much the connecting cable allows the transmission of noise. This is even before we begin to draw attention to the common coupling point, the AC power supply. Another story entirely.

 

References:

 

Passion for Music – Accuphase interview

 

Exasound measurements – Galvanic Isolation

 

PDFAS (Playback Designs Frequency Arrival System)

[wisnon]

Reclocking Dacs that d the clean up...do they sound real or artificial? Does an exact facsimile equate to the real thing? Just asking.

 

In the end you hav to listen to these solutions to judge if you can live with the compromises.

[Hiro]

I guess that's where dedicated audio servers (with USB outputs) like the Auralic ARIES and Aurender X100 fit in... I'd think they solve the common computer-related problems wholesale.

[Panelhead]

Or playback from RAM. With the proper firmware and software this yields even better results than extracting from a spinning disc.

[bmoura]

I guess that's where dedicated audio servers (with USB outputs) like the Auralic ARIES and Aurender X100 fit in... I'd think they solve the common computer-related problems wholesale.

 

Right. That's part of the appeal of Sony's excellent HAP-Z1ES DSD Music Server. The files and DAC are all in one box. No USB and computer needed once the files are transferred.

 

I think we will see more products along these lines in the future. It definitely allows better control of the variables.

[Hiro]

Right. That's part of the appeal of Sony's excellent HAP-Z1ES DSD Music Server. The files and DAC are all in one box. No USB and computer needed once the files are transferred.

 

I think we will see more products along these lines in the future. It definitely allows better control of the variables.

 

Too bad though the Sony doesn't have a USB output like the two above devices. It could then be used with a stand-alone DSD DAC of your choice.

[bmoura]

Too bad though the Sony doesn't have a USB output like the two above devices. It could then be used with a stand-alone DSD DAC of your choice.

 

It could be used with a DAC of your choice. I find that most Sony HAP owners don't seem to have any need to use another DAC but the one in the server! And then there are always the upgrades from Red Wine, Modwright and others if you want to "hot rod" the HAP a bit....

 

Red Wine Audio Components

[One and a half]

Reclocking Dacs that d the clean up...do they sound real or artificial? Does an exact facsimile equate to the real thing? Just asking.

 

In the end you hav to listen to these solutions to judge if you can live with the compromises.

 

It's more of a consistency with the product. The DAC under test for now on the road for example, plays cleaner with an ifiUSB. The DAC shouldn't need this device, maybe on another source the ifiUSB is not necessary.

[One and a half]

Or playback from RAM. With the proper firmware and software this yields even better results than extracting from a spinning disc.

 

The control of data is in a closed loop feedback system is superior than an open ended data stream from a computer. It's just that CD players in the low bottom end market have very poor error correction, jitter et al and this is then applied to all CD players.

There are some threads here that still prefer the use of a tried and true CD (SACD) over computer playback. With DSD being so resolving, finer details are often lost in EMC muck which if unfiltered or untreated doesn't sound that great.

[Miska]

Right. That's part of the appeal of Sony's excellent HAP-Z1ES DSD Music Server. The files and DAC are all in one box. No USB and computer needed once the files are transferred.

 

That's just a computer + DAC in a single box. You can do the same with computer if you want to, get rid of the USB many ways. You can use either I2S straight or a PCI/PCIe sound card.

 

Aries is a computer with USB output, or alternatively S/PDIF that seems to be internally on USB (you can see the XMOS chip in the Aries internal photos).

[bmoura]

That's just a computer + DAC in a single box.

 

The way Sony has designed it is very clever. With the ability to change songs with just the turn of a knob, it feels much more like an appliance than a computer + DAC.

 

A much easier, more convenient way to listen to music for the non-techies out there.

[bhobba]

It is therefore immaterial to be concerned with linear PSU, USB Cables, USB filters (ifiUSB et al), provided the DAC that does clean up the front end.

 

I own a Playback Designs that does most or maybe even all those things. You can clearly hear differences in USB cables, as well as if you feed it via my Off-Ramp or use its internal USB. Being an upmarket DAC the Off-Ramp doesn't make as big a difference as lesser DAC's - but a difference there is.

 

EMM makes a DAC they claim totally strips jitter - and I have no reason to doubt the claims. But its the same with that DAC as the PD.

 

Why? Well there are guys like Steve Nugent that makes the Off-Ramp, and he is one of the most qualified digital engineers on the planet, being the team leader that developed the Pentium, and sill holds a number of key patents. He has posted a few things such as common-mode rejection that's involved, but evidently there is more to it. However knowledge of those things is what gives him a competitive edge in the marketplace so don't expect him, and others in the know, to tell all any time soon.

 

That said - why worry. Simply listen and make you own mind up - let the engineers worry about the technicalities.

 

I only know one DAC that is USB cable immune - its the Phasure. But the guy that designed it went to uber lengths to achieve it - eg its own separate low noise power supply powering just the USB.

 

Thanks

Bill

[One and a half]

I own a Playback Designs that does most or maybe even all those things. You can clearly hear differences in USB cables, as well as if you feed it via my Off-Ramp or use its internal USB. Being an upmarket DAC the Off-Ramp doesn't make as big a difference as lesser DAC's - but a difference there is.

 

EMM makes a DAC they claim totally strips jitter - and I have no reason to doubt the claims. But its the same with that DAC as the PD.

 

Why? Well there are guys like Steve Nugent that makes the Off-Ramp, and he is one of the most qualified digital engineers on the planet, being the team leader that developed the Pentium, and sill holds a number of key patents. He has posted a few things such as common-mode rejection that's involved, but evidently there is more to it. However knowledge of those things is what gives him a competitive edge in the marketplace so don't expect him, and others in the know, to tell all any time soon.

 

That said - why worry. Simply listen and make you own mind up - let the engineers worry about the technicalities.

 

I only know one DAC that is USB cable immune - its the Phasure. But the guy that designed it went to uber lengths to achieve it - eg its own separate low noise power supply powering just the USB.

 

Thanks

Bill

 

This reply is similar to this post at least you are consistent. For the life of me, I would have changed at least five USB cables on the MPD-3 and the sound is much the same if not identical. Adding an ifiUSB and purifier, made things worse. By accounts using an OFF RAMP, with an external power supply, an extra cable, and using S/PDIF with no clock master, you're saying the sound is better when you add in more connections to the source, rather than use a single USB data line?

 

This implies the USB on the Playback to be absolutely dreadful, which is quite contrary to what I hear, for example brushes on Shelby Lynne's "Anyone who had a Heart" are clearly defined as such and not castanets like some DACs I hear, and this is using the 'bad' USB input. Besides, S/PDIF won't pass DSD through the MPD-3, so the discussion is rather pointless from there on in.