I agree that the Oppos are great bargains. I've heard a 93 and a 95SE and they sound good, but I preferred my then source (Sony SCD-XA5400ES) for SACD and Redbook, and then moved to the PWD then upgraded that, and the PWD was so much better a source (and also preamp) than the SCD-XA5400ES. . . . The Digital Lens in there alone makes a significant difference, and the output stage is excellent. I think the PWD Mk II is in another league, and ought to be, it's frickin' expensive.

Will, the 95/105 series are, in my view, on an entirely different plane in audio quality, and the DAC sections are years ahead of the 83. Your Tranquility is bound to be much better on that basis alone. The 105 really could be your only front-end, though. At least, for a while. The problem, of course, is that the DAC technology arms race is going to obsolesce equipment rapidly, so any "investment" in non-upgradeable hardware really is just sunk cost. Sit back, enjoy what one has, until something comes along sufficiently compelling to pry open the wallet. One of the great things about our Decware equipment is that it is more resistant than most to obsolescence.

And yes, Lon, you are right, the PWD ought to be good for its price. And is.

Interesting! I have been playing phone tag with him. Had the feeling something was up. I agree. If they decide to put it out, it is very likely to be time-proof for the most part.

Makes sense Fireblade. The price gap is indeed significant. And while there are less expensive DACs that do native DSD, one thing I would not want to lose is the quality of PCM I already get out of the PWD. The Invicta would clearly impose no such compromise.

I don't know, Oppos really haven't appealed to me, but they do look like big bang for buck if they sound better now than what I heard before, and if you really want computer audio et al. (I don't).

For me, with a lot of time now with both Mk I and Mk II PS A PW DAC, the digital lens in this (and in my PW Transport) makes the difference. I wouldn't change DACs until there's a stand-alone Digital Lens to use in concert with the DAC. Those lenses really make a difference. (And as Greg mentions, coupled to the killer analog output stage of the PW DAC). For Redbook and for my video sources that's the sound still amazes me and I can upgrade around it and receive sonic improvements.

More and more I find vinyl to be the most pleasant listening experience and I anticipate spending more and more time spinning records. Thanks in part to the ZP3. What a machine.

I don't blame you, I would if I were at all interested in digital files, but they actually repel me, I'm not ever planning on going that route.

Spinning a record I bought thirty-five years ago and it never sounded this good! I am loving vinyl. If I didn't have 15,000 plus cds I'd probably be an analog only guy now.

:D

Me, if I were a better musician I'd probably just have a boombox.

Lon quipped:

Quote:


Good words Lon; wouldn't we all!

beowulf asked:

Quote:


I have narrowed my DSD-capable DAC down to two: the Mytek and the Invicta [price gap of over 100%]. I wish PS Audio would roadmap a DSD-capable PWD, but Paul has explained several times he does not intend to. Sooooooo, looks like we're headed that way. I already have almost 300gb of DSD files, so this is starting to become a no-brainer. So, anyway, one answer is that, if PWD had a DSD-native roadmap, i wouldn't even be considering this.

The other part of your question, the connection, is interesting. Invicta can use SDXD cards, which theoretically can hold 2tb of files, but commercially, I think only 256gb ones are available, and they are pretty expensive. So, yes, in all likelihood, USB or SPDIF will be the connector. I wish I had more confidence in the Lumin network player, because from a feature set, it is pretty complete.

I got a chance to hear some native DSD recordings on a fine system through the fairly reasonably priced Mytek 192, and I have to admit, it was very impressive. Whether there is enough prospect for commercial success of the format remains to be seen, but there is enough good material out there to make one want to play it. And I have almost 300gb of DSD files that I would now prefer to play natively, rather than converted to PCM as I now do.

Based on the current review of the exaSound e20 in The Absolute Sound, I have added it to my list. That review sums up nicely the significant source material advantage that DSD files have over even hi-res PCM files.

All these mid-to-higher-end DAC makers offer 30-day in-home trials, which seems a good compromise to address the widening product base, in the face of a shrinking retail market and networks. I plan to start with the exaSound.

Fireblade, thanks. As noted above, the Invicta is already on my list. And it does already support native DSD.

Hey, I appreciate the input. Keeping up with this stuff can run one ragged. Cheers!

Thanks beowulf. I know about the Lumin; in fact I mentioned earlier that I had taken it off my list, because I did not feel confident in the maker. Part of that lack of confidence came from the manner in which one of their astroturfers completely carpeted an audio forum with promotional posts about the product. Part of it came from the "promises everything including a better sex life" approach. But if you read where they are from actual users, they have a ways to go. The 6Moons review indeed illustrates that fact as well; there are many "gotchas" waiting for the early adopter.

I used to put more stock in 6moons, but over time, I have gotten tired of wading through their purple prose that passes for erudition, and the endless gobs of data that does more to mask the lack of meaningful information than anything else. Still, it is a very positive review, and I will continue to watch the Lumin to see if it continues to make progress. I am marginally concerned about its reliance on the minimserver architecture. I have downloaded and tried minimserver as a server for my PWD, and I was less than impressed. 6Moons pointed out some of the idiosyncrasies of the Lumin software ecosystem, and it well illustrates the adage that the 'U' in 'UPnP' doesn't stand for what you think it does. Software development seems to be the bugaboo of many of these devices, and it is a significant challenge [witness the JRiver offering of a Mac OS X version of their media center]. I suspect that like most such things, long term success will depend in great part on market traction.

I also have learned the hard way that as great as network streaming can be, and as problematic as USB can be, neither is good or bad, just different. My PWD over network can be great, or over USB, because it really solves USB with the Digital Lens, but over network, it can actually be occasionally flaky due to software interactions and the challenges of gapless playback. PS Audio is working to solve all that, but it's something of a work-in-progress. So, ironically, USB is a rock-solid alternative on the PWD, and many other DACs.

On the Mytek, which supports USB, the alternative is Firewire [a la the Weiss and others]. Firewire is not as difficult as USB, but when properly implemented, both sound very similar. The Invicta offers an SD card interface, which solves the computer interface problem altogether, albeit on a smaller storage scale for now, but also provides very good audio performance.  The exaSound addresses its jitter issue differently, but quite effectively.

I am starting out my trials with the Mytek, in part because Blue Coast has a great promo [same 30-day trial, plus a $200 credit for DSD downloads], they are local in the Bay Area, they and Sony use the Mytek as their demo rig, the aforementioned Firewire interface is nice on the Mac, and I expect it will make a nice headphone DAC/amp for my desktop computer, if I eventually settle on a different machine for the main listening station.

beowulf wondered:
Quote:

McGowan has recently indicated that DSD may be on the update/upgrade path, and from a chipset architecture perspective, it is certainly feasible. But I have no doubt there are challenges, so I am not expecting it "any day now." I sure wish they would, because as far as SQ goes, I haven't had any other reason to seek out alternatives or additions to the PWD.

Brett,

What USB to SPDIF solution do you use?

Thanks,

Will

@beowulf: Interesting comments from Chord, so it sounds like they were talking about their stuff, not the industry in general. I also did not realize they were dependent on the Twonky ecosystem, so that makes sense. But I think Colin Pratt is spot on with this; my network is now feeding some large hi-res files very reliably, with increasingly rare hiccups:
Quote:

But it is true that in an abstract sense—as Brett notes above—the data problem is a data problem whether over USB or Ethernet. In my case, I need to tweak my network layout to make sure the full audio chain is on its own little subdivision of the network, and not bothered by other stuff, and generally going through like switches. This is analogous to the problem of choosing and setting up a computer to not be distracted by other stuff like keyboard interruptions, WiFi, background processes, etc., so it does not add to the inherent USB problems. Conceptually similar, but pragmatically different with very different solutions [though all benefitting from a musical DAC with sophisticated jitter reduction capability].

@Brett: Nice rig. I really like the Empirical Audio line, and their approach to the computer audio field. BTW, for some very good "starter" tips, check this page at their site, including this somewhat surprising assessment: "Mac Mini has the least problematic USB ports." I demo'ed both an Off Ramp and the Pace Car [replaced by the Synchro Mesh re-clocker]. I thought both were pretty good, but I ultimately preferred the Audiophileo at that time going into my PS Audio Digital Link III [itself heavily modded, and a very very good DAC], and eventually ended up preferring the approach of the PWD altogether for solving the clocking problem. Rather than re-clocking, I chose what I would call PSA's "de-clocking" approach with the Digital Lens.

Not sure I would go as far to say: "the data is coming through an ethernet pipe or a USB pipe the problem as it relates to audio quality is essentially the same." I think this is conceptually correct, because computer data streams in general are not designed for music playback, and every version of those data streams requires some serious "reconstructive effort." But the two approaches are very different, and USB is a more difficult problem to solve in my view. And with all due respect, USB->SPDIF is not state of the art, but given the improvements in the 5th generation of the OR, my guess is that it probably is the best of that class. Like I said, I am a fan of EA, and their engineering but USB->SPDIF is part of the problem itself. Indeed, if one can use the Off Ramp in HDMI I2S mode to a DAC that has that protocol [PWD, W4S], it's going to provide a better result. I never got to try this with my PWD, because I was moving away from USB in my move to the PWD, but given the approach of the PWD, I would bet the OR impact would be less, if perceptible at all. In theory, any jitter reduction upstream should be a good thing, but if one component actually removes the computer-induced jitter altogether, irrespective of whether the computer is a good or bad USB source, does it matter if an additional in-stream device [the OR] has made it better before it gets to the Lens? Don't know. I am a big believer that "everything matters," even if I cannot hear it, but my wallet cares. This very question was raised by John Darko in his Digital Audio Review of the PWD:
Quote:
My guess Brett is your answer is to divide the component requirements up into manageable options and invest in the best of each that one can afford, e.g., transport, reclocker, DAC, etc., because, as you say, there is no one perfect do-it-all box. I think this is a valid strategy. The Off Ramp solves a problem that only exists if one is using USB playback from a computer, and one does not have a digital lens or its equivalent. Of course, if you have a wonderfully musical DAC, that is not async, then the OR or something like it is absolutely essential. I had a delightful Promitheus DAC, NOS, and it sounded beautiful. It benefitted significantly with outboard USB control. Same with the PSA DL III.

Beowulf,

the OR5 is hooked up to my Win 7 PC using a standard belkin 5ft usb cable. It sounds best when running Foobar 2000 with the kernal streaming plugin. Streaming from the kernal bypasses the audio manipulation of the Windows 7 mixer. Most of my audio files are ripped using exact audio copy to uncompressed WAV files.

I've got a lot of FLAC files as well and convert those back to WAV before listening through Foobar. Foobar can play FLAC files on the fly, but interestingly this does not sound as good as playing WAV converted from FLAC. Low bitrate MP3s sounds the worst, but all in all it is a matter of detail loss. Nothing coming out of this setup ever sounds harsh or irritating.

As for the computer itself, I got a wild hair and the crazy notion that it would be a good idea to build a computer that could run Crysis as well as pull double duty as a machine quiet enough to be worthy of a serious audio room. Obsession took hold and what follows was the result.



The first order of business was to consider a way to provide adequate cooling and ventilation while keeping the noise down. It was decided that water cooling was the way to go for most of the computers heat generating components. Still, within the case would be two fans and a water pump. The HDD also was a noise consideration.



My solution was to design an intake and exhaust muffler so to speak. The chambers above and below the large middle chamber are the intake and exhaust mufflers later to be lined with sound absorption foam. You can see two fan holes on the bottom right and top left.



The side panels bolt into place with countersunk weather strips to keep it air tight. When closed the side ports are the only way for air to get in and out.



The primary method of heat exchange is done by this little radiator. It has four quiet fans, yet they are noisy enough that I decided to place it outside of the audio room. It sits mounted on the wall in the hall way and is inaudible from the room. Kind of neat to pump the computers heat away via water like that. BTW this computer is a space heater. The power supply is by necessity 1000 watts. The audio room stays nice and cool though.



Here is the right side with adhesive acoustic foam placed on the door and inside the case.



Here's a larger pic of the innards http://i.imgur.com/UWfnpWj.jpg

Even the power supply, motherboard, and RAM is water cooled. Yet still the internal temp sometimes reaches 90F. On board sensors are monitored and fan speeds depend on how warm it is. I have two saved bios modes. One where everything is at normal clock speed and minimal voltage for keeping it cool and quiet. The other setting is super blasto overclocked for running games and stuff. Even at full blast this computer stays cool and is relatively quiet.

Weighting around 100 pounds this computer is easily the least practical thing I own. It's actually the second computer case I've built using wood (in this case maple). Some day I want to build another using CNC and good plywood.






The OR5 is powered by a wall wart which is plugged into a Power Plant Premier. It does it's thing and pushes SPDIF through a DIY silver interconnect with BNC jacks. VHAudio has an terrific 75ohm silver coax cable at $100 a meter.



The digital input of my DAC has a silver wired torroid transformer with a 100v teflon  protection cap. The input transformer feeds a CS8412 or CS8414 receiver chip. The DAC has two switchable inputs each with their own transformer.



Audio Note DAC Kit 1.2

Larger picture http://i.imgur.com/iMMhU2z.jpg

I bought it about 8 years ago and have studied and modded just about every aspect of it. I practically have the schematic memorized now.

The power supply feeding the digital board is all Black Gate FK and N caps and Cree diodes. The power regulation uses ultra low noise AD797 op amps. I rolled through about 5 different kinds of opamps and surprisingly they have quite an effect. The AD797s really cleaned everything up and made for more dynamics.

The cans between the digital board and tube output stage are silver wired transformers which serve to convert the current (rather than amplified voltage) outputs from the DAC chip into voltage (I/V conversion) for the tube grid. This way from the moment the digital stream is converted to analog there is no transistor amplification. The DAC chip has no oversampling and no digital brick wall or analog filter. The limited bandwidth of the I/V transformers acts as a filter to prevent ultrasonic distortions. They also provide galvanic isolation keeping grounding paths pure. Adding these brought the DAC to a whole different level.

The tube output board power supply uses a 6x4 rectifier tube, polypropylene filter caps, choke, and gas tube regulators. The tube filaments are heated with DC.

The tube output stage has 6922s running in SRPP. The 3 volt output is hot enough to drive my Zen directly.

All teflon coated silver plated copper hookup wire, and cotton insulated silver signal wire.

Pale-Rider,

I like the way you mention conceptualized. I think with so many sources of information about audio that conceptualizing audio systems in to granular processes is the best way to cut through a lot of the BS we are subjected to in reviews etc.

I'm not privy to the the real world shortcomings of USB vs Ethernet as it relates to providing data to the clock. Conceptually though, it would seem the job of the USB or Ethernet receiver (on the master clocking component) to buffer the incoming data so that it is then presented to the master clock in such a way that it can perform it's job well.

In other words if the receiver is doing its job, then the master clock will behave the same regardless of how the data was received.

Another way to look at it is to consider that all digital clocks that perform according to the SPDIF standard are essentially the same. Some are simply better at timing than others. Some claim that a CD transport is the only way to go, but even here data from the CD drive is received, buffered, clocked, and then transmitted as SPDIF. And knowing that makes the concept of belt driven CD transports seem a kind of silly. After all a CD is not very well balanced and any vibration is a huge deal when we are talking about accurate timing in pico seconds. Thus the necessity of the buffer.

A pico second is one millionth of one millionth of a second.

For reference the OR5 master clock keeps timing accurate to within 390 pico seconds. 250 pico seconds if using the I2S output.

The main reason that jitter is reduced in I2S is due to the data and clock being transmitted on separate channels, where as SPDIF is sent as an infusion of data and clock on a single channel. The added complexity of processing into SPDIF fudges the timing a bit.

Re-clocking and digital lenses are band aids that are often necessary but to be avoided if possible. Assuming you have a happy master clock, the digital stream from that point on should be very carefully preserved all the way to the DAC chip. By carefully preserved I mean minimal alteration or interaction.

Brett,

Great work and achievements, congrats!

I'm curious why you don't take advantage of the latest version WASAPI Event style Plug-in for FB2K, designed precisely for lowest latency, best sound streaming from FB2K in a Win 7/8 based PC environment.  As in Kernel streaming, it bypasses the mixing processes from Windows altogether, allowing for an external DAC's exclusive mode of utilization.

I've tried several approaches (not Kernel, though), and the Event Style is the best I've found. It allows for very low buffering, therefore virtually eliminating latency issues (important, especially for a laptop), and the sound quality is just great (other things being equal).