I'm not sure about the performance of the PSA unit, but Musical Fidelity's X-DAC also features a USB input.

Regarding wifi, I find it interesting that Linn has chosen not to include wireless capability in the new Klimax DS streamer. They claim that wired ethernet sounds better, which I suppose makes sense. Also, Linn clients probably already have structured wiring in their homes.

I use an Apple Airport Express (connected to a DAC via optical) for streaming, and the sound quality is indiscernible from CD for me.

I've never tried USB for playback, but I'd really like to, especially since Macs have upsampling capability.

Quote:
Regarding wifi, I find it interesting that Linn has chosen not to include wireless capability in the new Klimax DS streamer. They claim that wired ethernet sounds better, which I suppose makes sense.

Actually, that does not make sense, (at least not to me). 802.11b or g (wireless wi-fi protocols) are reliable protocols, which means that they contain error correction, checksum and retransmit flags within its specification. Therefore, the digital data will either be delivered fully in tact or it won't be delivered at all just like Ethernet. Hence, there is no sounding better or worse than Ethernet. Either there's going to be sound, which means packets will be delivered fully in tact or there won't be sound at all. There will be nothing in between. At worst, you will have dropoouts in the data stream, but there won't be any sound degradation due to using wi-fi.

I guess it makes more sense to someone with less sense.

Quote:
I guess it makes more sense to someone with less sense.

Sometimes less is more.

Quote:
I'm not sure about the performance of the PSA unit, but Musical Fidelity's X-DAC also features a USB input.

In addition, there are also the various Headroom units, either the stand alone MicroDAC or the internal DACs available on their Desktop, Home and Max amps - all have USB inputs.

Based on everything which has been on this forum plus the numerous equipment reviews which have been published in various magazines it seems that at the present time the best way to use a computer in an "audiophile" music system is to stream the music via one's wireless network.

And the best way to stream music, IMHO, is to use the SlimServer software and a Slim Devices Transporter. Sure the Transporter doesn't have the blessing of all the high end audio magazines (see the December 2007 issue of The Absolute Sound, in which their survey of PC music servers completely ignores the Slim Devices products) but it's by far one of the best servers out there and compared to some of the other high end music servers, one of the most cost effective.

All DAC's will do a nice job when connected to a PC via an audio card with an S/PDIF out.

All those tested so far that I have seen seem to have difficulties with data arriving via USB.

I wouldn't so easily dismiss that hardwired sounds better than wireless. While it "shouldn't" make any difference, data arriving by USB also "shouldn't" be any different. Yet it is and there are apparently technical reasons why this is true.

My impression is that we are all still learning how to transmit data effectively for audio purposes.

That being said, I wouldn't hesitate to buy a wireless system as they can sound superb - that is, as good as the DAC that does the conversion from digital to audio.

Elk,

The reason USB sounds different than SPDIF is because it's a PCM protocol being sent via the USB interface and this protocol has not been encapsulated properly within USB to ensure reliable clocking.

When we talk about wireless wifi vs ethernet, we're not talking about streaming PCM. Rather, we're talking about delivering TCP datagrams via the IP protocols. Same as browsing the Internet. Wi-fi has been developed SPECIFICALLY to handle the TCP/IP protocol, hence it doesn't fall into the trap of implementation and modification of the interface to handle a non-native protocol.

In other words, as far as wi-fi is concerned, it does not distinguish between streaming audio and browsing the Internet. The audio streams are encapsulated within TCP and UDP datagrams. Therefore, there can be no sound degradation. You either hear it the way it is sent or you don't hear it at all.

This is a great example as to why I state that hardwired and wireless "shouldn't" sound different.

Yet, we have learned that accurate data transmission is not the same as transmission of data resulting in great sound.

Accordingly, I simply feel we need to be open to the possibility that hardwired sounds better than wireless.

I haven't experienced a difference, and have heard excellent wireless, but I am unwilling to definitively state that there is no difference. I don't have the unquestioning faith in digital that some others do. YMMV.

It would be fun to learn why Linn takes the position that it does.

Hi Elk,

In this particular case, it's not a matter of wireless "shouldn't" sound different. In this particular case, wireless CANNOT sound different.

As I've stated before, you're not transmitting a PCM signal over wireless. (if you did, I would agree with you that there is quite a good chance that it will sound worse.) However, in this case, you're transmitting ONLY TCP/IP over wireless and THEN you're transmitting PCM to your DAC through a WIRED (albeit not Ethernet)internal, SPDIF or USB connection.

Quote:
In this particular case, it's not a matter of wireless "shouldn't" sound different. In this particular case, wireless CANNOT sound different.

Nor can amplifiers, interconnects, etc.

Perfect sound forever!

Sorry, but we have heard such pronouncements too many times which have later turned out to be wrong.

I prefer to keep an open mind.

Time and experience will tell.

Ignorance is bliss.

Merry Xmas.

Quote:

Quote:
In this particular case, it's not a matter of wireless "shouldn't" sound different. In this particular case, wireless CANNOT sound different.

Sorry, but we have heard such pronouncements too many times which have later turned out to be wrong.

Something I offer up vis-a-vis WiFi connections from Ayre's Charlie Hansen, which I have no opinion of myself, is that systems will react differently to being bathed in 2.4GHz energy. Although the level is low, audio circuits can be terribly non-linear in this region. RF energy that fings its way into the circuit's input via the feedback loop will therefore have an unpredictable effect. In a system that is vulnerable in this manner, Ethernet might sound superior to WiFi.

John Atkinson
Editor, Stereophile

Quote:

Something I offer up vis-a-vis WiFi connections from Ayre's Charlie Hansen, which I have no opinion of myself, is that systems will react differently to being bathed in 2.4GHz energy. Although the level is low, audio circuits can be terribly non-linear in this region. RF energy that fings its way into the circuit's input via the feedback loop will therefore have an unpredictable effect. In a system that is vulnerable in this manner, Ethernet might sound superior to WiFi.

John Atkinson
Editor, Stereophile

Hi John,

Something like that would be very simple to test: Put a wireless 2.4ghz phone next to the DAC and see if there is any interference. The phone will have much greater RF than a wi-fi device. So, if a DAC reacts, then it needs to be shielded better. This is one of those rare instances where there doesn't need to be much guessing.

Quote:
John Atkinson based his opinion assuming that WiFi is already not a part of our homes.

Or our neighbors' homes! Scanning my 'hood this evening from my listening room, I can log on to 4 other Wi-Fi networks other than my own. :-)

But I was offering Charlie Hansen's position as a datapoint, is all.

Quote:
Apart from argument of Ethernet vs. WiFi, am I correct to believe that at this time the best way is to connect DAC via TosLink/Coaxial and not USB?

Yes indeed, if you have a soundcard using the ASIO protocol. The S/PDIF protocol was designed by "real audio engineers" whereas USB, like all things computerese, was designed by computer "hardy boys," who regard audio as a trivial problem not to be taken seriously.

Not kidding. One reason I recommend the Squeezebox/Transport and the Sonos gear so highly is that you don't have to make any apologies for their audio performance, whereas the Roku Soundbridge, for example, was clearly designed by engineers unfamiliar with what matters when it comes to optimal audio performance.

John Atkinson
Editor, Stereophile

John,

It seems to me that if SPDIF was designed by "real" audio engineers, they would have included proper clocking and error correction within the protocol so that we wouldn't spend the next 30 years trying to figure out how to get around the clocking issues.

Quote:
Ignorance is bliss.

Why is this gratuitous personal attack warranted?

How did it further the discussion?

<sigh>

Quote:
It seems to me that if SPDIF was designed by "real" audio engineers, they would have included proper clocking and error correction within the protocol so that we wouldn't spend the next 30 years trying to figure out how to get around the clocking issues.

S/PDIF was designed when most thought "bits is bits"; as long as the data gets there it doesn't matter how precisely in time it arrives.

We subsequently learned that there are other factors that influence accurate sound reproduction, such as jitter.

We similarly may subsequent learn that hardwired connections offer benefits that wireless does not.

Science and human understanding is not static. There remains much to learn.

Quote:

Quote:
Ignorance is bliss.

Why is this gratuitous personal attack warranted?

How did it further the discussion?

<sigh>

It's not a personal attack at all. I have tried to explain to you how wireless data communication works. You stated that regardless of how it works, you choose to believe that it works some other way. There is nothing more for me to say in that respect. There is no magic to wireless or wired communication. If you choose to believe that there is magic involved, then that's your choice, to which I say that ignorance is bliss. It makes life easier, I guess.

Quote:

S/PDIF was designed when most thought "bits is bits"; as long as the data gets there it doesn't matter how precisely in time it arrives.

We subsequently learned that there are other factors that influence accurate sound reproduction, such as jitter.

We similarly may subsequent learn that hardwired connections offer benefits that wireless does not.

Science and human understanding is not static. There remains much to learn.

Clocking is nothing new in communication protocols. It's been around for a long time before CD's and sPDIF were developed. The only reason why I think that clocking would not have been included in the original sPDIF spec is because they didn't have enough bandwidth to accommodate it. In which case, they should have redesigned the interface.

Re: hard wired connections certainly offer benefits that wireless does not such as a more reliable connection over greater distances that prevents frequent requests for retransmits, thus reduces a chance of dropouts, which would sonically manifest themselves as cut outs in the audio stream. However, once the stream is established, it's the same as a wired stream from a quality stand point.

I see all this talk about audio over wireless v. wired...audio over USB v. optical...but no talk about the audio PLAYERs that were used on the computers! While I'm no Mac expert, I am understanding that audio in Windows XP is literally mutilated unless some steps are taken.

Windows has what Microsoft calls the kmixer. All audio routed through Windows runs through this kmixer. This kmixer effectively up or down converts all audio signals to 48k. If Windows Media Player is used to playback audio files, there's really no way around this if going through a USB cable to a DAC (we're not talking about ASIO sound cards here). There are various software solutions out there that have the ability to utilize plug-ins to enhance the audio. Some such players are WinAmp and Foobar2000 and bit-perfect/bit-accurate plugins.

These plug-ins and certain audio chips allow for truly bit-accurate playback...something Windows Media Player does not allow for.

Now, Windows Vista is a different situation as apparently Microsoft has completely rewritten the audio engine to be above and beyond what XP ever was able to be. However, I refuse to use Vista for various reasons so I don't bother myself with these items there.

I find it interesting that literally no mention is made of the METHOD of playback via the computer in such a technical magazine and such a technical review? Why is this? Was this overlooked for some reason?

Curious more than anything as I believe it's a VERY important part of an accurate review of ANY audio via a computer.

The Absolute Sound had a really terrific article a few issues back on this same topic.

Quote:

Quote:
It seems to me that if SPDIF was designed by "real" audio engineers, they would have included proper clocking and error correction within the protocol so that we wouldn't spend the next 30 years trying to figure out how to get around the clocking issues.

S/PDIF was designed when most thought "bits is bits"; as long as the data gets there it doesn't matter how precisely in time it arrives.

This was really my point. Yes, S/DIF and AES/EBU (AES3) have problems due the data being embedded in the bit clock, but the bits that come out at one end are the same that went in at the other. That was a given with the design of the protocol. This is not necessarily the case with USB-connected DACs.

As Wes Phillips noted on our home page a week ago, the only way of guaranteeing this with XP (other than kernel streaming, which is not for the faint of heart) is to set the audio data format to 24 bits even with 16-bit information. Windows will still mess with the two LSBs, but as these are now empty with 16-bit audio, the bits transmitted vis USB are indeed the correct bits.

John Atkinson
Editor, Stereophile

Unfortunately, I'm not subscribed. Do you happen to know which issue as I'd like to purchase the pdf version from their site?

Quote:
Unfortunately, I'm not subscribed. Do you happen to know which issue as I'd like to purchase the pdf version from their site?

I'll have to look for it. It was either last month's issue or the month before... Once I find it, I'll post the info.

Quote:
It's not a personal attack at all. I have tried to explain to you how wireless data communication works. You stated that regardless of how it works, you choose to believe that it works some other way.

I stated no such thing.

You may wish to review both my comments, together with others that provide possible explanations as to why there may be differences in sound between the two formats.

We are simply acknowledging the possibility that there is a difference. No magic invoked or implied.

Quote:
There is nothing more for me to say in that respect.

Thus, this would have been a great place for you to to stop rather than go on to mischaracterize and negatively label another's thought process and knowledge.

Back on topic:

As a practical matter Linn may also not want to spend time troubleshooting customer wireless networks. Linn is a high customer service operation that sets up its servers in its customers' homes. Most problems people have with their home networks are with wireless systems. Linn may not want to guarantee the operation and sound of their equipment if operated over wireless.

Also interesting is that the Linn unit does not include digital inputs of any kind other than ethernet. Thus it cannot be used as a DAC in the conventional sense.

Quote:

Back on topic:

As a practical matter Linn may also not want to spend time troubleshooting customer wireless networks. Linn is a high customer service operation that sets up its servers in its customers' homes. Most problems people have with their home networks are with wireless systems. Linn may not want to guarantee the operation and sound of their equipment if operated over wireless.

Also interesting is that the Linn unit does not include digital inputs of any kind other than ethernet. Thus it cannot be used as a DAC in the conventional sense.

That sounds a lot more plausible. What gets my goat is that if this is indeed the case and they simply don't want to bother with wireless troubleshooting, all they have to do is just say so, rather than start spewing out BS with regard to sound quality.

As far as labeling is concerned, if you don't want to be labeled, then I strongly recommend agreeing with everything I say. A general nod in my direction will go a long way.

Quote:
As a practical matter Linn may also not want to spend time troubleshooting customer wireless networks.

I was speaking to Gilad Tiefenbrun about this yesterday, as I was checking some technical aspects of the Klimax DS. He said that while they feel a Wi-Fi network on its own has sufficient throughput to work with hi-rez data, it gets too squirrely once you take into consideration that in a town, there may be several networks competing for airspace in your home. As I said in an earlier posting, I can log on to as many as 5 neigboring networks from my listening room. A wired connection gets rid of all those potential problems.

John Atkinson
Editor, Stereophile

That makes perfect sense. They don't want to troubleshoot network connectivity issues. They shouldn't say that wired Ethernet SOUNDS better than Wireless. That's completely misleading and it causes Elk to start believing in Santa Claus.

uuhhh...pardon my seeming stupidity..but..what about buying recording and playback AD/DAC's from the pro audio industry? ones that operate strictly via USB? Or am I missing something here? Or, can't anyone build a freakin' buffer these days?

I bought this little neat ensoniq, I think (I'll have to check). It is a stereo XLR 24/192 AD/DA recording box, via USB only. $179.00. I was going to mod it out, and do LP-to-24/192khz, in unlocked DVD-A format ($50 for the software), for use on my Denon DVD-A player.

Oh yes. I forgot to mention. I am putting my own clock design in it. It is relatively immune to outside sources of jitter, like RF, for example. Nada. Zip. No effect.

Quote:
They shouldn't say that wired Ethernet SOUNDS better than Wireless. That's completely misleading and it causes Elk to start believing in Santa Claus.

Actually we don't have any indication that Linn takes this position. Erik reported this - certainly in good faith - but I don't think that any of us have seen or heard Linn state this directly.

Again, Alex, all I wrote is "we are all still learning how to transmit data effectively for audio purposes". I did not attack you nor belittle your statements. Why are you deliberately misstating what I wrote and why the vendetta? If this is your attempt at being funny, I doubt many of us share this form of humor. It is long past time for you to stop.

Back to the topic:

I would enjoy knowing what happens - if anything - in a wireless network that can affect the sound (leaving troubleshooting behind as an issue). Unless packets are being dropped or are arriving too late, multiple wireless networks in the area should not make any difference.

Additionally, is there any reason to get excited over USB other than that a USB port comes with one's computer? Soundcards with S/PDIF can be had for very little money and are easy to install and configure. For less than $30.00 and a half hour one can have an S/PDIF digital out.

Quote:

Quote:
They shouldn't say that wired Ethernet SOUNDS better than Wireless. That's completely misleading and it causes Elk to start believing in Santa Claus.

Actually we don't have any indication that Linn takes this position. Erik reported this - certainly in good faith - but I don't think that any of us have seen or heard Linn state this directly.

That's true. We've taken this statement at face value.

Quote:

Again, Alex, all I wrote is "we are all still learning how to transmit data effectively for audio purposes". I did not attack you nor belittle your statements. Why are you deliberately misstating what I wrote and why the vendetta? If this is your attempt at being funny, I doubt many of us share this form of humor. It is long past time for you to stop.

I'm just messing with you. Let's kiss and make up

Quote:

Back to the topic:

I would enjoy knowing what happens - if anything - in a wireless network that can affect the sound (leaving troubleshooting behind as an issue). Unless packets are being dropped or are arriving too late, multiple wireless networks in the area should not make any difference.

I can't think of a good reason why wireless network would be detrimental to the sound unless the audio equipment is not shielded properly and gets affected by the RF interference from the Wireless networks. Medical equipment used to be sensitive to it, but they got that under control now. WiFi is too prevalent and you don't want your heart/lung machine to stop functioning because some bozo wants to get on the Internet.

Quote:

Additionally, is there any reason to get excited over USB other than that a USB port comes with one's computer? Soundcards with S/PDIF can be had for very little money and are easy to install and configure. For less than $30.00 and a half hour one can have an S/PDIF digital out.

When I first heard that DACs were going to incorporate USB for music streaming, I got very excited because I thought that finally, the days of jitter were behind us. Little did I realize that they wouldn't re-engineer the DACs to take advantage of the USB, but rather carry the same, flawed protocol over a different interface.

Furthermore, most laptops don't have a digital out, so a USB or a firewire is the only way to get your laptop to stream audio to an external DAC.

Quote:
...this morning my laptop found 28 wireless networks excluding my own.

Wow!

Quote:
Elk, you are right about soundcards with S/PDIF. However, some people believe that it is hard to get bit perfect output with them and then as AlexO wrote that for laptops there are not many options. Can someone recommend a bit perfect soundcard with S/PDIF for laptops?

Both excellent points.

I know next to nothing about laptop peripherals but I immediately think RME for solid audio interfaces.

Check the computer audio section below for some excellent postings on getting accurate digital outputs from sound cards.

I'm not an Engineer, even though I slept at a Holiday Inn once. Long story on my personal version of a 'good clock'. No hints!

As for buffers with the capacity to buffer well and or correctly, I dunno, in specific terms, much of what I learned at the C, Pascal, Assembler, and etc levels is still true, it is all enabled at the higher levels, for most folks. I'm interested in it at the hardware level and can understand it, for the most part. As for integrating a buffer to the data stream, it can be difficult due to drifting clock speeds, at either end. Basically, tagging the level of fill in the buffer, to high/low levels..and asking for 're-send' and allowing the other end of the buffer to dump out a bit..and then restabilizing and then filling again..etc. The specific mechanics of such with respects to current design are lost to me due to the nature of changes in hardware, over time. But the basics remain. Clocking systems tend to stabilize to one level ..and then the buffer might work correctly. Basically, you need to be in a situation where the 'fill rate' is too high,and then drain off from that. This comes down to purposely speeding up the data rate to the buffer, and then asking for re-sends, as the other end dumps out in sync with the main clock. Keeping the power supply of the buffer separate from the rest of the system might be a good thing too, due to the uneven current loading of the system's PS from the buffer section.

I just went and found the box..it is a "E-MU 0202/USB". I bought it a year ago, for $154.99. Now I'm remembering. It has non-sample converted real sample rates of 44.1/48/88.2/96/176.4/192khz, at up to 24 bit depth...on a USB 2.0 connection. Ultra low jitter clock, of 100ps rms. uh, right. Sounds kinda high to me......I can buy 20ps units for about $.50.

A/D converter is the AK5385, D/A converter is the CS4392.

The unit can be listened to 'real time' with the input signal.

Main chip is the Soundblaster (creative labs) chip.

A rockin' machine for $154.00. The reg price was $249.00...but the Canuckistanbolian Dinero was lower then.

It has a mic input via the left XLR only, for High z, or line level, you go in the neutrik multi-xlr, with a TRS (Tip-Ring-Sleeve) 1/4" jack, into said center of XLR,and then a 1/4" TRS for the right channel, again, at line level, or high-z. So balanced in, high Z line level STEREO, is achieved via 1/4" TRS connectivity. Balanced out, AC coupled, done via TRS 1/4", as well. Easily converted to signle ended, standard output.

I'm looking to mod the entire PS system, the input and output op-amps, the clock,and most specifically the analog and digital PS components of the A/D converter and the D/A as well.

This should make it a stellar little performer. Frightening, if used at a show, off a PC or laptop, to send 192/24 recorded audio to a system. or to, let's say, my highly modded DEQX, at 96/24.

But it's primary purpose is to allow me to transcribe LP's into 'test sources of high repeatability and reliability' audio signals. There is a DVD authoring software package out there, that is $50..and allows one to make DVD-A audio disks..and there IS a 192/24 DVD-A stereo audio format, IIRC, and something like my cheap but highly modded 1390i Denon DVD-A player says..in the manual that it DOES accept that data format.

Everything is there, just gotta break down and find that software on the net, again.... Anyone know what software package I'm alluding to? I found it last year, but I've had difficulty finding it again. The existence of that software is what made me buy the two units, the E-Mu and the Denon.

I'd been waiting for a reasonably priced 192/24 stereo INPUT USB machine to come out at really good price, and the E-Mu sure does the job...at least on paper. We shall see.

The only inexpensive DVD-A authoring software I know of is Minnetonka Audio's DiskWelder Bronze ($100.00 for either Mac or PC).

Minnetonka Audio

I have a copy and it works great.

Quote:
Elk, you are right about soundcards with S/PDIF. However, some people believe that it is hard to get bit perfect output with them and then as AlexO wrote that for laptops there are not many options. Can someone recommend a bit perfect soundcard with S/PDIF for laptops?

If your laptop has a PCMCIA slot, the Audigy 2 ZS has SPDIF (optical) out. While it is a 1/8" jack, a simple adapter can be used to take that to a tos-link connection.

Within the software of the card, there's a tab where you can check for 'allow bit-accurate playback'.

I've used it in my CarPC set-up for bit-perfect playback out to my Alpine 701 processor with great results for a long time.

That would work. However, that does call for a sound card purchase, rather than using a built-in sound card from your laptop.

A while back I attempted to compile some information about bit-perfect. I asked multiple times over on AVS forums for some feedback, but alas, no one seemed interested. This is the information I compiled: http://www.mp3car.com/wiki/index.php/Bit-perfect

I've listed in there some audio codecs that I believed from research were capable of bit-accurate playback. However, that's not been updated in over a year so I'm sure there's more info. I'm not familiar with any onboard sound on a laptop that is capable, but if there is I'd like to know so I can update my wiki.

Quote:
However, that does call for a sound card purchase, rather than using a built-in sound card from your laptop.

True. However, a $30.00 purchase is trivial.

Besides the whole purpose of a hobby is to fritter time and money.

I use the PS Audio DAC with the USB input, with a MacBook and a lossless compression scheme, in my home theatre. I haven't tried it in my dedicated
2 channel system (where the really high end gear resides). It sounds at least as good as the Universal Player playing the same material in the HT system, probably because the DAC is really outstanding. I haven't been able to discern a difference when I try the disc player via the PS Audio.

The PS Audio DAC is a wonderful sounding unit.

Is anyone aware of a review/test examining its jitter specs when supplied by USB?

I've kind of been holding back to comment in this thread until I had a chance to read the Jan issue, but now that I have, I feel that I might have some things to add. Returning to haroon's initial comments, I was actually a little surprised when I read the DAC1 usb review, and a little more surprised that JA commented on follow up posts and not the initial point, i.e. performance of the Benchmark unit. The feeling I got from the review was not really negative at all towards performance, but rather that there was a problem with 16bit playback on two of the three setups (as JA pointed out most likely to do with software/computer setup).

Thus I'm not sure we should yet discount the benchmark (and trash usb audio in general) before JA responds as to whether he and the guys at benchmark were able to troubleshoot the problem.

Secondly, as I've done a bunch of research into usb solutions I figured I'd add some resources both for research and products. The top three places I've found are the benchmark thread at headfi where Elias Gwinn has been kind enough to answer all sorts of questions; Steve Nugent and his forum over at Empirical Audio; and Gordon Rankin over at Wavelegnth Audio.

Both Steve and Gordon have their own usb product lines which, while different are both highly regarded. Another less expensive solution might be the King Rex USB unit which was reviewed over at 6moons.

I'm sure the PS Audio unit sounds great, although from what I've read their implementation of usb is closer to the bel canto dac3, and thus is a little less thorough than the benchmark. This is not to say, however, that it doesn't sound as good or better than the benchmark (a comparison I can't attest to) but just something to keep in mind.

Ok, I'll stop here for now, but hope to hear back from John about troubleshooting the problem with the benchmark.