Pacific Valve's take on USB audio. What do you think?

[funkle]

Vic Trola (is that his real name?) has posted this video:

http://www.youtube.com/watch?v=iAYohdX7BhU

There are some interesting points here and on his site, some of which seem to go counter to the general census here. For instance he says a computer as a source will never sound as good as a CD player or proprietary media server here

http://www.pacificvalve.us/USBHELP.html

His explanation does seem reasonable. It was my understanding that a CD ripped to a hard drive was better, because of the error correction that could be applied during the rip, but he's saying it all falls apart in the streaming. His solution has you going USB (using their PC only software) > SPDIF > DAC. Seems like a lot of conversion. I'm interested to see what your thoughts are.

 

[The Computer Audiophile]

No comment :~)

 

[Wavelength]

Funkle,

 

The problem today is that USB does not come completed for you. You have to either develop it or buy it.

 

CD's really cannot sound as good as computers for the very reason you suggest. You can get close if you bypass the entire drive output and run it like a computer in ATAPI mode. But then heck why not just use a computer and still you run into the problem with error recovery.

 

Look a lot of people are just going to tell you what you need to know to buy their products. It's up to you to go through the jungle of crap to get what you are looking for.

 

Thanks

Gordon

 

[ar-t]

"Digital Audio Level 1 is using a USB DAC connected to your computer. It is easy to install and does not matter what version of Windows or MAC you are using. Its great for background listening and entertaining. It should not be used for serious listening."

 

Maybe someone should point out there is life after a PCM2702.

 

But he only goes to 5................why no 11?

 

"But when you are already at ten, and you want to go higher, where can you go? You can't. So, that is why ours goes to eleven."

 

[audioengr]

Outdated and uninformed IMO.

 

I do feel that locating the conversion to I2S or S/PDIF outside of the computer in an external device (which may be the DAC) is a huge benefit. This allows the clocking and associated logic to be powered from a low-noise, fast-reacting power source. This can only improve jitter.

 

Steve N.

Empirical Audio

 

[idiot_savant]

I love bucket analogies.

 

This one, however is terrible.

 

The whole "different size lego blocks" problem is completely and utterly irrelevant, because the DAC has... a buffer.

 

Which ( to stick to the analogies ) is another bucket in-between the source and sink buckets, which drains at a constant rate, no matter whether the source bucket slows down and speeds up or not...

 

your friendly neighbourhood idiot

 

[firedog]

IMO the whole approach seems to be deceptive - "PC's can't reproduce music well b/c they are computing devices". It may be good salesmanship, but I don't think it is some kind of factual analysis of digital audio reproduction.

 

 

First,according to my definition, if something runs an OS, has memory, and a drive, it's a computer. So his "Cyber Server" is essentially a simple computer. It has an OS, CPU, memory, drive etc.

 

Again, he also recommends using the Netgear Ready NAS to add HD space for his "Cyber Server" if you need it. What is that, other than another "computer" put into the mix? It also has it's own OS, memory, and HD.

 

His definitions are also very self serving - "jitter is what happens when music moves" ; basically saying that "all methods other than the one I'm suggesting are high in jitter"(my paraphrasing in quotes). But no, moving music off a HD doesn't cause jitter. If it did, even his method wouldn't work.

 

His "digital audio levels" are also amusing. Note that he puts quality CD players above ANY usb playback. I'm not sure then that he has even listened to quality USB playback, as all of us with experience know that high quality USB playback tends to be as good or better than even high end CD decks.

 

Note what he also brushes over in his salesmanship - the simplistic, probably not very convenient UI that his method forces you to use. Not something I'd want to browse and listen to my music collection with.

 

There are lots of ways to get good sounding digital audio. USB can be one, and his "Cyber Server" may also be one. Certainly on paper a simple computer/streamer using an SSD and ethernet seems to be a good way to get a quality signal to DAC. His device may sound good. The problem is that he is trying to sell it by saying that every other method is, by definition, inferior.

 

[audioengr]

"the DAC has... a buffer.

 

Which ( to stick to the analogies ) is another bucket in-between the source and sink buckets, which drains at a constant rate, no matter whether the source bucket slows down and speeds up or not..."

 

99% of DACs have no buffer that has the functionality you are talking about. There is no buffer if you are driving I2S directly and little jitter reduction. This all happens before I2S. Even if you are using S/PDIF, the buffer is very shallow and specifically there to work with the PLL in the CODEC chip. It is the PLL that speeds up and slows down to track the incoming stream.

 

A buffer that accepts a high input rate and outputs a lower rate is a buffer that is broken and will eventually overrun.

 

It is possible however to have a type of DLL that is crystal-controlled that tracks the incoming rate and makes adjustments over a long period of time, say seconds. My Pace-Car reclocker in mode 3 works this way. It is the only one of its kind I believe.

 

DACs are asynchronous using the S/PDIF input and synchronous using the I2S interface, usually internal, but sometimes external. There may be some reduction in jitter due to the PLL, but not much. Even the best chips are so-so, and most of the incoming jitter gets through. Even cascaded PLLs dont make that much difference IME.

 

There are a lot of misconceptions about how DACs work and how digital data is transmitted.

 

Steve N.

Empirical Audio