Can a PC match the quality of the best CD players?

If you are looking for a fanless solution that is really dedicated, or near dedicated, for audio, look at the Serener ITX PCs. With the NEC spinpoint drive, its dead quiet (that doesn't give you a huge amount of storage, but its networked to a 1TB Buffalo terastation in a remote location. I'm using one and my only complaint is that the blue LED on the front is like a friggin' laser.

http://www.logicsupply.com/product_info.php/cPath/49/products_id/372

Here's my narrow and non-technical view:

Each time a CD/Transport reads data off a silver disk, its doing it in real time. Yes, there is error correction and yes there is error tolerance. But, its subject to the vagaries of reading in real time or near real time.

Computers read blocks of data off a hard drive a lot faster, and those data storage systems are built with much better error detection--yes, disk sectors go bad, but how often do you see that? When you rip responsibly (i.e., EAC), you end up with a copy of the disk that has had each block read over and over and compared to make sure you minimize faulty read data that might occur on a one-time only play.

Onto the computer. There are several things you can do to make computer audio sound better:

- Use an external DAC
- Optimize the digital connection; in some cases this may be a USB path to the DAC itself. In my case, it means a USB audio device (waveterminal U24) to a DAC via coax digital. I tend to dislike sound cards generically; I think a computer is a noisy environment in which to perform that format change.
- Use decent software. Foobar, for example. Make sure you are bypassing the kmixer in windows.

YMMV, but I think a decent computer can sound as good as a $5K transport.

Actually, you don't need a burner to read CDs... ;)

The issue here is that if you proceed like some have recommended, *everything* behind the USB cord is isolated from the audio chain and--unless your computer is so bad it blows the buffering for the serial output or can't keep up with reading--will not affect the audio quality.

And, there are folks--like audioengr--who do audiophile mods for USB devices.

Sorry, a CD burner is the lingo used to describe a drive that "burns" CD blanks--i.e., writes to them. To rip a CD, you need a CD reader.

A CD transport gets to read the data on a CD once and, even if there is some buffering, the bits read off the CD are basically what goes to the DAC. So, a transport is subject to the vagaries of power supplies, glitches, whatever. The disk has to be spun precisely, because it just gets one chance.

On the other hand, the software that controls a CD reader on a computer can tell the reader to read the same block of data over and over again. A good ripper--like EAC--does that and compares the data it gets, over and over again, until it is statistically satisfied that the copy created is a complete and accurate duplicate of what is on the CD. Because a computer can read the data created off a hard driver--where much more sophisticated error correction can be employed--timing sort of ceases to matter. The data is spit out asynchronously over USB--no timing information.

That is why I say everything behind the USB cable is irrelevant. The timing is supplied by the USB audio device, which buffers the data and outputs it based on its own clock.

To give you a very concrete example, I've got several "unplayable" CDs--stuff that won't read in my DV50S, my Theta David, my Sony SACD player... But, I can rip those with EAC and get a perfect set of .wav audio files off of it. Takes forever, and I wonder if replicating a bad CD is worth the wear and tear on a CD ROM that is basically running for 24 hours straight, but it works...

Hmm... I think I've made some blanket statements in the past that USB was effectively jitter-free; that was based on my understanding that USB was an async protocol that had to be buffered and reclocked. I conceded, however, that jitter could be introduced in reclocking the stream.

Looks like the serious tech-heads at the Asylum have looked at some measurements and are concluding that USB may not be the panacea... Check out this thread:

http://www.audioasylum.com/forums/pcaudio/messages/9640.html

I am not sure I quite understand this, and I'll still stand by my statement that my PC-based rig sounds as good as several different transports I've used. Just thought I'd make sure the record was a correct as possible...

For those looking for cheap network storage, this looks interesting:

http://www.trustedreviews.com/article.aspx?art=2484