If you're one of those people in search of the holy grail of audio fidelity, there's no doubt that using a PC as a complete front-end solution has probably crossed your mind at one time or another. Saving your entire music library to a hard drive and having all your favorite tracks just a few clicks away is certainly appealing, but what about the sound playback quality? Can it compete with dedicated disc transports costing thousands of dollars?

If you haven't made the move to using a PC as your front-end player, perhaps you've been deterred by the fact that PC's lack the dedicated audio engineering that we find in high-end disc spinners. Or, like me, you brought a cheap CD player and modified it to the nines and are now reluctant to invest your time in starting afresh. Such was my case until a couple of months ago when my aging Pioneer PD-S801 gave up the ghost, leaving me scrambling to find a suitable replacement.

I'd invested so much time into the PDS-801; just about every aspect of the machine had been changed somehow. Modifications to the unit included a directly heated triode output stage, fitting a low jitter master clock, replacing all audio critical electrolytic capacitors with ultra low ESR types, and replacing the stock power circuitry with ultra low noise wide bandwidth voltage regulators. Most of the inspiration for these modifications came from cruising DIY audio forums, where other obsessive-compulsive audio crazed folk like me tend to hang out.

Frequenting such places again in my time of need, I noticed that the buzzword in audiophile circles regarding ultimate digital playback now revolves around using PCs to store and playback music rather than the very best standalone transports that money can buy. It seems the buzz is primarily about three things. The first is the prospect of bit perfect data retrieval when using a suitable lossless format to burn your compact discs to a hard drive. The second is using DRC (digital room correction) to help compensate for listening room resonance and reflections. The third, using software based digital crossovers, thus overcoming passive crossover insertion losses and allowing for a more cohesive integration of drive units in multi-driver speakers.

My previous experiments using a PC with mid-budget consumer grade soundcards fell short of providing the resolution, sound staging, and detail retrieval of the modified Pioneer player. I'd put the differences down to the rampant levels of noise present inside of a PC case. After all, when it comes to soul-stirring audio reproduction, ultra low noise clean DC power is a must, and that's not something that we associate with your typical computer PSU. Computer PSUs are primarily designed to supply huge amounts of current on demand, within a certified noise band of course, but nowhere near the quality we find in a dedicated linear power supply. Hence, serious audio playback requires a soundcard designed to deal with the shortcomings of the PC's internal environment.

This leads us back towards pro audio gear used by recording engineers such as the M-Audio and Lynx range of soundcards. Most of the physical differences between pro audio solutions and your basic consumer oriented product can be put down to better components, trace routing, voltage regulation, and power supply decoupling. In addition, the pro cards feature low latency drivers that bypass Microsoft's K-Mixer and can be used with specialized software allowing all sorts of signal rerouting and manipulation. This adds up to making the pro audio offerings flexible enough for people wanting to engage DRC in a fully customized multichannel setup.

Although user reports on some of the internal pro soundcards are very favorable, my interests are stoked by external affairs. An external box presents far more interesting possibilities and flexibility to me when it comes to power supply and output stage modifications. Both are things that I'm too twitchy to leave alone and unchanged until the unit either dies under the knife or gives me what I want in terms of sonics.

One such solution revolves around using the Texas Instruments 270* range of USB - I2S and S/PDIF converter chips, which are used in several commercial outboard DACs that are rumored to be capable of upstaging even the most expensive standalone players. Better still, a range of attractively priced DIY DAC kits based on the Texas Instruments receiver chips are available that utilize levels of engineering found in commercial products costing much more. The unfortunate upshot with the TI 270* family of converters is that they're designed for two-channel use only. Those demanding external multichannel audio units will have to look towards Pro FireWire audio boxes or standalone units like the Behringer DCX2496, which has more functionality than most of us will ever need. If two-channel playback is sufficient then Logitech's Squeezebox music streamer also deserves a mention. Both the DCX2496 and Squeezebox are products that have been thoroughly adulterated by DIY masterminds and there are plenty of commercial or DIY modification packages available for both units that elevate their performance.

We aim to put some of these products to the test in the coming months while also focusing on commercial loudspeakers, disc players, and amplifiers for a range of budgets from pocket friendly to the spare-no-expense league. Today, we will take a brief look at two DIY DAC kits that we've built up and have been subjectively listening to for the past few weeks. We'll also be looking at PC-based DRC in the form of a software package called Audiolense 3.0 using some open baffle single driver speakers from 3D Sonics. If any of this tomfoolery interests you, read on....

The Test System
Comments Locked


View All Comments

  • ccd - Monday, December 1, 2008 - link

    In an attempt to keep the article from getting too long, the author has skipped for a number of controversies where audiophiles have basically agreed to disagree.

    1) Tube vs Solid State Electronics: Go to a site like audioasylum and you can hear this debate rage ad nauseum. This is a subset of the whole analogue versus digital debate. There are "golden ears" who will swear that only SET provide the best musical reproduction and others who will claim that tubes lend a warmth to some music that is pleasing, but not accurate.

    2) Full Range vs. Speakers with Crossovers: Some argue that crossovers distort the sound and others who say you can't hear it.

    3) Tweaking: There are those of us who think that tweakers just like to tweak and that most of the tweaks really are not audible.

    If you want Exhibit "A" for the solid state position, you need look no further than the Linkwitz Orion speaker which is a 3-way speaker for which some relatively inexpensive SS electronics is recommended. This speaker is hands down the best speaker I have personally ever heard. It also represents a trend in speakers with crossovers toward having dipole mid-ranges.

    A much more appropriate article, IMHO, would have been to discuss the viability of the PC as a single box solution for sound. In the past, there have been a number of obstacles to this. First, the soundcards reputed to have the best sound reproduction (ie, Lynx) are both expensive and not designed for use by audiophiles. For example, volume control with high end sound cards is cumbersome. Second, anything beyond 2-channel sound with a Lynx card becomes extremely expensive. Third, the use of a PC as a single solution is limiting because a PC is not setup to switch between sound sources: you are limited to your CD drive and the music on your hard drive. Fourth, even the best cards lack the ability to handle certain music formats.

    Some of this may be about to change with the introduction of the HDAV soundcard which was mentioned in a previous Anandtech article. An article on that soundcard and how it affects the ability of the PC to act as a soundsource without an external CD player would have been a much more interesting and appropriate article for Anandtech.
  • quanta - Friday, December 5, 2008 - link

    To throw into the tube vs solid state amplifier argument, there seems to be two words that will throw vacuum tube into the... vacuum: Pritchard Amps. It has built quite a reputation on making warm sounds with transistors.

    That aside, didn't AOpen tried building a series of vacuum tube PC motherboards (eg: AK79G, AX4B-533)? Judging on how quickly it went out of production, it seems there aren't enough audiophiles to even care about the differences to sustain the market anymore, especially with the growing generation who has destroyed their eardrums with loud MP3 players...
  • slashbinslashbash - Monday, December 1, 2008 - link

    This has got to the most ridiculous article I've ever seen on AnandTech. Far worse than the digital camera reviews -- at least there's inherent objectivity in the images produced. This whole review is just a bunch of meandering, billowy BS. Has the reviewer conducted ANY double-blind tests of these various audio components? I quote one small part of the article:

    "I find the battery input to provide cleaner, tighter bass notes, more perceived air throughout the mid-band and high frequency range, and better stereo imaging."

    This is either a case of the placebo effect, or the battery causes the components to output slightly higher signal levels than the DC power supply. "Cleaner, tighter bass; more air in the mids and highs; and better stereo imaging" is a classic symptom of "one system is 1 dB or so louder than the other." Slight increases in loudness are not perceived as such, but rather as improvements in clarity and imaging. Pull out your SPL meter and watch the peaks on some program source with the different power sources. This is the first step in setting up a robust listening test. Matching levels is critical.

    A friend and I once held a critical listening comparison of 2 CD's: one the original, one "24-bit remastered". At the first listening test, we were able to tell which disc was playing with surprising reliability. Investigating further, we were surprised to find that the remastered disc had wildly varying SPL levels from the original; it even varied from song to song. Once we matched the levels, we were unable to perceive any difference between the 2 CD's, based on single-blind tests (tester knew which was which, listener did not, tester was behind a partition wall the whole time and was never in view of the listener).
  • phusg - Tuesday, December 2, 2008 - link

    Very good point!
  • mindless1 - Monday, December 1, 2008 - link

    Both the battery pack and wall wart were before a reasonably good (though even with a lowly LM7809 that'd be enough to negate the factor you mentioned) regulation stage that presumably remained a constant, it is not likely the end result would be higher SPL levels with one than the other. You had a significantly different situation when you and your friend started out with two different versions of the same recording.

    However, whether it was a placebo effect or not we can't say since we weren't there to hear it.
  • slashbinslashbash - Monday, December 1, 2008 - link

    First, let's give the author the benefit of the doubt and say that he did indeed hear the differences he described. (Removing the placebo effect explanation, for now.)

    Now, you seem to be positing that the SPL *wasn't* affected by the power supply, because the power supply was sufficiently regulated; but the bass somehow magically became tighter and the mids and highs became more airy.... due to the power supply. You are saying that the power supply still caused a difference, but that the difference was more subtle and mysterious. I posited a simple explanation; yours is more complicated. Of course, either explanation is possible, but we have no way of knowing, due to the complete lack of scientific rigor in the author's discussion.

    I agree, of course, that the situation is completely different from the listening test that my friend and I performed. I used it only as an example, however, of the well known acoustic phenomenon that "subtly louder sounds better" (not to mention an example of how to perform a reasonably robust audio test).

    So, in effect, I was throwing down the gauntlet to the author: perform a listening test with the two different power sources. First, test the SPL output with both setups and make sure that they are identical. If they are, then have a friend switch between one and the other, giving no indicationwhich is currently hooked up. Do this 10 times; if you can identify the correct power source 8 times or more, I will believe that you can hear the difference. Write them down on paper and do not show your answers to your friend until the test is completed, and do not communicate in any way with your friend during the entire test (ideally, have both him and the equipment hidden behind a partition; and the test sequence should be created by a randomizer).
  • RagingDragon - Saturday, December 6, 2008 - link

    The reasoning behind "battery sounds better than wall wart" is that the wall wart is a switching power supply, as such it imparts some degree of noise and ripple in the power; whereas, the battery would provide more uniform power; therefore, the battery should provide clearer more uniform sound. I haven't tested this hypothesis, nor do I care enough to do so, so I'll refrain from commenting on its correctness.
  • Spivonious - Monday, December 1, 2008 - link

    That reminds me of the time I took my original Peace Sells vinyl, original Peace Sells CD, and the 24-bit remastered Peace Sells CD, and played them all on my dad's $10k+ tube system (complete with 24-bit DAC of course).

    Which sounded the best? The vinyl. Isn't it silly how after all of this digital technology that we still can't imitate an analog source?
  • CSMR - Monday, December 1, 2008 - link

    You can imitate an analog source. The people who mastered the CDs didn't want to. If you ran the vinyl into your dad's player and recorded it on PC and made a CD you would get something audibly the same as the vinyl (assuming the sound card and CD player are any good).
    But what about taking a digital recording and making it sound like a record? Anyone know of any good vinylify DSPs?
  • murray13 - Monday, December 1, 2008 - link

    No one can doubt your enthusiasm Rajinder! I enjoyed reading about what you found out.

    Have you tried using the DRC on a 'good' PC audio card? Now that is something I would like to hear.

    As I personally use headphones, using DRC is problematic at best...

    The one thing that has ALWAYS bugged me about Audiophiles in general is that they use recordings that they were not there to hear when recorded. If you don't know what something is supposed to sound like, how in the world can someone say that one thing or another sounds more 'realistic' than something else, when talking about the n'th degree. {rant mode off}

    With all this said, I look forward to reading the next installment.

Log in

Don't have an account? Sign up now