You are reading the older HTML site

Positive Feedback ISSUE 64
november/december 2012

Back to the Computer Audio Drawing Board... Again!
by Tom Gibbs

I would probably never have gotten into this whole computer audio thing without my know-it-all brother's constant insistence that I was "missing it," and especially without his offer to me of a computer for use as a media server. Of course, that relatively cost-free and painless humble beginning has in less than two years of constant upgrades morphed into my third media PC, which realistically is only serving as my music server since I've yet to take the plunge into visual media via PC on the big screen. The most current PC is probably the best yet in certain regards, but still has unresolved issues (we'll get to that later)—I'm pretty confident we'll be building Mark IV from the ground up in no short order to address some of the shortcomings of the third iteration.

I've really tried to embrace the whole "less is more" philosophy in my approach to audio reproduction, but this whole computer audio thing has had me rethinking that approach, and to the total detriment of my already-thin wallet. My most recent revelation came after ripping another one-buck acquisition, Natalie Merchant's Ophelia CD. I'd seen this disc for as little as $2.50 on several occasions, but bristled at paying much more than a buck. What unbelievable stupidity on my part—this record is so superbly crafted, with such excellent songwriting, musicianship and Natalie Merchant's incredibly liquid vocals—it's easily worth every penny of the original suggested list price! The revelation came after having ripped the disc to level 8 FLAC with my usual setup using Exact Audio Copy; after a couple of days, when I actually got around to listening to the ripped files, I was dismayed that I could hear this constant high-frequency noise on the opening title track. At reference level playback it was particularly nasty—I checked the CD again to confirm that the surface was clean, and it was indeed pristine! A second rip with EAC yielded identical results; at that point, I played the CD directly and was equally shocked that no noise was present at all—what a conundrum!

Problem No. 1: Accurate Ripping

Now, I've already gone and done the whole FLAC-to-CD-to-WAV comparison thing after the TAS flap last year, and couldn't really find any significant or apparent differences between the half-dozen discs that I chose specifically for that purpose—they all sounded uniformly excellent. All over the blogosphere I regularly see the "EAC is crap!" rants that are invariably followed by the "dBpoweramp is God!" pronouncements—up to this point, I frankly bristled at the thought of shelling out any additional cash for a ripping program—EAC is free, and had never given me any cause for concern with ripping accuracy. This recent instance was, however, particularly irksome, and with some trepidation, I downloaded the trial copy of dBpoweramp, which is fully functional for 21 days and about $50 should I decide to keep it. You can get a single-machine copy for about $30, but if you need the convenience of having it on multiple machines (really, really convenient to have on a laptop, for example) it'll cost you the full fifty.

After ripping Ophelia again, I was stunned at the difference in overall clarity and especially delighted that there was no noise. Of course, this got the wheels turning, and I know this sounds borderline OCD, but I've already been spending lots of late nights re-ripping many of the files I've already ripped at least once before but now with dBpoweramp (about 70 discs in less than a week!). And I've been taking the time to take a listen, and so far the results have been not much less than thrilling! Hey, and with dBpoweramp, you can download a plug-in that will allow you to rip all your HDCDs at full 24 bit resolution—and they sound pretty darn spectacular—I think this puppy is a keeper, and is definitely due for a full-blown review. Which, by the way, turns out Ophelia is also an HDCD, and within a few minutes we had the music playing in 24-bit glory—brilliant!

I know I'm really trying to do the whole great-sound-on-the-cheap thing, but as I found with JRiver Media Center, sometimes you have to give a little to get great performance, and I totally have no qualms with spending a little extra for really accurate rips. I look back at all the countless tweaks and gadgets and gizmos I bought to improve analog playback over the last 30 years, so it's only logical that I (we) should expect to pay a little for top-notch playback and performance.

Problem No. 2: Perception

Let's face it—one of the biggest reasons most audiophiles would balk at having a computer in the listening room is one of perception—computers don't look or behave like the average high-end source component. Now if we really think about it, some of those source components already in our systems utilize a pretty high degree of computerization—I remember reading years ago, when the first Sony CD player came out, that in 1960's technology, the processing power of the microchips inside would fill a three story building three city blocks long. But computers just don't look as though they belong in high-end audio—if I'm contemplating a component purchase and there's a fan in the case, by nature I look elsewhere. Now, my current PC is a step in the right direction—it's in a horizontal case that's standard rack width with a thick aluminum faceplate—it almost looks like an audiophile source component. Unfortunately, that's where most of the comparison ends; while it most assuredly is feeding my DACS with magnificent sounding music and powering the software that drives the storage and organizational capacities of my system, all one has to do is hit the power switch – when the whirring of all those fans kick in, you immediately know why it's been relegated to the room next door!

The Real Problem: Noise!

The first and foremost reason most audiophiles would bristle at the thought of having a computer in the listening room is—plain and simple—Noise! Whether computer-generated rf or simply just the operating noise of the mechanism itself, the noise generated by the average PC or Mac can be almost overwhelming. I'm not too familiar with tablets yet—the concept seems very promising, but they're pretty limited in terms of storage; I think they hold much more promise in terms of controlling your computer and/or music server remotely. And most laptops are generally pretty noisy, as well.

I currently have three compact DACs in use in my system, the HRT Music Streamer II+, Centrance's DACport LX and the Audioquest Dragonfly (review coming soon!), and of the three, only the Music Streamer is dead silent in my system. I currently have my computer and external drives in an adjacent room to eliminate the drive and fan noise from the big boxes, but any cabling scheme I've come up with to get the analog output from the USB DACs to my preamp's inputs has been met with absolute failure—with the exception of the HRT Music Streamer. It's a real champ, and behaves the most like a true high-end component in my system—in terms of sound quality alone. If I could only get it to work with WASAPI Event Style, all would be well. The other two devices, which work perfectly with WASAPI, each suffer from a degree of rf hum induced from some source I've yet to determine, and I've been down the checklist countless times and in seemingly countless cable configurations trying to eliminate the problem with no luck thus far. Maybe at some point in the near future some elegant kind of wireless solution will become available to eliminate the long cable runs (whether interconnect or USB or coax digital). And I'm absolutely certain that there are no inherent problems with any of the DACs I'm currently using—they all perform quite admirably, although I can't help but wonder if both the Centrance and Audioquest units were in slightly more robust cases, then perhaps they'd offer better noise rejection.

Throughout my experimentation with cabling and computer placement, I've discovered that when the computer is physically present in the listening room, the noise problem goes away for all devices. I've really been working at trying to isolate whatever is causing the noise, but the relatively small amount of DAC-generated hum is much less intrusive than the unbearable fan noise from the computer case. Anyone who'd like to chime in on the situation, please feel free to send me an email!

Moving Towards a Solution

I can't help but feel that throughout this process, I've flipped and flopped, zigged and zagged and changed my position on what works and what doesn't so many times that it probably strains whatever "street cred" I may or may not have at one time had. I'm sorry, people, but there's a pretty steep learning curve involved here! The bottom line is that slowly—and by repeated experimentation—I've gently tweaked my digital playback to the point that it really rivals any level of satisfaction I still get from analog sources. The real issue here for me is pretty simple: if the computer is now considered a true audiophile source component, then we need to take whatever steps necessary to bring it up to speed and make it more at home in the listening room environment.

I'm already contemplating my "Mark IV" rebuild, and I've targeted some key areas for improvement: 1) decrease processor related noise, perhaps with a much quieter cooler (no fans!); 2) the graphics card must be of the "0 dB" variety; 3) solid state hard drives (no spinning mechanisms!); and, 4) all case fans must be exceptionally quiet! As I mentioned earlier, the computer case of my current media server looks a lot like an audiophile component, so I'm perfectly OK with it, and if the rebuild goes as planned, hopefully it will perform quietly and up to my expectations. I'm ecstatic with the music playback, especially after adding dBpoweramp into the mix—now if we can only get that damn computer to behave more like it belongs there!