This site seems to be slowly but surely growing into a nexus where regular audiophiles access and also share information. Occasionally, I receive unsolicited contributions like this one. While publishing them doesn't automatically mean I agree with everything stated (or had the time to personally corroborate what is presented as facts), I'm still very much in favor of adding them to our formal content. If you wish to correspond with today's author, his signature at the end is linked to his e-mail as is our custom for all our regular contributors - Ed.
One thing that some audiophiles like to do is learn about the process that makes music playback in our homes possible. My interests in particular were to learn how a digital system (really) works, then go on to find the best ways to play back digital music.


While learning about the digital process, I learned two things: all of the previously employed concepts were flawed; and these flaws were great in number. It wasn't just the playback gear either, it was the entire recording and playback process. I didn't realize how deep the problem was. A series of recent, very positive developments concerning digital playback made my journey all the more exciting. And since no one has ever written an article explaining all the limiting factors in the digital process, I felt that I should step up to the plate - the "timing" was perfect. Understand that I am not a designer of audio equipment, I am simply a music-loving audiophile who got a little curious.


I have isolated eight areas of concern. Each paragraph begins with a summary of recent developments, then cites a few companies working in these areas. Note that this list is concerned over why digital struggled, not why home audio struggles. All systems -- digital or analog -- benefit from better isolation, signal wire and power amps. These categories therefore have been omitted. So here are the key areas that affect digital audio's performance as far as I have been able to decipher:


1: Recording studio power
This was a surprise. Up until the mid 1990s, most recording studios were using unbalanced/noisy power - they had no choice and probably weren't concerned anyway. Then along came Martin Glasband (now of Equi-Tech) who along with his associates made formal proposals to the U.S. government concerning an upgrade of our national electrical code. The government acceded by responding to a real need of the high tech industry seeking cleaner, less noisy power. Glasband's group proposed balanced power and the government followed up. This not only helped hi-tech but high-end audio as well. Noise that had remained inaudible to analog recording systems became obvious to high-resolution digital recording systems. By going balanced, these systems gained an increase in dynamic range of 16dB or more. This is one reason why today's digital recordings sound better.

2: Recording techniques (new recordings) / remasters (older recordings)
A biggie. Because digital sound is so clear and revealing, it cruelly shows up bad judgment in regard to microphone placement and tonal balance. As a result, (serious) recording engineers had to become musically literate and the old analog recording techniques had to be ditched. The entire recording process had to be revamped. Multi-point miking was replaced by one-point purist (or at least fewer) mikes. Recording levels had to be completely readjusted. At the console, higher resolution gear and higher sampling rates -- to at least twice the Nyquist requirement -- were being used, a great development. The sampling methods also improved - the latest includes "Faulkner Downsampling" as used by Hyperion and others. A final development was the use of quieter, less resonant microphones. All of these efforts have paid off wonderfully as you can hear on any recent recording by Harmonia Mundi, ECM, Reference, Water Lily, EMI & Chesky among others [WaterLily records purely in the analog domain - Ed.].


Recent digital works are among the greatest recordings ever made - maybe the greatest. The remastering process has greatly improved too, due to lower jitter interfaces, better storage media and more accurate analog-to-digital converters. Brand new remasters sound noticeably better than their remastered siblings from just a few years ago. But this process is not over. Future remasters will sound better still.


3: Data placement method (on disc)
I never would have thought of this but as I said, the problems were manyfold. We now realize that a freshly burned CDR (from a hard drive) can sound noticeably better than a mass-produced CD. You see, the pits on a CD are stamped on the disc, then covered by two layers of material. The pits are in effect "buried" beneath the surface, making it difficult for even the best lasers to pick them up. To make matters worse, the pit edges produced by this process are not as sharp as they should be. Lasers produce sharper pit edges than those produced on a stamp/mold machine. An added benefit of burning is that there are no layers covering the bits, they're right on the surface. All of this allows for a more accurate pickup process. SACD and DVD-A have an advantage in that their pits are placed closer together than Redbook but when RB is burned to a CDR, this narrows the gap, pun intended. Software programs to use here include ExactCopy and Nero. Make sure that the "ripped" data is stored losslessly, then "burned" at 1x speed. Finally, use any gold or black CDRs - Mobile Fidelity now makes archival-quality ones.


4: Playback media
This area concerns the upkeep of the discs themselves. They get dirty, dusty, oily from our skin and even magnetized just being near electronic circuitry. So we need to clean and "demagnetize" our CDRs - Bedini, Walker and Furutech have products to help in this category. They make noticeable improvements. A wave of the future will be to drive digital signals direct from a hard drive. All of the issues in #3 and #4 simply go away when replacing a manually loaded, optical plastic disc with a permanently fixed magnetic hard-disk. Besides, there are other reasons to do this (see #6). VRS and Linn are among the few companies that make "high-end" hard drives.


5: Internal & external noise contaminants
A very serious problem. Remember from #1, digital is very sensitive to noise and the sources here were many. It's a fact that any noise introduced to an analog system winds up as background noise. Not so with digital. Noise in a digital system is far more damaging as it commits a far more serious crime - it distorts the waveform. It should be readily apparent that distortion is much worse than background noise. Areas of concern:
  • Incoming AC stock power cords: We knew these were potential problems but we had no effective way to deal with them. We do now with the latest wave of power conditioners and AC cords. It wasn't too long ago that we still weren't sure if these products were helping or hurting our systems. But recent efforts by Nordost, Equi-Tech, Richard Gray and Shunyata among others have changed that as major progress has been made in cleaning up AC pollution. These products will help an analog front-end and even preamps and power amps but are much more effective on digital components that are more sensitive to noise.
  • Radio frequency emissions: Besides outside noise, there's also inside noise in the form of radio frequency emissions - RF. These are being generated by the switching circuits of the digital gear itself. Stillpoints makes high-tech sheets (ERS cloth) that clear up the RF fields - they work.
  • The power supply: No surprise here - a major source of noise. A recent development to deal with this issue is to separate the power supply from the digital processor. For years it was thought that to get the best out of digital, you had to separate the CD transport from the digital processor. The idea of separation was a good one, it's just that we were separating the wrong things. Outboard power supplies were found in high-end equipment but only with preamps & power amps. We now realize that this approach is even more effective with digital gear.
  • Component parts: Somewhat of a surprise. Component parts and construction methods were either noisy or coloration-inducing on their own. New measures to combat this problem are multi-layered PC boards, cryogenically treated circuits and MusiCoat, a micro-film coating that reduces the colorations caused by the hard plastic casings of ICs.

All of these measures help digital reveal its potential which includes an ultra-quiet background, lots of inner detail and expanded dynamic range.


6: Digital timing errors
Jitter. The sources here were many - warped discs, the eye pattern, clunky transport mechanisms, output filters, improperly pressed software. But the primary reason? Tying the system master clock to the transport. This flawed idea gave these other areas opportunities to corrupt the (critical) time dimension of the music waveform. The read-in process alone was causing problems even on great players. Only a few picoseconds worth of variation causes distortion. Many companies tried to help matters via time-base corrections. That wasn't enough because the timing clock before and after this circuit were tied together. So why not go all the way - isolate the transport by placing the master clock in the digital processor - right next to the DAC?


By doing this, the master clock becomes immune to any external influence. The transport still needing a clock to guide it could have its own "local clock". This separating process has already begun. Note that this also includes single-box players, not just separate DACs. While single-box players do not have interface jitter, they still suffered timing issues - their master clock was still tied to the transport. The latest feature in separate DACs is "real-time re-clocking" which obliterates the transport-is-the-problem issue. Even if you're not in the market for brand new digital gear, you can still improve matters by buying an outboard "re-clocking" unit (for separates). These new units work far better than the "jitter reduction" devices of yesteryear. Apogee's Big Ben is one example. Note that hard drives don't and never did have timing issues. Any timing irregularities die at the input buffer. How nice. Out-of-sync clocking at home and in the recording studio, along with insufficient recording sampling rates, were the two biggest obstacles facing digital audio and the primary reasons why it wasn't living up to its potential.


7: Digital filtering
This is the only point on this list hotly debated because the opposing ideas all seem to have merit. Let's not forget that the job of the digital filter in your DAC/CD player is to create all the music you hear above 2kHz. It's not just to filter out ultrasonic noise. What is surprising is how radically different the methods are to carry it out. Note that the region right above 2kHz is where the ear-brain is the most sensitive. There are two methods:

  • Oversampling/Upsampling. Standard oversampling used for many years was a waste as it turns out. In a 4x oversampling scheme, there was only one real waveform sample followed by three dummy samples. We were wasting 3/4 of the sample. Couldn't we put these to better use? We did, finally - enter socalled upsampling. This was only half an idea, however. It needed help in the form of high-power averaging. This technique, borrowed from the low-bit DSD camp, allowed us to mathematically fill in those empty slots with real values along with increasing the output word length from 16 to 20 bits. Upsampling was a only a preliminary step toward achieving this goal, with the averaging calculations taking over and producing the higher levels of resolution. dCS was the first of eventually many to upsample Redbook CDs.
  • No oversampling. Within a few months of dCS going up back in the late 90s, Audio Note went down. They were the first to embark on this less-is-more school of oversampling (none at all). Others have followed, most notably Zanden. Proponents of this theory claim that it better preserves the integrity and timing of the waveform.

'I'm still confused over which way (up or down) is better - but I do know they both made significant improvements over prior techniques.

8: Linestage preamplifiers
A shocker. It was always known that digital processors -- separated or inside a CD player -- produced enough output to drive power amps directly. But some form of attenuation was still required and with limited choices, most of us were using active linestages - tube or transistor. Good as they sounded, they were unnecessarily boosting the current. Along came transformer-based line controllers (TLCs). These all-passive units might have been new to high-end audio but not to the pro recording industry - they've been using them for years. TLCs don't amplify. They are different too from passive resistor-based line controllers (RLCs). Where RLCs burn off volume as a means of attenuation, TLCs transform voltage to current - a much better way. They're also free from the impedance mismatch issues that RLCs suffered. They are the most transparent link yet for digital sources. Companies that make them include Sonic Euphoria and AVTAC.