Home Entertainment

 

HiFi Digital Audio

June 25, 2010 By Geoffrey Morrison



Digital Audio
Analog Sound

Everyone has music on their computer.

The question though, is what’s the best way to get it from there to your ears?

We take a look at how it all works and poke at three DACs vying for your attention while we're at it.

Let's start at the beginning: If a tree falls in the woods and no one's around, does it make a sound?

Technically, no. It will cause changes in pressure in the air, but without something to hear it, it's not sound. It's our ears that register and our brain that recognizes these changes in pressure as sound. It's from this core simplicity that everything gets annoyingly more complex.
    
In the early days of audio recording, the changes in pressure were recorded verbatim onto a medium. Light bulb lover and DC power enthusiast Thomas Edison is credited for coming up with the first practical audio recording device, the phonograph cylinder. The compressions and rarefactions in the air could now be stored and played back with reasonable fidelity. Hi-Fi was born. OK, Hi-ish-Fi.

Fast-Forward

Audio quality got better and better, even surviving 8-track and cassette. Then the folks at Philips and Sony got the bright idea to ruin it all. Why have direct representations of the sound waves, when we can make them digital? No one bothered to ask "why," and digital audio was born.
    
At CD quality, every 44,100 times a second, the sound wave (blue line) is sampled (red dots). Each sample is given a numeric value. During playback, the wave is reconstructed based on these points.CDs are 16-bit, 44.1kHz. This means that 44,100 times each second the sound wave being recorded is "sampled." The analog-to-digital converter (ADC or A/D) takes a snapshot of the sound wave and assigns it a number. That number has a resolution of 16-bits, or 65,536 (see Figure to the right).

So the sampled point on the wave can have a value anywhere between -32,768 to +32,768. That's pretty much it. So 44,100 times each second the ADC sees what the sound wave is doing, and assigns it a number. On playback, a digital-to-analog converter (DAC) does the opposite. It creates an analog waveform based on the number it reads. But we're getting ahead of ourselves.
    
Why 16/44? Most people are taught that the human ear can hear from 20 Hz to 20,000 Hz. This is pretty much crap. Maybe babies can hear that, but everyone starts losing their hearing from that point on. But that's not to say we can just have 20 Hz – 12 kHz or something and call it a day. There is a pretty widely held audiophile mantra that the frequencies above the hearing spectrum affect the frequencies we do hear. This could all be why no matter how good the CD is, it still never sounds "real," like you're there, despite being "perfect sound, forever" and supplying all the frequencies we can technically hear. Early experiments with "brick wall" filters at 20,000 Hz also resulted in odd audio artifacts.
    
A Swede named Nyquist figured out that you need to sample at twice the frequency that you want to reproduce. So the maximum frequency possible recorded at 44,100 is 22,050 Hz, enough above 20,000 to allow for a bit of a roll off instead of a digital wall. It's also, conveniently, what the equipment available at the time could do. So there's that. That's also pretty much why we have 16-bit and not 32 or more. Remember, getting a computer to do anything 44,100 times a second was serious business in the late ’70s. So that's what we're stuck with. At least as far as CDs go.
    
FFx2

Micromega AirstreamAs with anything, though, technology moves quickly. Computers are infinitely faster, storage space infinitely larger, and 30 years of development mean that we don't have to be stuck with 16/44. You'll often see 48 Hz as a frequency in movie soundtracks. DVD-Audio claimed sample rates of 96 and 192 Hz. Even better, 24- and 32-bit audio is widely available on the pro-audio side.

With the demise of DVD-A, this is harder to find on the consumer side, but it's out there (check out Dennis's article Downloads on the Up and Up). It's often the bit depth (more available numbers to assign) that most audiophiles will say is the important part. It makes the audio sound smoother, less "digital."

> (Play)

All of this just addresses the recording and storage of the audio. The playback is just as important, being an equal part of the chain. There is a school of thought that all CD players sound the same and that because it's "digital," it's always perfect and that's that. Well... sort of.
    
It is true that if you get the bits from the CD perfectly to the DAC and that perfectly converts the bits back into a wave, then you'll get exactly what was on the CD. But that's a lot of “ifs.”
    
Simaudio 300DThe first "if" assumes that all CD players will read every bit on a CD. True, they should, but even though there is extensive and exceedingly clever error correction built into the CD data, it's not perfect. So ideally you need a player that does its best to get every bit off the CD as accurately as possible.

Does this mean you have to spend $10,000 on some fancy transport? No, but it does mean that your $100 DVD player probably isn't up to snuff. Remember, this is not vinyl. This is not just a plastic "record" with a linear stream of 1s and 0s. There is a LOT more going on. If the tolerances are off on any number of parts (the drive motor, for one), little variances will add up to errors in the sound.

Worst case: dropouts like we've all heard on a really scratched CD. In a less-severe case, you're just missing data, which reduces fidelity.
    
So even if we assume the data coming off the disc is perfect, it's the next step that gets really messy. The stream of 1s and 0s is being thrown at the DAC at a pretty high rate of speed. The DAC needs to piece these 16-bit chunks back together to create the waveform originally recorded.
    
Here's an analogy: Say you're playing mini-golf. You know the hole with that damn windmill? I hate that one. You need to time your putt in such a way that by the time the ball reaches the windmill, the blade is not in front of the hole. Time it right, ball goes in. Time it wrong, your kids mock you. As I said, I hate those holes. Digital audio is the same, except that the blade is going past the hole 44,100 times a second, and the hole is only the size of the ball. Each ball has to get through at exactly the right time or the re-created waveform will be different than the original. In other words, packets of data are being thrown at the DAC. If the timing is a little bit off, you'll get errors and errors mean a change in the sound. This is known as jitter. Minimizing jitter is the key to quality digital audio.

Enter the Computer

These days, the vast majority of people's audio library is virtual and digital. Storing audio on the computer isn't new, but being able to do it without it sounding like crap is. Faster processors and greater storage mean you can have uncompressed or losslessly compressed files of your entire library. The trick is getting this audio to your ears. Computer speakers have come a long way, but they are a poor substitute for a great set of speakers and a kickass amp.
    
Wadia 150The first benefit of importing your music is the ability to minimize read errors. If you use iTunes, make sure you enable "Use error correction when reading Audio CDs" in the Import Settings menu. This will ensure that you get as accurate a copy as possible. It might take a few moments longer to encode a CD, but it's worth it. If you don't use iTunes, this type of option is going to be available in any decent encoder.
    
Getting the music to your audio system, though, presents more of a challenge. Sure, many sound cards have digital outputs (coax, for example), but personally I wouldn't trust the sound cards in most computers. Computers are fantastic at a lot of things, but a high-end audio device, they are not. There is way too much noise (of the electrical variety), interference and potential timing errors possible inside a computer. Best to get the stored files out and into something audiophile.
    
A better option can be USB. The trick is, as with all digital audio, minimizing jitter. Each of the three products you see reviewed in this article has different ways to minimize this jitter, as you'll see when you read those sidebars.
    
Another method, used by Wavelength Audio and Ayre Acoustics' USB DACs, is called Asynchronous USB. This is where the DAC itself controls the flow of data. This intrinsically lowers jitter by allowing the DAC to keep time with itself instead of relying on the computer for timing or rebuilding the timing internally.
    
Of course each company says their method is amazing, and most have compelling data to prove it. As far as I'm concerned, though, minimizing jitter by some means is the biggest step. From there it's up to your ears as to what you think sounds best.

Backup!!!

I'll end on this: If all your music is on one hard drive, you are asking for hurt. To quote an old adage relayed to me by Micromega's John Bevier: "There are two kinds of hard drives; those that have failed, and those that are going to fail." External USB backup drives are insanely cheap (1 terabyte drives were under $100 as of this writing). A NAS, or Network-Attached Storage drives are under $200. These let your whole family back up data, and most allow you to access or store data on them from anywhere in the world.
    
Buy at least one and backup all your music, movies and especially pictures. If you don't, you stand to lose all of it, and how much of that stuff can you easily get back? How much is irreplaceable?

 

Check out these USB DAC Reviews:

Micromega AirStream Review

Simaudio MOON 300D Review

Wadia 151PowerDAC mini Review

Comments

Post new comment

  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
  • Images can be added to this post.
  • Glossary terms will be automatically marked with links to their descriptions. If there are certain phrases or sections of text that should be excluded from glossary marking and linking, use the special markup, [no-glossary] ... [/no-glossary]. Additionally, these HTML elements will not be scanned: a, abbr, acronym, code, pre.

More information about formatting options

Local Guides

 All Guides
   Alabama
   Alaska
   Arizona
   Arkansas
   California
   Colorado
   Connecticut
   DC
   Delaware
   Florida
   Georgia
   Hawaii
   Idaho
   Illinois
   Indiana
   Iowa
   Kansas
   Kentucky
   Louisiana
   Maine
   Maryland
   Massachusetts
   Michigan
   Minnesota
   Mississippi
   Missouri
   Montana
   Nebraska
   Nevada
   New Hampshire
   New Jersey
   New Mexico
   New York
   North Carolina
   North Dakota
   Ohio
   Oklahoma
   Oregon
   Pennsylvania
   Rhode Island
   South Carolina
   South Dakota
   Tennessee
   Texas
   Utah
   Vermont
   Virginia
   Washington
   West Virginia
   Wisconsin
   Wyoming