You must have been wearing earplugs for the past few years to not notice that the MP3 phenomenon has emerged as the de facto format for audio delivery over the web. MPEG, or more properly the Motion Picture Experts Group (www.mpeg.org), encompasses a family of standards for encoding audio and video into digitally compressed files, one of which is MP3. Most MP3 software includes the ability to record files from analog or digital sources (such as CDs). And generally, if you leave the default settings alone, you're going to get satisfactory results. But for those tweakin' and tinkerin' gearheads among us (you know who you are), what's exactly going on in there? Let's pop the hood and find out.

The more accurate way to describe the MP3 recording process is to use the term "encoding." The method MPEG chose to use for encoding audio is called perceptual encoding, which analyzes the way the human ear perceives audio. The encoding algorithm tosses away audio data presumed to be redundant or unimportant, based on the perceptual model. MPEG Audio Layer-3 (the long name for MP3) uses this kind of perceptual audio coding algorithm to compress digital audio file sizes without sacrificing sound quality. The magic of MP3 is the fact that digital audio (from a CD for example) can be shrunk by a factor of 12 while maintaining CD sound quality. Or so the theory goes.

Previous to MP3 (and other perceptual encoding methods like Sony's ATRAC compression for minidiscs), the only way to reduce the size of your digital sound file was to reduce its sampling rate and/or the resolution of the samples. But that results in grainy, brittle, unacceptable audio quality you can easily hear in a side-by-side test. MP3 is more clever than that.

Keep in mind that MP3 is NOT a lossless compression scheme. When you encode an MP3 file from a digital audio source, the encoded file actually tosses away data. This is called lossy compression. Purists insist that an encoded MP3 file isn't "really" CD quality, because signal degradation occurs when the audio is encoded. When the mp3 player decodes the audio, the resulting file is not a bit-for-bit copy of the original CD quality sample.

What do we mean by CD-quality? This overused term defines audio sampled in a 16-bit/44.1kHz format, the standard for CD audio. Any time data compression enters the picture, you are actually compromising that pristine gold standard. The audiophiles are right. But for the layman, it's very difficult to tell the difference. And if you'd rather spend some time living your life than waiting for audio to download off the net, the tradeoff is well worth it. Consider this: one minute of 16-bit/44.1 kHz digital audio can weigh-in at 10 megabytes when uncompressed. The same audio when encoded as an MP3 can shrink to around 1 megabyte.

The amount of "headroom" available for sending data over an Internet connection is known as bandwidth. As much as you'd like your Web-based audio to stay in pristine CD quality condition, bandwidth forces your hand when it comes to delivering audio fidelity over the web. For instance, a user connected to the net over a T1 line has more available bandwidth (much more) than a user who dials up to an ISP over a 28.8 K modem. One is a big fat pipe, the other a little pipe.

The larger your bandwidth, the faster you can download files off the net, and the less streaming audio will hiccup and choke on you. You can never have enough bandwidth, especially for multimedia. It determines the available bit rate for streaming or downloading; restricted bandwidth is why compression schemes like MP3 have become so popular in the first place.

Bit Rates
One of the key parameters in your MP3 encoding software you must get to know intimately if you want to flexibly produce audio for the web is known as the "bit rate."

When your MP3 software encodes audio, it uses an algorithm to analyze a soundfile, isolate its distinguishing characteristics, and build a reasonable likeness to the original in the form of a smaller file. Your MP3 player then decodes this file and reproduces the sound of the original.

The bit rate tells you how fast audio data is sent from the storage medium (where your file resides) to the decoder. Another way to put it: bit rate is the average number of bits consumed by one second of audio. Bit rate is expressed in terms of Kbps, or kilobits per second (that's 1000 bits per second). The most common bit rate for MP3 audio is 128kbps, or 128,000 bits of audio data per second. A file encoded at 128 kbps is generally considered CD quality and is safe to use as a benchmark rate.

Logically, the higher the bit rate used when encoding mp3s (that is, the more bits per second) the more audio data per second is coming down the pipe. In digital audio, more means better - which means audio of higher fidelity, closer to the original sound. If your MP3's are going to be streaming over a live connection, keep in mind that higher bit rates require faster connections; otherwise, the sound file is going to "break up" in transit because you don't have enough bandwidth.

Go to Page 2

We invite you to discuss issues or post questions on the Pro Sound World Wide Users Group