Of course this wouldn't be a very good explanation of how music compression works without an explanation of bit rates and what they mean. A bit rate at its most basic, describes the amount of storage space per second of audio that a compressed audio file is allowed during the coding process. A minute of audio encoded at precisely 128kbps will therefore always end up occupying 937.5KB.
In more technical terms, what happens in the encoder is that the information that is judged to be useless or inaudible is stripped out to a lesser or greater extent, depending on the specified bit rate. Choose a lower bit rate and the compression algorithm will be applied aggressively, and more musical subtleties will be lost along the way. Specify a higher bit rate and that compression will be less noticeable.
Of course this assumes that we're talking about files that are encoded using a constant bit rate (CBR) here. When you encode a track at a constant bit rate, each and every second of the music is allocated an equal amount of storage space. The trouble with this approach is that music isn't constant in its complexity. A piece of classical music such as Mahler's 5th Symphony, for instance, has many quiet, simple sections but also sections of massive complexity. Variable bit rate (VBR) is designed to take account of this, varying the amount of storage allocated depending on its complexity.
Instead of specifying a bit rate when encoding, with VBR you instead specify an arbitrary quality level, which tells the encoder how aggressively to apply its compression techniques. Thus, during a loud, busy passage of audio, a higher bit rate is used by the encoder, and during simpler sections, a lower bit rate can be employed. Music files encoded with VBR should, therefore, be more space efficient and provide better quality for the space they take up on disk.
Note that many encoders will offer the option of specifying a bit rate and VBR. This is misleading. Think of the bit rate figure as a guide to roughly how large the final audio file will be.
It's also worth bearing in mind that the higher the quality level, the less effect VBR is likely to have. MP3, for example, can't allocate bit rates higher than 320kbps (see next page), and this acts effectively as a cap to MP3 VBR.
How the codecs differ
Although there are many different compression file formats, they all utilise the same basic subset of techniques, as already described. Yet each codec, and its fans, claims some sort of advantage over the other. When Microsoft launched its own compression file type - WMA - it claimed it offered better quality at a given bit rate than MP3. Ogg Vorbis claims superior quality at lower bit rates than other codecs. Fans of Musepack say it's the best bar none at higher bit rates.
It's all a bit irrelevant if your music player doesn't support one format or the other, of course, but increasingly players are straying outside the confines of the more traditional MP3 and WMA limitations to include formats such as Ogg Vorbis and AAC, with mobile phones and smartphones having a slightly different take on format support as well. But if they all utilise similar techniques, what's the difference and, more importantly, how do you choose which is best to encode your music in?
Over the next few pages you'll find a rundown of what makes each of the major codecs tick...