Basically it's all about compressing the files, dropping some of the inaudible data, and the encoding of duplicate data. A good example is comparing a one minute Wave file recording in high quality might take up to 10 Megs of space. The same file when converted to an MP3 only takes about 1 Meg. That's a huge improvement over the size of the files but there will be some loss in data.
There are two basic ways to compress any files. Lossless and Lossable. Lossless is like a zip file that will compress the original and then uncompress an exact match. Lossable compression (used by the MP3 file format) means that it's fairly close to the original but not a perfect match.
MP3 files can be encoded at different bitrates which is measured in Kbps (kilobits per second). You can record MP3 files at either 128 kbps (which is OK quality), or 160 kbps for good quality playback, and 193 kbps for very good quality on most systems. Using the best possible original source can make a big difference in the end result too.
For those with very high end stereo playback systems might want to up the bitrate to 193 Kbps. This increases the amount of data and therefore can increase the amount of detailed sounds in the playback. Of course many people with less than perfect hearing will never notice the difference. And the method you use to playback the recordings will also impact the quality of reproduction of the original.
If you rip (record) a CD song to a Wave file, then convert to MP3; some of the data will be lost. If you record directly off the CD, since there is more data, the quality of the MP3 should be better. The more data, the better quality or playback. It can get pretty technical but there are a couple of basic factors to consider that impact the quality the most.
MP3 encoders employ several techniques to compress the data files:
Perceptual encoding and data compression. Perceptual encoding algorithms determine sounds below the minimal human hearing threshold, usually between 2Khz and 5Khz, and remove that data. Encoders use a psychoacoustic method which duplicates the ability of the human ear to determine which sounds are being masked by the louder sounds and aren't necessary to store.
There are also three basic compression methods that don't address the audio but the method used to store the data. These methods combine common elements and removes any repeated data. It's much more technical but that's the gist of it.
An MP3 cannot match the quality of a properly recorded Wave file since there is a greater amount of compression and some data is eliminated by that process. But the MP3 format was invented to make music files smaller and easier to download from the Internet. The highest quality music files come from either the CD originals, or from AIFF / WAV uncompressed or FLAC / Apple Lossless compressed downloads.
If you want the portability and smaller file size, MP3 files will fill the bill. To get the best quality, use a bitrate of 192 kpbs. If you absolutely need the best quality recording, stick with uncompressed file formats.
