There are a lot of audio file formats. Here are the ones you encounter most commonly.
Analog formats
Recorded sound consists of fluctuations in electrical current coming off of a microphone or mixing desk. Before computers, you translated that current into tiny smooth wiggles in the shape of the groove cut into a vinyl record, or tiny smooth wiggles in the alignment of magnetic particles embedded in tape. You can reproduce the original electrical current by dragging a needle along the groove, which vibrates a little magnet, or by running the tape over a little magnet.
Examples: Wax cylinders, vinyl records, reel-to-reel tape, cassettes
Pros: Analog formats can sound really great if your media are in good condition, and if you are listening through a good sound system.
Cons: Analog formats can sound terrible if the media get scratched, dusty, or demagnetized. You need to be very careful about physical degradation–every time you listen to a tape, you scrape a little bit of the coating off. You can’t make copies of analog media without introducing noise. And analog gear is expensive.
Uncompressed digital audio
Computers convert the electrical fluctuations coming off the mixing desk as a long string of ones and zeroes. You can’t represent a smooth wiggle in discrete increments with perfect accuracy, but if you use enough ones and zeroes, you can get close enough that no one can tell the difference.
Examples: AIFF, WAV, Compact Disc, Digital Audio Tape (DAT)
Pros: Uncompressed digital audio sounds great, as good as analog at high bit rates. Digital audio files can be copied perfectly as many times as you want. Digital media are cheap and durable.
Cons: Uncompressed audio produces enormous file sizes, ten megabytes per minute at standard CD quality, and many times that using higher audiophile settings. That’s a problem if you want to send music over the internet.
Lossless compression
Have you ever wondered how court stenographers can write down a word-for-word transcription as fast as everyone is talking? They have a shorthand that they write with special keyboards. The shorthand is much more compact than full English, but it can be converted back into full English without any loss of detail. You can do the same thing with audio. You exploit data redundancies in audio file formats, which makes it possible to reduce their file size without sacrificing any sound fidelity.
Examples: FLAC, Apple Lossless
Pros: Somewhat smaller files with exactly the same sound quality as the uncompressed versions.
Cons: These formats are only used by audiophiles and nerds, and not all software can read them. Also, the files are still pretty big.
Lossy compression
If you are willing to sacrifice some sonic detail and nuance, you can shrink your audio files down to a fraction of their original size. Lossy compression is like SparkNotes. The SparkNotes version of War and Peace is much shorter than War and Peace, and it gives you a pretty good idea of what War and Peace is all about. But you lose nuance, and you can’t reproduce the full text of War and Peace just from the SparkNotes. MP3s are the SparkNotes of sound.
Fun fact: image compression formats like JPEG use the same algorithms as MP3. A low-quality digital image is precisely analogous to a low-quality digital audio file.
Examples: MP3, M4A, AAC, Ogg Vorbis, all streaming formats
Pros: Lossy compression can reduce your file sizes exponentially. The more sound quality you’re willing to sacrifice, the smaller you can shrink the files. That’s crucial if you want do things like stream audio on the web.
Cons: At higher compression ratios, lossy compression sounds like garbage. The more the audio is compressed, the worse it sounds. (I don’t know how people can stand Spotify.) Also, once the audio quality is gone, it can’t be restored.
Other formats to know about
Digital audio workstation (DAW) session files: Each DAW has its own file format: GarageBand, Logic, Ableton, Pro Tools, and so on. A DAW session is really a folder containing the session file itself, a folder full of audio files, and various metadata. Be aware that DAWs can’t open files created by other DAWs. So, for example, you can’t open a Logic file in Ableton, or vice versa. Also, an older version of a DAW can’t open a session created with a newer version of the same DAW. If you need to convert a session from one DAW to another, the best method is to make stems.
MIDI: Contrary to popular belief, MIDI is not an audio format. It’s more like music notation the computer, a list of numbers specifying what notes should be played and when. To hear a MIDI file, you will need a software instrument to play it back, just like you need a human performer to turn sheet music into sound. MIDI files are really tiny.
Notation: Like DAWs, every notation program has its own specific format, so Sibelius, Finale and Noteflight files are not mutually compatible. The good news is that all of these programs can both read and write a generic format called Music XML, and they also read and write MIDI.
Love the Sparknotes analogy. These comparisons make perfect sense to me.
Sounds like a pretty good summary to me. “Compact Disk” is usually spelled “Compact Disc”. Also I would disagree with the blanket pronouncement that “Lossy compression sound like garbage”. That depends on the degree of compression. In my experience, using typical reproduction equipment, most people can’t pick the difference between a standard 256K AAC file and a lossless version.