Recording, mixing, mastering. That's the normal order of things. But when you start thinking more deeply, there are some lines that are just a little blurry.
Let's go back to the olden days of recording, when musicians would gather around the horn of an acoustic phonograph - no microphones involved, and no electricity either.
The process of recording here takes the sound waves from the instruments, collects them into the horn, and carves a wavy line in a wax disc (or in even earlier times, cylinder).
But there is a simultaneous process of mixing, where quiet instruments would be placed closer to the horn, louder instruments further away. If there was a soloist then they would be placed where they would stand out in the finished recording against the accompanying instruments.
I think it would be fair to say that there would not be any specific mastering process. Following recording, the next step would be directly to duplication.
Flash forward to the late 1940s and early 1950s when it became possible to make original recordings on magnetic tape. A band or orchestra could now sit in their preferred layout, and multiple microphones could be positioned to capture their sound. These microphones could be mixed to give a pleasing blend.
Here again therefore the processes of recording and mixing take place simultaneously.
But let's flash further forward to 1967 when Ampex created the world's first 16-track recorder. This made it possible to record each instrument and vocal to its own track. We call this 'multitrack recording'. Recording and mixing can now become separate stages of the process - First get a good recording, then when the musicians have gone home take as long as you like to create the very best mix that you can.
So that accounts for recording and mixing. What about mastering?
This is where things can become confusing.
Let's consider the recording studio of the 1970s. A band would make a multitrack recording. The engineer, advised by the producer, would mix it to stereo. The finished stereo recording would often be referred to as 'the master tape'.
The meaning of this is that this one tape is the very best version of the finished recording that exists, and will ever exist. From it will be made a safety copy, which due to the cold hard facts of analog recording will be degraded compared to the master, and a production master from which vinyl records will be made.
Aha - but then comes the process of vinyl mastering. The vinyl mastering engineer would further process the recording so that it would sound at its best on a record. A vinyl cutting master tape would be made so this process only had to be done once.
Let's hop to the present day...
When recording, let's say a band, into digital audio workstation software, the aim is to capture each instrument and vocal to the highest level of quality possible, with a certain amount of creativity according to the needs of the production.
When recording is complete, the multitrack session will be mixed to stereo. Mixing is the creative and time-consuming process of making decisions on how the instruments and vocals will be blended. When that decision-making process is complete, the result will be 'bounced', as we say, to a stereo file.
In the 1970s, we would have called this the master version but, because in digital audio infinite numbers of identical copies can be made, this one stereo file doesn't have that level of status. So we can just call it the mix.
In theory, the mix could be ready for commercial release as it is. But, just as in the old days vinyl mastering engineers could apply an extra layer of gloss to a recording as well as making it technically fit for vinyl, the modern mastering engineer will take a stereo mix and improve it still further, without reference to the original multitrack recording.
So there we have it. In summary...
Recording: The process of capturing the original sounds into the digital audio workstation.
Mixing: Blending the multitrack recording into stereo.
Mastering: Improving the stereo mix for commercial release.