First, I want to clear out the difference between mixing and mastering:
In mixing your overall goal is to get all the sounds to fit together in the track and emphasize the feelings and emotions of the track.
In mastering your overall goal is to make the track ready for distribution and making sure that the quality of the track stays the same on all sound systems from small shitty speakers to very good sounding surround sound speakers.
Also in mastering the later years (last 20-30 years). Mastering engineers has started to correct any mistakes that should appear in the mixes they get.
This is where many people get lost because many people think you can do mastering in the mixing process and you can to a certain extent. But then many people think “I do not need mastering just a limiter on the master buss then”. That’s where you are wrong. You still need all the other steps in the mastering chain to making your track ready for distribution.
1. Editing minor flaws.
2. Applying noise reduction to eliminate clicks, dropouts, hum and hiss.
3. Adjusting stereo width.
4. Equalize audio across tracks for optimized frequency distribution.
5. Adjust volume.
6. Dynamic range compression or expansion.
8. Peak limit.
In the first step you can see that the mastering engineer is correcting any mixing mistakes.
In the second step the mastering engineer is reducing annoying noise and clicks as good as he can.
In the third step the mastering engineer is adjusting the stereo width of the mix, so we don’t have any sound cancellation or have the mix too mono.
In the fourth step the mastering engineer makes sure that the frequencies in the mix is adjusted in best possible way to fit into the different distribution platforms while still keeping the quality (Spotify, SoundCloud, iTunes etc.) he does this because every platform is applying eq, compression and limiting to the tracks (also Facebook).
In the fifth step the mastering engineer adjust the volume of the track (loudness) so it fits in with the “standard”. The new standards are around -14 LUFS. Because the streaming platforms have signed an agreement that states the streaming platforms have to keep the sound levels around -14 LUFS.
(Note: you must look at the integrated LUFS metering)
In the sixth step the mastering engineer makes sure that the track has a good dynamic range there is not too much or too little. This is achieved by compression or expansion (expansion is the opposite of compression).
In the seventh step the mastering engineer sets the peak limit, where he makes sure that the track isn’t clipping. Also you want to introduce true peak limiting which I think you should read about by clicking on the picture below. There’s a reason I don’t cover it in this post as it’s because it would take too long.
In the eighth step the mastering engineer makes sure that the quality stays the same whatever if it’s a wave file, mp3 file or any other file. He does this by using dithering.
Dithering is about adding a little bit of noise to the track so that if the track is played on a sound system with lower quality or is transposed to a file with lower quality the sound will stay the same.
If you want a more in-depth explanation here is a link: http://www.earlevel.com/main/1996/10/20/what-is-dither/
Remember these steps can vary depending on the track that have to be mastered. Because all tracks are different.
And sometimes a track requires you to use multiband compression or little bit saturation to get the balance you need. Be careful because multiband compression and saturation can very easily ruin the tonal balance of the track which you don’t want.
Note it is also important to remember when you release an ep you have more than 1 track. That’s where it’s also the mastering engineers job to make sure that the loudness of all the tracks are equal and that the tonal balance of all the tracks fits together so you get an overall feel that the tracks belongs together.