Mastering is the final stage in the production of a record. It’s the final check on everything that’s been done to that point just before it’s released into the wild. This can vary from just a simple quality control process, where very little is done to the audio aside from listening, to something much more involved, where various digital and analogue tools might be used to get the mix to where the mastering engineer feels it needs to be.

In the majority of cases this process falls somewhere in the middle. The engineer will make a few choice changes to the mix to bring out the best bits and take care of anything they feel would compromise the song translating well across different formats and various real-world listening situations.

To do this a mastering engineer will use a fairly narrow selection of processes; namely dynamic processing (usually compression/limiting), equalisation (maybe adding bass or removing a harsh frequency), and stereo imaging (are some elements too wide, or not wide enough?) Most importantly this all takes place in an accurate listening environment that the engineer knows very well.

That is mastering, and that is the same regardless of the source material. The terms stereo mastering and stem mastering refer to the source material. Stereo mastering is done from a stereo mix, while stem mastering is done from the component stems that are used to create that stereo mix.

Stems are usually groups of instruments. For example; all of the elements of a drum kit might create a ‘Drums’ stem, the bass guitar and extra synth sub might create the ‘Bass’ stem, some filtered atmospheric sounds might be summed together to create an ‘FX’ stem.

All of these stems will be printed with the processing and effects used during mixing so that when they’re imported by the mastering engineer they can be summed together to recreate the stereo mix.

Referring back to the first couple of paragraphs above, stem mastering utilises the same tools and the same mindset but with access to the component parts that created the mix.

Let’s use the same processing examples above and see how mastering from stems might affect those decisions:

Would this specific element benefit from more compression?; Maybe removing a harsh frequency from this element without affecting the others is more transparent?; Is this element too wide, and this one not wide enough?

Stem mastering is not a mix fixing service, but rather another way for a mastering engineer to get the most out of a mix so that a song translates in the best possible way across as many real-world listening situations as possible. After all - couldn’t the mix engineer just add that little bit of extra top end that the stereo mastering engineer added too? Or does that defeat the point of having a fresh pair of ears make those crucial final adjustments?