Here is another piece I wrote for 301's blog to explain what 'Stem Mastering' is.

Working as mastering engineers, we tend to receive songs at varying amounts of mix standards, from the very accomplished, to the developing. Regardless of level, our job remains the same; to bring out the best in these songs and have them sound as good as they can, and on as many sound systems as possible.

Occasionally, we are sent a mix where a standard master may not be able to bring out the best elements in the track. In these cases, one would ideally revisit the mixdown, make the tweaks and send a new mix to mastering. But sometimes the client doesn’t feel confident or knowledgeable enough to address the mix issues, but also can’t afford the cost for a full mixdown.

This is where something that we call a ‘Stem Master’ might be a good solution. A Stem Master is a method to process sub-grouped elements of the song before combining them and mastering this sub-mix. In simpler terms, it’s like a halfway step between a mixdown, and a master, both in task and cost.

For example, a common issue we find is that some people are unable to get the definition and separation required between the Kick and Bassline in order for the song to be punchy. It could be because the bassline is a too loud relative to the kick, causing the mix to sound bottom heavy and sluggish. EQ at the mastering stage sometimes cannot remedy this because increasing the frequencies around the kick, is going to bring up those same frequencies in the bassline. The only way to solve the problem, is to go back to the mixdown. The problem remains however, the client may not be able to solve this.

Proceeding with a Stem Master, we then request the track to be bounced without the kick, and the second stem of the kick by itself. (Or perhaps, bounces of the Kick, Bassline, and the rest of the music, as stereo stems, all starting from Bar 1.)

This allows us to place the stems in parallel, apply the EQ and compression to each track, and combine these macro elements to a balance of our judgment. We then apply our treatments to the resulting ‘Mix’ to complete the master.

Another example could be that a mix we’ve been sent has a messy mid-range where synths (or guitars) are fighting against the vocal, and the song lacks a clean and clear vocal. Again, because the vocal and synths occupy a similar frequency range and clash, any EQ applied to bring out the vocal will also be applied to the synths, and any fix is limited. Not knowing how to solve this, our clients will opt for a Stem Master and send stems of the Music (no vocal or synth), the synths, the vocals. We can then apply EQ to these individual parts, combine them, and master it.

This technique should again demonstrate that it is a middle step between a full mixdown and a master, for those looking for a cheaper solution than getting the track mixed down externally.

It’s a semi-flexible system that is often used by our client to solve their problems due to it’s cost effective nature, but we need to place some caveats on what is permissible when choosing a Stem Master.

They are:

  • The max number of stems is 6 stereo stems (client to choose what combinations to send, depending on where they see problem areas in the original mix);
  • All vocal stems must have their intended Reverbs, Delays and effects in place. We will limit our treatments to applying EQ and a balance, and choose not to veer into what is deemed ‘production’ or a ‘mixdown’.
  • We can apply limited automation if placing the vocal level is an element the client has struggled with.

If you have any queries about this at all, feel free to leave some feedback below, and I'll answer it.