What Are Audio Stems? (Mixing & Exporting Audio)
Having access to proper audio stems is useful in music production, mixing, editing, scoring, and other productions, giving us a good balance between simplicity and flexibility. If you're wondering about what stems are and how you can use them or create them, you've come to the right place.
What are stems in audio and music production? Stems are submixes of a mix, typically in stereo, that combine similar elements together for further processing and/or editing down the line. For example, all the drums could be on one stem, all the vocals on another stem, and so on. Stems can be utilized for mixing, editing, scoring, and more.
In this article, we'll discuss stems in more detail and how they're used in audio.
What Are Audio Stems?
Stems are audio files (typically stereo) made up of multiple tracks of similar information. Many times, the stems of a mix session will coincide with the subgroups.
For example, we could have a vocal stem with all the vocal tracks, a guitar stem with all the guitars, a drum stem with all the drums, and so on.
Each stem of a specific project will have the same length, spanning the entire duration of the song. This way, a mix engineer can play back all the tracks in sync as he or she inputs them into a mixer or imports them into a DAW.
Stems can be either from an unmixed session (often the case if the mixing engineer specifically wants stems) or from a mixed session (often the case with stock music licensing).
When stemming down a project, it's always best to, at the very least, have a proper balance between the various instruments in any given stem. The balance of level between the tracks within a stem cannot be altered after the stem is created (though there are methods of “cheating” some balance shifts using EQ, compression, etc.).
They generally don't include effects return channels or other auxiliary channels or buses, though these returns may be stemmed for future use in some cases.
Mixing With Stems
Stems can be used for mixing, though multitracks are often preferred. We'll get to the differences between the two in the following section.
For now, let's assume we have stems for a project. This is commonplace with stock music licensing for video and can also be the case for music by itself if stems are all we have access to.
For simplicity's sake, let's assume we're working with digital audio, with a folder of stems we're importing into our digital audio workstation. The same general process can be followed by routing the stems from an analog source (like a tape machine) into an analog console or audio interface (for DAW compatibility). Still, for this section, we'll focus strictly on digital.
Alright, so we have our stems folder. Hopefully, everything we need is held within the stem, and all issues of phase/polarity, timing and tuning have been addressed appropriately.
We'll go ahead and import the stems into our session, adjust the session's tempo(s) and time signature(s) if need be, and ensure everything's looking good. Every stem should be the same length and sync perfectly in the session.
We can then go about fixing any of the aforementioned issues if fixing is required (phase/polarity, timing and tuning).
Gain staging is a great next step. This sets each track/channel at the same level with faders at unity and allows for proper headroom, plugin performance, and even A/B testing throughout the mix. We can use gain/trim plugins for this process, inserting them on each track to adjust the gain as necessary (we can use the preamp gain control on hardware consoles).
From there, we're ready to mix using our stems.
Now, if we're utilizing stems from a licensed piece of music for scoring to video, we're likely working with something that's already been mixed to a decent standard.
Of course, we can always alter the mix of the stems to our liking.
However, the big benefit of stems in licensed music is that we can effectively build the song to be longer or shorter and have elements weave in and out to our liking. Looping and shortening full mixes is very difficult, and stems make these processes much easier to accomplish.
The Difference Between Stems And Multitracks
The terms “multitracks” and “stems” are often confused. Let's clear up the difference.
The term “multitracks” actually refers to each and every track of a session, whereas stems are groups of similar tracks. For example, the “drum stem” would be a stereo audio file of all the drums in the mix, while the “drum multitracks” would include the kick in, kick out, snare top, snare bottom, rack tom, floor tom, overheads, etc.
So, the multitracks have each individual element of the mix, and each stem will be made of one or (typically) more multitracks.
Related Article On Multitracks
To learn more about multitracks, check out my article What Are Multitracks? (Audio Recording, Mixing, Playback).
Bouncing Down Stems
Bouncing out stems means we must first find all the different groups of tracks in the session.
For example:
- All lead vocals could be on one stem
- All adlib vocals could be on one stem
- All background vocals could be on one stem
- All bass could be on one stem
- All electric guitars could be on one stem
- All acoustic guitars could be on one stem
- All drums could be on one stem
We can go a bit further here and have separate stems for solo instruments (say, a guitar solo, for example) or split different percussion (drum kit on one stem, percussion on another, for example), but I digress.
We then bounce, export or otherwise record each of these groups of tracks (and potentially each bus and effects return) down to its own audio file (often stereo), spanning from the beginning of the song's arrangement to the end.
All stems will be of the same length, meaning there's a high likelihood there will be silence in sections of certain stems.
Producing stems isn't necessary if you'll be the one mixing and are already familiar with the current session. However, it can help to clean up the session whether you'll be the one mixing or not.
If you're composing music for licensing, you'll likely want to mix first and then stem down to give the music director as much flexibility as possible in scoring without overloading him or her with high numbers of multitracks.
Depending on your workflow, you may want to keep the original tracks in the session just in case you need to reference them in the future. This can be a good strategy if you're mixing your own material.
In this case, you can bounce each track as its own audio stem and subsequently mute the original track, bypass or turn off its plugins, and hide it from view.
Bouncing Down Multitracks
Bouncing down mutlitracks follows pretty much the same steps as bouncing down stems. The big difference is that we'll export each track of the session individually.
We're still ensuring we're bouncing the entire session length, but we're bouncing every track rather than combining several similar tracks.
If you're planning on mixing in a completely new session or want to send the session off to another mixing engineer, creating multitracks is warranted. Having a single audio file for each track (from start to finish) means we can import each “multitrack”, line them all up to the same starting point, and begin mixing. That said, it's probably best practice to multitrack your mix preparation tracks whether you're mixing yourself or not.
Call To Action!
Take a recording, production or even amix session you've completed and create your own stems. You can do so by soloing a group of insturments and bouncing them out or soloing individual subgroups and bouncing them out (assuming you've grouped like instruments with like).
Consider stemming all drums and percussion together, lead vocals, background vocals, bass, guitars, etc. all on their own stems.
Have any thoughts, questions or concerns? I invite you to add them to the comment section at the bottom of the page! I'd love to hear your insights and inquiries and will do my best to add to the conversation. Thanks!