The RPM Guide to stem mixing.  What is it, how do it do it and why you should care

The process of recording is full of techniques which may not have an immediate benefit towards the overall completion of your RPM album(s), but may be useful in other aspects of your musical endeavors. Stems mixing, also known as sub-group mixing, is one such technique. Generally used by mix engineers to mix down large groups of instruments into a mono or stereo track, this technique began as a way to simplify processing large groups of similar instruments, like drum sets, backing vocals and string sections, as well as circumvent track limitations. These days, when modern digital audio devices and increased storage allow for much greater track counts, stems mixing has developed into a great way to prepare your material for collaborations and other uses as it allow the end user to retain some multi-track capability without having to struggle with software incompatibly issues.

What Are Stems and Why Should I Care?

Stems are simply a group of separate instruments that are mixed down to a mono or stereo track. The best example of this can be seen in through the recording of a drum set. Typically, a basic mic'd drum set will have a microphone on the bass drum, the snare drum, and the hi-hat as well as at least one mic on the toms and two overhead or room mics. (Note- this set up is simply for the purposes of analogy, many other mic techniques can be used to mic a drum set!) This means that a typical drum set mic'd in this manner would have six microphones used, all going into separate tracks on a mixing board. Since the individual drums are all parts of the same musical element, it can be advantageous to mix all of them down to a stereo track rather than leave them as six separate tracks to be dealt with during the final mixdown. This not only reduces the CPU cycles you are using as your computer now only has to deal with two audio tracks for the drums rather than six, but it allows you to control the entire drum track with greater ease. Once you have mixed your drum track down it also becomes much easier to send it to a friend in this format so that they can use it to add say, a bass part, and so on. For collaborative purposes this means you can send over a simplified multi track session, with the important elements (say, drums, guitars, etc) separated into their own tracks. This gives the collaborator more control of the individual elements than if they had a single stereo file of all tracks, and it also eliminates compatibility issues if they aren't using the same recording platform as you are.

Recently it has become more common for bands like Radiohead and Nine Inch Nails to release stem mixes of their songs for fans to remix. This means the bands can distill the important element of the tracks down to a smaller number of files, facilitating the ease that the fans can work with the material on any platform while still allowing for the ability to work with separate instrument parts. Another example of stems mixing is seen in the popular Guitar Hero and Rock Band games. In this case, popular songs are separated into the elements that are to be followed during gameplay, (like the drum parts, guitar bass, backing vocals, etc.) but similar instruments are also combined together to avoid clutter and confusion. For instance, many of the Beatles songs feature overdubbed vocal harmonies by all members, but for game play purposes, all backing vocals are mixed together to form one track. Likewise, bands looking to contribute their own tracks to the new Rock Band Network will need to provide submixes of various elements of their own tracks in order to make them fit the standards for gameplay.

Awesome! Now How Do I Do This?

There's nothing difficult about creating a stems mix. Although the specifics will be different from DAW to DAW, the concept is universal. There are several ways to create a stems mix, but two basic methods cover most of the ground. In both situations you will need to begin by choosing the tracks you'd like to mix down. Lets use our mic'd drum kit from above as an example for the first method. Create a new stereo track. This can be an entirely new stereo audio track or an Aux track. (Some DAW's also allow you to create a Group track as well.) Now select all the various drum tracks you would like to mix down. Next, set the outputs of each separate element to feed the input of the submix track you just created. Arm the submix track for recording and off you go. The second method is similar, but involves rendering or bouncing to disc a selected group of tracks. This is useful if you're simply making a stems mix for the purposes of collaboration or remixing rather than for the purposes of using it in a final mixdown. Start by selected your submix elements, but don't create a new track. Many DAWs, like Ableton Live and Cockos Reaper, have a bounce to disc or render function, and you simply need to bounce the selected tracks down to a new stereo track. If you have recorded any automation, such as volume or pan changes in any of the sleeted tracks, your bounced file will reflect all the changes as well. In Garageband the only difference is that you need to solo the tracks you want to bounce down rather than selecting them, then choose "Export Song To Disk" from the "Share" menu. This will create a mix comprised only of material in the soloed tracks. In all cases, make sure to give your submixed tracks a reasonably descriptive name, like "Drum Mix" so you can keep track of what it is.

If you've ever found yourself wanting an easy way to collaborate with someone who doesn't use the same recording software that you do, publish a track online with a Creative Commons license so that others may remix and sample various elements from it easily, or possibly even submit a track tooth Rock Band Network, stems mixing could be your friend. It's an easy and useful method to make your mixes less complicated while still retaining a level of control over the final balance of the track.