|Michael Laskow contemplating stems in Studio West at Howard Schwartz |
Recording (HSR/NY) in 1987.
You’ve probably heard the phrase, “mixing to stems,” at some point in your career, especially if you’re providing music for Film and TV productions. There aren’t a lot of simple explanations of what it means or how to do it on the Internet, so I thought it might be timely for me to give you a basic explanation—albeit from the old school, which is the one I mixed in.
As with most things in audio, there are very few absolutes, and others may have varying techniques or opinions from mine. These thoughts are my own, based on my experience as an audio post engineer during the 80s at HSR/NY, the leading audio post house in NY (if not the world) at the time. I’m certain some techniques have changed in the digital era, but the basics should remain similar.
Stems originated (to the best of my knowledge) in the early post-production world to give editors and post mixers more flexibility in adding or subtracting mix elements, without completely remixing the entire track. Audio post engineers frequently created separate “stems” or sub-mixes of Music, Dialog and Effects, so they could be dropped in or out, and raised or lowered at will. That M.D.E. definition of Stems is probably the genesis of the term as it was originally used.
The simple explanation is that the mix engineer mixes to sub groups or several stereo pairs or tracks instead of mixing to only one stereo (2-track) mix, thus giving post production engineers the ability to quickly and easily drop sounds in or out without remixing the entire track.
As an example—and maybe an extreme example—let’s assume that you want to create Stems of a song’s mix. Assuming the song has bass, drums, rhythm guitars, lead guitar, stereo piano, stereo synth, lead vocal, stereo background vocals, and stereo percussion. Normally, you would just pan each of those elements within the stereo field (left, right, and anywhere in between), adjust individual levels, add effects, and you would create a stereo mix.
Mixing to Stems is virtually the same, but you would send each element (instrument) to its own pair of stereo channels (also called subgroups), that when played back at equal levels, sound exactly like your stereo mix, but give you the ability to drop each pair in or out at will. You could also raise or lower the playback level of any of the stereo pairs to change the mix balance, but without affecting the sound of the original EQ, effects and so on.
A simple way to do this would be to set up several stereo pairs of faders/channels/subgroups with the faders set at zero, and the pan pots for each pair panned full left and right. Let’s call them Pair A, Pair B, Pair C, Pair D, Pair E, Pair F, Pair G, Pair H, Pair I, and Pair J.
When you mix, the sound goes not only to the console’s stereo bus (the 2 track or stereo output of the console), but it also goes to the stereo sub groups that each of the instruments (or those stereo pairs of faders/channels mentioned before) are assigned or routed to. The bass would go to Subgroup or Pair A, drums to Subgroup or Pair B, rhythm guitars to Subgroup or Pair C, lead guitar to Subgroup or Pair D, stereo piano to Subgroup or Pair E, and so on. If you listen to all the stereo Subgroups or Pairs together (summed), your mix should sound exactly the same as the 2-track (regular stereo mix).
Using what I think is today’s common (and not necessarily accurate) definition of Stems, each of those Subgroups or Pairs is a Stem.
An even more simplified version of Stems—and maybe more commonly used by folks who aren’t high-end professional audio post engineers, but need some post mix flexibility—is to simply create a mix minus the vocals, or maybe a mix minus the lead instrument. Any variation you might anticipate needing in the future could be made into an “alt-mix,” commonly, and I believe often incorrectly called Stems in today’s vernacular.
Back in my days as a music engineer, and later as an audio post mixer, we used the count off or a click track at the head of song so the down-line mixer could later line the tracks up in time. The click or count was printed at the head of all the stereo Pairs/Stems. We would also print a range of audio tones (1K, 5K, and 10K) at the head of all tracks so down-line mixers could exactly set their faders to zero, giving them a more perfect restoration and recreation of the original levels of each Stem. I’m not sure if that’s still common or even necessary in today’s digital audio world, but suspect it would be a wise idea.
I hope this helps you better understand why Stems and alt-mixes are something you should understand and be comfortable creating at will. As TAXI brings you more and more opportunities to get your music DIRECTLY to music supervisors and leading ad agencies, you could very well need to use this age-old technique at any moment.
I just got home from having dinner with a top Hollywood music supervisor, and he told me that it’s very common for him to use Stems as he edits songs for placement in huge feature films. He feels that it gives him much more flexibility in presenting songs in the perfect light to the directors he works with. He also mentioned that he routinely uses Stems to edit and mix songs used in TV commercials!