What is Smart Mixing?
"Smart mixing" is a term introduced by Moldover, a controllerist digital musician who uses a sophisticated software and hardware configuration that allows him to make fast changes to the mix with a minimum of control movement. One component of this is his "smart mixer" Reaktor ensemble, which makes it easy to do basic automated EQ mixing with a simple fader move in Ableton Live. I wanted to create a similar effect using only native Live effects or freeware (in Live 7 it's possible to do with only native audio effects).
When mixing two (or more) tracks together, the sound can be muddy or boomy if the various frequency bands of the music are simply added together. Often, DJs will pull the bass or the highs out of one track and let the other track fill in, creating a better blend. When mixing out one of the tracks, the EQ must be readjusted to bring the missing part(s) back into the mix.
The idea behind smart mixing is to automate this a bit more. Basically, you can specify which track or tracks you want to be dominant in the low, mid or high bands, and allow the smart mixer to push the program content out of the way on the other tracks to make room as the fader is pushed up. Likewise, when the track is pulled out, the other program content automatically fills back in the empty band space.
We will want to be able to specify which tracks will dominate the low, mid and/or high bands of our master program when multiple tracks are mixed together. I call this "band suppression."
Smart Mixer Theory
Warning - this gets technical, and understanding it is not necessary in order to use smart mixing, but if you get your head around it, it will potentially make smart mixing more useful to you.
The approach we're going to take here (which is very similar in result anyway to Moldover's) is to use multiband compression to achieve the smart mixing effect.
Basically we're going to run everything through three sidechain compressors tuned to separate frequency bands - low, mid and high. I call these the "band suppressors" because they will serve to suppress content in their respective bands when keyed. The sidechain keying will come from three return busses, one for each band. I call these the "suppressor key busses." The content in each suppressor key bus is filtered to match the tuning of its respective band suppressor, and is fed into the sidechain input on the band suppressor. This has the effect of causing lows to suppress lows, highs to suppress highs, etc.
Since suppression affects the entire master program, including the suppressing track, we will use suppression feedback channels to add back content from the suppressor key busses into the master, essentially negating the suppression for the tracks that have their suppression sends turned up.
Implementing a Smart Mixer in Ableton Live
In Live then, we're going to add several new audio channels to implement all of this.
First, we add a channel to act as an input aggregator to just bring together the outputs of all the deck channels.
Then, we add channels to hold the three band suppressors, and we drop a sidechain compressor and an EQ in each one.
Then we configure three return channels to act as our suppression key busses. We drop an EQ in each one tuned the same as its corresponding EQ in the band suppressor it keys.
Finally, we add three channels to feed back the suppression busses to the master (with about -3dB cut).
So - since the suppressor key busses are returns, in order to have a track impose suppression on a band, you just need to turn up the appropriate send on that track. So if you have a deck channel 1 and 2, and you want deck channel 1's bass to dominate when decks 1 & 2 are mixed together, you simply turn up the low suppression send on deck 1. Its low-band program material will now key the low band suppressor and pull down the low-band content in the program. Since deck 1's low band material is also returning to the mix via the low suppression feedback channel, its low-end content dominates.
Phew. You get a gold star if you're still with me.
In the righthand bar, you'll find the routing diagram, a screenshot of the expanded Session view, and the download link for the actual Live set if you want to try it out. Please give me feedback if you use it or improve upon it!
The Live set uses the Slim Slow Slider freeware sidechain compressor VST - available here. Any sidechain compressor could be substituted. I am working on a Live 7 version that uses the new native sidechain compressor that comes with Live.
email me - belvario at biglimemedia dot com
TEB's Sonalksis mod:
"Ok after a few hours of tweaking I got results I wanted using the Sonalksis SV719 Gate and the SV517 EQ. I used the same basic setup as Belvarios template and used the gate and the eq in a rack. Settings for the SV719 are this:
Mode: Duck
Threshold: -30 dB
Hyst: 3 dB
Range: 80dB
Attack: 10ms
Hold: 15ms
Release: 80ms
EXT: ON
for the sv517 eq just make your own EQ curves and make sure they match on the return tracks. The only downfall is that when you turn a send slowly there is a small volume drop at about the middle, but for fast cuts this thing is phenomenal. Another cool thing to try is turn on the filter section in the side chain section for extra crazyness - you can chop up individual sounds between the 2 tracks with just a twist of a send."
Note from Belvario - try adjusting the feedback channel faders a little if the sends have a funky curve.