BEATS WITH KEVMid/side EQ, dynamic EQ and linear phase EQ each fix a different problem. Mid/side lets you EQ the center of a stereo mix separately from the sides. Dynamic EQ only moves when a frequency gets too loud, or when another track tells it to. Linear phase changes tone without shifting timing, and costs you latency for it.
I use mid/side on every bus. Dynamic EQ comes out whenever something only misbehaves some of the time. Linear phase I rarely touch.
What does mid/side EQ do?
A normal EQ treats left and right the same. Mid/side EQ splits the signal into the mid, what both speakers share (usually kick, snare, bass and lead vocal), and the sides, what differs between them (width, stereo reverb, panned parts). Then you EQ each one on its own.
One move I use a lot: lifting the mids and highs of the sides. It opens the mix up without making the vocal harsh.
How I set mid/side on my buses
I keep Waves Center on every bus. It separates the phantom center from the sides, and I set it differently depending on the bus:
- Instrument bus: I pull the mid down a little, to taste. The vocal cuts through without me doing more work at the bus level. It helps a lot when I’m mixing top-down, or fitting a vocal into a two-track beat.
- Percussion bus: about 1 dB more on the sides. The percussion sounds more exciting.
- Drum bus: the other way. More energy in the mid, so the kick and snare hit harder.
Should you cut low end from the sides?
On some instruments, yes. Cut the side signal between 50 Hz and 150 Hz and the low end gets more room. The whole mix sounds less muddy.
Don’t do it on everything. Go instrument by instrument and decide how much low end each one should keep. If mud is your main problem, start with the 200 to 500 Hz range first.
Go easy on any side move. Big changes to the sides can wreck mono compatibility. Check the mix in mono, or run it through Signal Check: the MONO FOLD chart in the SPACE view shows how many dB each band loses when the mix sums to mono.
What does dynamic EQ do?
A normal EQ move is always on. A dynamic EQ band only moves when its frequency crosses a threshold. That puts it somewhere between an EQ and a multiband compressor. Some people reach for a multiband like FabFilter Pro-MB for this job. Others put FabFilter Pro-Q 4 in dynamic mode.
Use it for problems that come and go:
- a vocal that’s only harsh on the loud notes
- an 808 that booms on one pitch
- cymbals that spike now and then
- sibilance, so a dynamic band works as a de-esser
A static cut would dull the whole track to fix a moment. A dynamic band dips only when the problem shows up, then lets go.
Can a dynamic EQ do sidechaining?
Yes. A dynamic EQ can listen to another track, the way sidechain compression does, and the two sound different. A sidechained compressor turns the signal down by compressing it, set by its ratio, attack and release. A sidechained EQ dips only the frequencies you set, and how fast it moves depends on its own attack and release, or program-dependent timing.
- Trackspacer 2.5 by Wavesfactory runs a 32-band EQ that reacts to a sidechain. Put it on the bass and feed it the kick, and the part of the bass in the kick’s range drops for a moment every time the kick hits.
- Soothe3 by oeksound takes a sidechain too. It’s often used to keep a vocal and the instruments from fighting over the same frequencies.
- Pro-Q 4 can trigger any dynamic band from an external sidechain.
One I like to try: I sidechain the background vocals to the lead with a mid/side dynamic EQ. When the lead is singing, the background vocals dip in the mid, where the lead sits, and lift in the sides. They get out of the lead’s way, and the part sounds wider. When the lead stops, they come back.
Linear phase or minimum phase?
A standard EQ is minimum phase. It has almost zero latency, but boosting or cutting shifts the timing of those frequencies slightly. Most of the time you won’t hear it.
A linear phase EQ avoids the shift. You pay in latency, and a faint pre-ring on sharp transients. Some plugins add a middle option. Pro-Q 4 has Natural Phase. Pro-MB, FabFilter’s multiband, has Dynamic Phase, which adds no latency or pre-ringing and only shifts phase slightly while the gain is actually changing.
I don’t use linear phase much. A lot of engineers do, mostly on the master and when they EQ one copy of a sound against another, like a parallel chain, where a timing shift between the copies can cause phasing. For tracking and mixing, minimum phase is fine.
Quick guide
- Mid/side: when the center and the sides need different things. Pull the mid down on an instrument bus for the vocal. Cut 50 to 150 Hz from the sides on some instruments.
- Dynamic EQ: when a problem only shows up sometimes, or one track needs to make room for another.
- Linear phase: on the master or a parallel chain, if you can live with the latency.
- After any side move: check mono.
The EQ Trainer covers the basics these build on: frequency, gain, Q and filter shapes, on your own track.