How to Open Up Space in a Mix
I’m writing this on a train up to NY. I’ve got K’Naan’s album “Troubador” playing through my headphones right now, and really vibing on the music. But as I often do while listening — even casually — I’m admiring the arrangement of sounds, textures, movement, and tones. Hats off to Manny Marroquin, as these mixes are incredible.
It’s got me meditating on how to “open up” a mix — to allow things to breathe and move while still being full and textured.
The reality of getting a mix to sound “open” is that it’s not one single thing.
Most people will attribute openness to the instrumental arrangement and the way the sounds are captured and EQ’d. This is true. But it’s also the space and dynamics. And the front to back perspective of the soundscape. It’s sort of the sum of everything.
Most obviously and most accessibly, the actual instrumentation of a song will define the openness of it.
Some things are meant to be more sparse and others are meant to be fuller.
Often times, it’s the change from sparse to full or full to sparse that gives a song momentum.
Solo banjo will invariably be “open sounding” whereas a heavily layered synth orchestra a la Stargate (like the chorus of “Black and Yellow” by Wiz Khalifa) is going to be “dense and full.” Neither is necessarily good or bad. However, there are instances where you find fully orchestrated records that still have a very open sound.
It’s more rate to find the opposite: a sparser record that is still dense and full — but engineers such as Young Guru and Mike Dean have managed to make that work.
So aside from arrangement, what else is there? Well, there is also this concept of “air.”
A lot of times, air is prescribed to the super treble frequencies, which I think is appropriate at times. But for me, air is very literal. It’s actual air. It’s very challenging to get a song that was recorded purely close capture and in dead spaces to sound “open.” I’m usually just doing my best not to make it sound claustrophobic.
If the space of a record is all happening within a one foot zone, then the way we perceive sound dictates that we will hear it that way. Whereas a record recorded with natural space, and different proximities, will be perceived as open because there is a wider psychoacoustic plane. We can brute force this illusion with reverb algorithms, but it’s not truly the same.
Compromise + Negotiation
Now we get to the mix concepts. “Open” and “full” are a negotiation.
With too much content flying about we get a dense, muddy mess. With too little intersection of sound we get disjoint and thinness. Mixing in this regard is about compromise and negotiation.
The “yes, I can take a little 2 kHz out of the cymbals because the guitars are already possessing that space.” The cymbals would most likely sound better with that 2 kHz content — that’s the body zone where the bell rings out. But in context of a screaming guitar, that part of the cymbal is really not coming through anyway — and might not be needed. Similarly, a low pass on those guitars might be good for the drum overheads — so the treble from the cymbals really rings through clearly.
This idea of putting things in their own space is a fundamental one. It’s one of the primary principles of mixing a record. But I think there’s a tendency to overemphasize this separation concept.
As important as it is to allow things to breathe by creating room for them, one has to remember that sounds often reinforce each other as well. So taking too much away is not really a good thing. It’s important to be decisive about what really has to go.
EQ vs Dynamics Processing
Part of this tendency to over-separate elements may come from the fact that EQ is played up as the tool for this, while dynamics processing is often overlooked for this process.
The reality is, if you have a long sustaining snare or clap tail eating up all the space for the acoustic guitar, you would have to EQ the life out of that snare to get it out of the way. But by using a little dynamics processing, you may find no need for EQ at all in this regard.
And this goes both ways. In the case of the overly elongated snare tail, you would want an expander to shorten it. In the case of, say, an acoustic guitar stepping on a vocal, it’s very often the peaks of the strums that are jumping forward. A little compression to tame down the attack will naturally separate the acoustic guitar and vocal without any need for EQ at all.
It’s all about space. That’s the left, center, right of the pan sphere, the front to back of the imaging, the top to bottom of the frequency range, and the dynamic space.
So while I’m writing this, listening to “Wavin’ Flag” by K’Naan — I’m taking note of the space, dynamics and tone — and how that all comes together to create an open mix.
- In the verse we hear K’Naan’s voice very forward. The ambience for the voice is primarily a delay rather than a reverb.
- The drums are a little behind that with a bit of plate sounding verb coming off the snare, and a good amount of the attack on the drums held down.
- The kick is fairly forward, but it’s in a very different sonic plane than the vocals.
- In the far back is an ambient pad that’s just kind of filling things out and really giving a sense of depth.
Then the piano comes in — just in front of the drums and just behind the vocal — but still very close. It’s tonally pocketed around the vocal very nicely. It doesn’t sound like any content is missing from the piano, but I suspect if we heard it in solo there’d be a couple frequency notches missing. My guess is they’re right around where that guitar is primarily living — the one that shows up in the chorus. At the risk of being absolutely wrong, I’d guess just shy of 1 kHz.
It’s all of these things combined and occurring throughout the mix that allows it to cohesively fit together with a sense of openness. Hopefully I can get Manny to weigh in with his thoughts!