Because "won't sit" can be several different problems, and each one needs a different fix. Here are the four I hear most. The vocal might jump between too loud and too quiet. It might sound like it's in a different room from the music. It might sound like it was recorded on different gear from the music. Or its tone might be fighting the track. In every case the symptom is the same: the vocal feels separate from the music, like it's been laid on top rather than placed inside.
So before you touch a plug-in, work out which one you've got. Here's what a vocal that sits actually sounds like, then the four possibilities, then a twenty-minute way to find yours tonight.
What "sitting" sounds like
Picture a stage. The singer is in the middle, in the spotlight. The band is around them, supporting, cradling the lead. The singer is clearly the focus, and they're clearly part of the same scene. Nobody in the audience thinks they're watching a singer on a screen in front of a band.
That's a vocal that sits. It has the spotlight, and it's connected to the scene. A vocal that won't sit has stepped out of the scene, either forward, back, or sideways into a different room. Each of the four possibilities below is a different way of stepping out.
First, a confession
When I started out in the late 90s, recording to a Tascam four-track and then on Pro Tools at RMIT, my mixing process went like this. Drums first: gate, compress, EQ, samples, snare reverb, room reverb, balance, refine. Hours. Then the bass. Then guitars, keys, synths, strings, percussion, and on and on. Then, finally, the lead vocal, and I'd think: well, I guess I just compress it, add some EQ, find a nice reverb and automate the level.
I had my ratio backwards. The listener connects with the vocal more than with anything else in the record, and I was giving it the least attention of anything in the record. I've since heard the same thing in hundreds of demos and home recordings. If your vocal won't sit, check the ratio first. Then work out which of these it is.
1. Volume: it jumps between too loud and too quiet
This is a dynamics problem. The gap between the vocal's loud moments and its quiet moments is bigger than the music around it can hold, so it pokes out, then disappears.
The fix is two stages, in this order:
- Control the dynamics. Clip gain on the phrases that are obviously too loud or too soft. Gentle levelling. Then compression to narrow the gap between loud and quiet. On my own mixes that's often inline compression plus parallel compression and parallel limiting, until the vocal is genuinely consistent.
- Then ride the fader. Once the dynamics are controlled, the fader moves should be small, a few dB either way, just floating the vocal over the song. I do those rides on a fader on a control surface, not with a mouse, because it's a performance.
Ride the fader first and you end up chasing a vocal that won't hold still.
Two things to listen for before you move anything:
It's often just the front of the word. A singer who hits the start of a word hard makes the whole vocal sound too loud, when really it's the first few milliseconds. Draw automation to pull down just the front of that word, or to lift the middle of it. That's detail work, and it's where I use the mouse.
One word can colour a whole section. A single word poking out at the start of a phrase makes the listener feel the whole section is too loud. When a client tells me the vocal's too loud in the chorus, I don't turn down the chorus. I listen first. Very often it's the first two words, or the first phrase. Is it the whole section, or a moment?
And one more: sometimes a level problem is really a space problem. A vocal in a different acoustic space from the music gets heard as too loud or too quiet, even when the level is right. Which is the next one.
2. Space: it sounds like it's in a different room
Reverb pushes a sound to the back of the stage. So does a darker tone. A dry, bright vocal sits forward; a dark one drenched in reverb sits further back. If the vocal is in a small chamber and the music is in a big hall, or the music is almost dry, your ear hears two rooms. Then it decides the vocal is too loud, or too quiet, or just not part of it.
There are two ways this happens.
In the mix. The reverb and delay on the vocal don't match the space the rest of the music is in. Choose them against the track, not against the soloed vocal. If the music is tight and dry, a lush hall on the vocal will lift it out. If the music is big and ambient, a bone-dry vocal will sit in front of it like a sticker.
In the recording. This one's harder. Say the drums were tracked in a good studio, the guitars and bass went in direct through amp simulation, and the vocal was recorded in the lounge room. That vocal has the lounge room in it: the plaster walls are a reverb, baked in, and you can't turn it off. If you don't like it, you have to reduce it, with a de-reverb plug-in or a transient shaper turning down the decay. Both help. Neither is free, which brings us to the honest one.
3. Quality: it sounds like it was recorded on different gear
I wish I could tell you a $200 interface and a $300 microphone in a bedroom gets you the same result as a vintage valve microphone into a classic console preamp, through high-end converters, in a great room. It doesn't. They're miles apart, and it's hard to overstate how much of that is the room. A mid-range mic and preamp in a bad room still sounds like a bad room.
Now put that vocal on top of music built from samples and virtual instruments recorded in world-class spaces, and you've got a quality gap. The music sounds expensive and the vocal doesn't, so they don't sound like they belong together.
I'm not telling you this to make you feel bad about your gear. I'm telling you so you can stop blaming yourself. To a point, it's not you. I'd estimate around 95% of the problems I see enthusiasts and early-stage producers wrestle with simply don't exist in a proper facility. They never come up in conversation.
Here's the principle that matters: at any level of production, as soon as you're doing corrective work, you're degrading the quality. You can mask the problems to a degree, but every fix costs something. So your real options are:
- Make it a feature. Tape saturation fills in harmonics. Lean further into a degraded, lo-fi tape sound and the slight lack of fidelity becomes the aesthetic. If your room is ambient and you can't control it, play to the room rather than against it.
- Re-record somewhere better. A better-sounding room beats any plug-in.
- Hire better gear for a weekend and record at home with it.
- Treat the room. Buy acoustic treatment, or make some.
Singing softly helps more than people realise. Billie Eilish can record vocals in a bedroom because she sings very softly, so the room barely joins in. Sing with any volume and you excite the room's resonances, and its problems end up all over the vocal.
4. Tone: it's fighting the track
A tone problem sounds like a vocal that's too harsh, too thin, too boomy on certain notes, or dull against a bright track. The temptation is to fix it with EQ straight away. Best practice is to go back to the source first, in the order the sound travelled.
- The singer. We can only work with what we've got. Can they engage the diaphragm more, improve their breath control or their enunciation? A better tone at the source beats any amount of processing.
- The room. If the tone is mostly right but certain notes suddenly bloom or vanish, that's more likely the room than the singer. Those are the room's modes interacting with the vocal. Every room has resonances, and depending on where you stand, some notes get louder and others almost disappear. Walk around the room singing those notes, feel where it sounds fullest, and sing there.
- The microphone, as an inverse match. A bright microphone accentuates the sibilance in a sibilant voice. A dark microphone makes a dark voice even duller. So a bright, sibilant voice wants a darker mic, and a dark voice wants a more present one.
- Then processing. A resonance suppressor, which automatically finds and tames frequencies that keep jumping out, across the presence range, the high mids and the low mids where most of the vocal's body lives. Or look at a spectrum analyser, find the peaks and the dips, and use a dynamic EQ or a multiband compressor to tame what's too loud and lift what's missing.
What makes it worse
Most people reach for the EQ, and in principle that's right. An equaliser is frequency-dependent gain, and the idea comes from telephones. Over a long cable run, the treble in someone's voice disappears, so at the receiving end the line was equalised: the treble boosted back until it sounded like the person who spoke. EQ's job is to turn the source back into the source.
Here's the problem. If you're mixing at home, your listening environment is almost certainly working against you, and your EQ moves end up correcting your monitoring, whether that's your room, your speakers or your headphones, instead of your vocal.
- Dull headphones, and you'll add treble the vocal didn't need.
- Bright headphones, and you'll cut treble it did need.
- A spot in your room where the low end cancels out, and you'll add warmth the vocal already had.
Then you play it in the car and it's a different song.
My rule of thumb: do 70% less. If you were about to boost 10 dB, boost 3. Your mixes will translate much better. Then check the spectrum analyser, because it tells you the truth when your room doesn't. And if you can, calibrate your monitoring. I'd do headphone calibration first: room calibration takes more effort and money, and speaker calibration software is only as good as the room and the speakers it's correcting. I use IK Multimedia's ARC On-Ear on my headphones, and I love it.
The other usual suspects:
- Too much reverb, which muddies the vocal, accentuates the sibilance and makes the words harder to understand, as well as pushing it back.
- Compression pushed to extremes without knowing what it's doing.
- Exciters, which add treble that turns harsh.
- Too much saturation, which adds mud.
- Too much automatic resonance suppression, which leaves the vocal thin and slightly unnatural.
They all have the same root. You're making the vocal sound good in your environment, and your environment may be lying to you. Everywhere else, the problem gets worse.
Tonight, in twenty minutes
Treat this as an experiment, not as "I'm finishing the song tonight".
- Mute the vocal and listen to the music. Does the music feel right on its own? If it doesn't, the vocal isn't your problem yet.
- Solo the vocal. You're not listening for how it fits. You're listening for problems in the vocal itself. Pull off all the processing. Listen on more than one set of speakers, and look at a spectrum analyser for anything obviously wrong.
- If the room is in the vocal, deal with that first. Reduce the recorded reverb before anything else.
- Pick one section. The first verse, or the first chorus. Bring in the key harmonic instrument, the piano or the guitar, and nothing else. How does the vocal sound against it?
- Shape the vocal against that one instrument. Compression, parallel compression or limiting, saturation, EQ, reverb, until the vocal sounds great with just that one part.
- Bring the rest in one at a time, in order of priority. Ask yourself: if I could only have one more instrument, what would it be? Bring it in and get it sounding good. Then ask again.
- Notice when it falls apart. If the vocal was sitting and then one instrument comes in and it stops sitting, you've found a masking problem: that instrument is standing where the vocal needs to be. That's the same mechanism behind a muddy mix, and the same fix applies.
Three questions people ask next
Why does my vocal sound on top of the mix? Usually because it's in a different space from the music, or it's brighter and drier than everything around it. A very common case is a vocal recorded over a purchased backing track or a pre-mixed stereo file, where you can't mix the vocal and the music together because the music is already finished. Then the job is bringing the two into the same space: share reverbs and saturation across the vocal and the backing, so both carry similar characteristics. Match the space before you touch the level.
Why is my vocal too loud in some places and too quiet in others? Because its dynamics are broader than the music's. Control the dynamics first, with clip gain and compression, then ride the fader. And check whether it's the whole line or just the front of one word.
Should I just turn the vocal up? Only after you've worked out which of the four problems it is. If it's space, quality or tone, more level makes it stick out more, not sit better.
Where to go from here
Every one of these comes back to listening for the relationship between the vocal and the music, not the vocal on its own. If you want to talk it through with people who do this every day, request access to the Academy of Audio community. If you've made things, hit the ceiling, and want to learn the whole process in a room with people who do this for a living, that's what the intensive is for.
And check your ratio. The vocal is what people listen to. Give it the time.
