my doubles kept stealing the centre because i was scared

I mix doubles quieter than the lead, usually a bit wider, and I keep their dynamics related to that lead so they don't take over.

1. Keep the lead in the centre - that's the vocal
2. Add the doubles on purpose - quieter, panned enough to be width
3. Let them have a job the lead isn't already doing
4. Keep their dynamics with the lead so they move as one

I used to be scared of doubles.

Adding more felt like it would take away from the main. Panning them wider felt like it would pull the listener off the middle. And once they were out wide, every timing and pitch thing would be sitting there in the open, not hidden by the rest of the mix. So I'd either skip them, or I'd add them and keep them close and loud so they still felt safe. That's how they stole the centre. They weren't width. They were another lead sitting on top of the first one.

I had to get over that and accept it's part of the process. A double is supposed to be a little human. If I hide it in the middle so I don't hear the imperfections, I also don't get the width. If I never add it because I'm protecting the main, I never get the size.

why my doubles kept stealing the centre

A double at lead level, parked near the middle, doesn't come across as bigger. It comes across as the vocal moving.

That's what the fear was doing. I'd add one, maybe two, then pull them back toward the lead so they wouldn't distract anyone. Level was doing more than pan. When they hit a word with the lead, the middle got huge and smeared. When one was a bit late or a bit loud, the whole vocal stepped sideways.

Quieter than the lead, they make it bigger. Same level as the lead, they compete with it. What actually pulled people off the middle was two loud vocals fighting for it. Once the lead was clearly the centre, a bit of width around it made the middle feel more stable, not less.

The mute test is still what I use. Lead only. If that doesn't sit in the track, doubling isn't going to help. Doubles can't be the thing that makes the lead sit. They can only be size around a lead that already has a place. That's the same idea as making room in the track first.

I also had to accept that wider does make the imperfections more audible. You hear the timing. You hear the pitch. That's the point of a real double. Cleaning the obvious mistakes is mixing. It isn't a reason to skip the take, or to hide it on top of the lead.

real doubles first, doubling plugins on a copy

A real double is a second take. Same words, slightly different timing and pitch. That's the human width.

A doubling plugin is a fake of that, built from the lead. Delay, pitch, stereo. No extra performance. MicroShift, Doubler, a chorus, a thickener - same job. It's a copy with a trick on it.

I get the recorded double in time and quieter than the lead before I reach for a doubling plugin.

If a phrase on the double is a little behind, I usually leave it. That's what makes it feel like a second take instead of a copy. If it's late enough that you hear two of every word, I move that phrase, not the whole take. And I keep it quieter than the lead. I used to turn it up to match so I could hear whether it was working. That just put another vocal in the middle.

If I'm using a doubling plugin, it goes on a copy or a send. The lead stays the dry vocal. Same idea as space: don't make the centre become the effect.

Harmonies are a different job. They're notes, not width. I still tuck them. I still don't let a harmony out-sing the melody because it has a nicer tone.

keeping them moving with the lead

Quieter and wider only holds if the doubles don't jump forward the second the chorus gets excited.

I used to ride that by hand, syllable by syllable. It worked until I actually wanted to mix. That was the same relationship I already use when dynamics work is split into parts. When I started using tether, I could set the lead as the driver and leave it. The bus is still how I group them. Tether is what keeps their dynamics with the lead through the phrase.

Tether doesn't pan for me. It doesn't decide how many doubles I add. I still have to get over adding them and putting them where they belong. It just keeps the balance I already chose.

Each double still gets the same kind of chain as the lead - stabilise, then the rest - so I'm not relating a wild copy to a controlled centre.

when i use a real double vs a doubling plugin

I use a real double when I want that human width, imperfections included. I use a doubling plugin when I need a little extra I didn't record. I don't stack both at lead level.

If the song doesn't need width, I don't double. A quiet verse with a fake stereo vocal was how I used to make something intimate feel like a preset. Most of the time doubling is a chorus or a hook for me, not a default on the whole take.

questions i still get about vocal doubling

How quiet vs the lead? Quiet enough that if I mute the doubles, the song still has a lead, and if I mute the lead, the doubles sound too small to be the track. If the double can carry the song alone at that level, it's too loud.

Won't panning them wider pull people off the lead? That's what I was afraid of. What pulled people off the lead was two loud vocals in the same place. Once the lead is clearly the centre, a bit of width around it is size. Hard-panned on a sparse mix can work. Hard-panned on a dense chorus was how I used to lose the lead in the sides.

Double the whole take or choruses only? Choruses and the phrases that are supposed to feel bigger. Verses usually stay closer to one vocal.

Should doubles be compressed the same as the lead? Same idea, not always the same settings. I want them controlled enough to relate. I don't want a smashed double next to a living lead, or the opposite.

Real double or a doubling plugin? Real if I have it. Doubling plugin if I don't, on a copy, under the lead. Both at once, both loud, was just a crowd.

I stopped asking the double to be impressive. The lead is the vocal. The double is the size around it, even if that means you can hear it's human.