Vocal production usually ends up with a stack: the lead in the centre, doubles panned out, harmonies wider still, and a reverb or delay filling the space behind them. Put a saturator across the whole bus and every one of those gets the same treatment.
That’s a problem, because they don’t want the same treatment. The lead needs to stay intelligible and the things around it are where you want the size.
What the lead can take
The lead vocal is in the mid channel, and it’s the part of the arrangement with the least room for harmonic distortion.
Consonants are what carry intelligibility, and they’re already the harshest part of the signal. Adding upper harmonics to a lead vocal makes the sibilance worse before it makes the vocal bigger. Push it and you get a vocal that sounds aggressive on a good system and painful on earbuds.
Low-order saturation on the mid is still worth having. Even-harmonic character of the kind a tube stage adds gives a lead more body without moving the top end much, and a small amount of soft clipping catches the peaks that would otherwise force a compressor to work harder. Keep it modest. If you can hear it as an effect on the lead, it’s too much.
What the sides can take
The side channel is where the doubles, the wide harmonies, and the reverb tails live, and it will take considerably more.
Drive there does two useful things at once. It adds harmonic content, which makes the doubles denser and gives them presence they didn’t have. And it compresses, because every saturator is a compressor with a soft knee: the loud parts of the side channel get held back, so the ambience sits more evenly behind the lead.
The result is a vocal that sounds bigger without the lead itself being touched. The intelligibility comes from the centre, which you left mostly alone, and the size comes from the edges, which you drove.
This is the case where the difference between mid/side and stereo saturation is least subtle. Try the same amount of drive both ways on a stacked chorus vocal and the stereo version thickens everything into a single mass, while the mid/side version keeps a clear lead sitting in front of a wall.
Character choices
Different saturation curves suit the two halves.
On the mid, tube-style even harmonics or a gentle console-style stage. Both add weight low down and stay out of the sibilant range.
On the side, you can be less careful. Tape-style saturation with its high-frequency roll-off is useful when the doubles are bright and you want density without more top end. Harder clipping suits a wall of stacked harmonies that’s meant to sound like an effect rather than like people singing.
Watch the de-esser
Saturation generates new high-frequency content, which means it can create sibilance that wasn’t in the original take.
Order matters. A de-esser before the saturator sees the original signal and misses harmonics the saturator adds afterwards. A de-esser after it catches everything, at the cost of also pulling down some of the brightness you just added.
Both are defensible. The common approach is to de-ess the lead vocal before the bus, so the raw sibilance is handled at source, then keep the saturation on the mid light enough that it doesn’t reintroduce much. If you’re driving the mid hard for effect, a second de-esser after it stops being optional.
Sibilance is worth being fussy about because it’s the failure that survives every playback system. A slightly dull vocal sounds fine on a phone. A harsh one is unlistenable on everything.
Level matching, and why it’s not optional
Saturation raises perceived loudness even when peak level is unchanged, because it fills in the gaps between peaks and adds harmonics the ear reads as presence.
Turn on a saturator without compensating and it will sound better. It sounds better in the same way turning anything up sounds better, and you’ll keep driving it until the vocal is a mess.
Match the output to the bypassed level and compare again. Some of what you liked will still be there, and some of it was volume. The part that’s still there is the part worth keeping.
A starting point
- Solo the vocal bus and split your thinking: lead in the mid, everything else in the side.
- Drive the side until the doubles feel dense, then stop before they smear.
- Add a small amount on the mid for body. Less than you think.
- Match levels and A/B. Then check the lead against the full mix, since a vocal that sounds thin soloed is often correct in context.