The vocal sounded dull.
So you did what everybody does.
A high shelf on the vocal,
plus four decibels
from about ten kilohertz up.
Air, they call it.
It opened up.
It sounded expensive
on your monitors.
Then you sent the mix
to the singer.
She played it on earbuds
on the bus
and sent back one line.
Why does every S hurt.
Here is the mechanism.
An S is not a note.
It is air forced
through a gap in the teeth,
a burst of noise
with most of its energy
somewhere between about
five and ten kilohertz,
higher for some voices.
You recorded her
with a bright condenser,
pointed straight at her mouth,
so that jet of air
went right into the capsule.
The S was already
the loudest thing in that range.
Then your shelf
lifted that range again.
The S grew,
the rest of the word did not,
and the reverb
took every S you sent it
and spread it across the room
like spilled sugar.
At a low level on good speakers
you barely noticed.
Up close in earbuds,
that range sits right in your ear
and the hiss jumps out.
The fix is not
to throw the air away.
It is to treat the S
separately from the voice.
A de-esser
listens only to that band
and turns it down
only in the moment an S spikes.
The vowels keep their shine.
Set it by soloing what it removes
and backing off
until you hear only the hiss,
not the word.
Put another one,
or a cut in that band,
on the reverb send,
so the room
does not echo the hiss.
And next time,
angle the microphone
slightly off her mouth,
toward the nose or the chin,
so the jet goes past it.
Then check every mix
on the earbuds
your listeners actually use.
The brightness you added
was meant for the voice.
It went to the loudest thing
that was already there.
– Serguey Asael Shinder
Top comments (0)