Where to High-Pass a Vocal (and When Not To)
Vocal mixing · 4 min read
For most singers a high-pass filter at 80–120 Hz with a 12 dB/oct slope is right. But the correct number comes from the lowest note in the performance, not from a preset: find that note, put the corner just below its fundamental, and never let the filter touch notes the singer actually sang.
Most vocals
80–120 Hz
12 dB/oct · set it from the lowest sung note
The 60-second answer
Find the lowest sung note and its frequency: E2 is 82 Hz, A2 is 110 Hz, C3 is 131 Hz, E3 is 165 Hz. Put the corner about a third of an octave below it, roughly the note times 0.7. Then sweep up until the vocal thins and back off 20%. Slope 12 dB/oct for a lead, 24 dB/oct only if there is real rumble.
What the corner frequency actually does
A high-pass filter is not a wall. You are already down 3 dB at the corner, and a 12 dB/oct slope is still doing something an octave above it. So a filter set at 120 Hz on a singer whose lowest note is A2 at 110 Hz is not below the fundamental, it is on top of it, taking several dB off the lowest notes in the song. That is why the same setting sounds fine on one singer and thin on another: the note range moved, the filter did not.
Why you high-pass at all
Below about 70 Hz a vocal track holds nothing you want: floor rumble through the stand, air conditioning, traffic, plosive energy and proximity boost. None of it is audible as tone, but all of it eats headroom and triggers your compressor, so the compressor spends its gain reduction on a truck outside instead of on the words. Filter first in the chain and your threshold suddenly means what you set it to.
Setting the filter, step by step
| # | Do this |
|---|---|
| 01 |
Put the filter first, before the compressor Anything after the compressor is too late, because the rumble already caused the gain reduction. First slot, ahead of dynamics and saturation. Slot 1, pre-compressor |
| 02 |
Find the lowest sung note Not the lowest note the singer can reach — the lowest one in this song. E2 is 82 Hz, A2 is 110 Hz, C3 is 131 Hz. Note → Hz |
| 03 |
Set the corner below that fundamental Multiply by 0.7 as a starting point: A2 at 110 Hz gives a corner near 77 Hz. Round to something sensible. Corner = note x 0.7 |
| 04 |
Sweep up until it thins, then back off 20% Raise the corner slowly with the beat playing. The moment the chest disappears, note the number and drop it by a fifth. Thin point x 0.8 |
| 05 |
Choose the slope for the problem 12 dB/oct on a lead, because it is gentler through the fundamental. 24 or 36 dB/oct only when there is real low-frequency noise. 12 dB/oct default |
| 06 |
Different corners per layer Doubles 20–30 Hz above the lead, harmonies 120–150 Hz, adlibs and effect layers 200 Hz and up. Only the lead needs the bottom. Lead lowest, adlibs highest |
The trick: sweep up until it thins, then back off 20%
Nobody can see where a high-pass should go on a spectrum analyser, because the energy that matters is not the loudest energy. Do it by ear in one pass instead. With the beat playing, set the corner at 40 Hz and drag it slowly upwards. For a while nothing happens. Then the vocal loses its chest and steps back from the speakers — on most singers somewhere between 130 and 200 Hz. Note that number and set the filter 20% lower in Hz: thinned at 160 Hz means your filter goes at 128 Hz. You have now removed everything the filter can remove for free, with a margin above the point where it costs you body. Do it once per song and you will watch the number move a lot, which is exactly why the preset was wrong.
Stop guessing the chain.
The Vocal Chain Bible is 88 real vocal chains from Grammy-winning engineers — every plugin, every setting, in order, with free alternatives for each one. 5+ years, 100+ interviews, one PDF.
Get the Vocal Chain Bible — $48 →Mistakes that kill it
- 120 Hz on everything. On a baritone that is several dB off the lowest notes in the song. On a soprano it is doing nothing at all.
- A 48 dB/oct filter under the fundamental. Steep filters ring and smear near the corner. Use gentle slopes where the music is.
- High-passing after the saturation. The saturator already used the rumble to generate harmonics that live above the filter and will not be removed.
- Using the filter to fix plosives. A plosive is a thump with plenty of energy above 100 Hz. Fix it with clip gain and mic angle.
Keep going
FAQ
Where should I high-pass vocals?
80–120 Hz covers most singers, but set it from the lowest sung note: corner at roughly 0.7 times that note's frequency, with a 12 dB/oct slope.
Should I high-pass every vocal track?
Every layer, at different frequencies. The lead sits lowest, doubles and harmonies higher, adlibs and effect layers at 200 Hz and up.
When should I not high-pass a vocal?
When the vocal is the lowest thing in the arrangement, and on very close breathy takes where the proximity lift at 100–150 Hz is the sound you wanted.
Who's behind this
Ivan Meshcheriakov · IvanFromRnD
5+ years ago I moved to LA with zero connections and started building from nothing — 14-hour days, meeting people and interning at studios. It took me 5+ years to meet and interview 100+ Grammy-winning engineers — then I turned it all into the Vocal Chain Bible.
Read my full story →