How to Make a Vocal Chop That Sounds Like a Record
Vocal mixing · 4 min read
Slice the sample on the consonants, not on the grid. Tune each slice to a note in the chord playing underneath — not to the root every time — and lock the formants so a chop pitched up five semitones does not turn into a cartoon.
Formants
LOCK AT 0
Slice on consonants · pitch to the chord, not the root
The 60-second answer
Drop the phrase into a sampler and slice at every hard consonant (T, K, P, S). High-pass each slice at 200–300 Hz, de-ess 3–4 dB, lock the formants, and put a 5–15 ms fade on the start and 20–60 ms on the end of every slice. Then play chord tones, not the root.
Why chops sliced on the grid sound wrong
The grid does not know where a word starts. A sung syllable builds for something like 10–40 ms before it reaches full level, and the consonant in front of it is earlier still. Cut on the bar line and you get half a consonant hanging off one slice and a vowel with no attack starting the next. The pattern then feels slightly late and slightly soft no matter how hard you quantise. Cut at the consonant and every slice opens with a transient, which is what makes a chopped voice behave like a percussion instrument.
Why formants matter more than pitch
A basic pitch shifter moves the whole spectrum, including the resonances that tell your ear how big the singer's mouth and throat are. Shift up five semitones with formants following and you do not hear the same person singing higher, you hear a smaller person. Shift down and you get a monster. Formant locking holds those resonances still while the pitch moves, so one sample covers an octave and stays one human being. No formant control? Keep shifts inside about three semitones.
The chop, step by step
| # | Do this |
|---|---|
| 01 |
Pick a phrase with hard consonants One or two bars, clear diction, little reverb in the source. A wet sample smears as soon as you shorten the slices. 1–2 bars, dry, mono |
| 02 |
Clean it before you chop it High-pass 150 Hz, de-ess 3–4 dB, cut any obvious room resonance. Every problem gets multiplied by the slice count. HPF 150 Hz · de-ess 3–4 dB |
| 03 |
Slice at the consonants Put a marker just before each T, K, P or S, then nudge it back to the nearest zero crossing. Ignore the grid here. Markers on consonants |
| 04 |
Map to keys and lock formants One slice per key, one-shot or gate mode, formant control at 0 so pitch and formant move independently. Formant 0 · one-shot |
| 05 |
Play the chord, not the root For each chord, trigger slices on the root, the third and the fifth. The chops then change colour with the harmony. Root · 3rd · 5th |
| 06 |
Fade every slice 5–15 ms in, 20–60 ms out, or an amp release of 40–80 ms. Clicks are missing fades, not bad slicing. 5–15 ms in · 20–60 ms out |
| 07 |
Print it and treat it as a synth Filter automation, a 0.8–1.2 s plate at 15–25%, a dotted-eighth delay, and 2–4 dB of sidechain from the kick. Print → filter → space |
The trick: chop on the consonant, then play the chord instead of transposing
Almost everyone maps a sample chromatically, plays the same key over and over and shifts it up and down for variation. It sounds like a sample being transposed, because it is. Do this instead. Write out the chord progression under the section, and for every chord trigger three slices: root, third, fifth, plus the seventh if the chord has one. You are no longer transposing a sample, you are voicing a chord with a human voice, and the pattern follows the harmony instead of fighting it every second bar. Stack that on top of consonant slicing, so every chop opens with a T or a K rather than a vowel, and the part reads as a rhythmic instrument made of a voice.
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
- Slicing on the grid. Every slice starts mid-vowel, the pattern feels soft, and quantising harder makes it worse.
- Letting formants follow the pitch. Three semitones up and the singer has changed body size. Lock it or keep the shifts small.
- Chopping a reverb-heavy sample. Each slice carries the room of the one before, so fast patterns turn to soup.
- No fades. Clicks at slice boundaries are the fastest way to make a good idea sound like a first attempt.
Slice on consonants, lock formants, play chord tones, fade everything. Four rules and a chop stops sounding like a preset.
Keep going
FAQ
Where should I cut a vocal to make a chop?
At the consonant that starts the syllable, snapped back to the nearest zero crossing. Not on the bar line by default.
How do I stop vocal chops sounding like chipmunks?
Lock the formants so they stay put while the pitch moves. Without formant control, keep shifts inside about three semitones.
Should vocal chops be mono or stereo?
Chop in mono so the slices stay predictable, then build width afterwards with a stereo delay or a doubled pattern panned as a pair.
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 →