The Vocal Throw: How Producers Make One Word Fly
Vocal mixing · 4 min read
A throw is one word pushed into a long delay while the rest of the vocal stays dry. Automate the send up for the length of that word only, then straight back to nothing. Use a 1/4 or 1/8 dotted delay with 30–45% feedback, and throw the last word before the drop, not the word on it.
Throw time
1/8 DOTTED
375 ms at 120 BPM · feedback 30–45%
The 60-second answer
Send the vocal to a delay return. 1/8 dotted (60000 / BPM / 2 x 1.5 = 375 ms at 120 BPM), feedback 30–45%, 100% wet return, high-pass 400 Hz, low-pass 5 kHz. Draw the send automation at -inf everywhere, then up to 0 dB for the syllable you want to throw and back down before the next word starts.
Why the throw works on the ear
You spend the whole verse hearing a dry, upfront vocal, which sets an expectation about the size of the space. When one word suddenly repeats away into a room that was never there, the contrast is the effect — not the delay itself. Run the same delay on every word and there is no contrast left. Two or three throws per song is usually enough.
Why you automate the send and nothing else
Automating the plugin mix knob changes the balance of repeats that are already sounding, so the tail gets chopped when you pull it back. Automating the send only affects what enters the delay. Once the word is in, the repeats decay naturally even though the send is back at -inf. That is why a send throw sounds like a real hall and a mix-knob throw sounds like someone turned a knob.
Making a throw, step by step
| # | Do this |
|---|---|
| 01 |
Set the delay to a long note value 1/4 for a big open moment, 1/8 dotted for something that stays in the pocket. At 120 BPM that is 500 ms or 375 ms. 1/4 = 60000 / BPM · dotted 1/8 = x1.5 |
| 02 |
Feedback between 30 and 45 percent Around 35% gives you four or five audible repeats before it disappears. Above 60% the tail runs into the next section. Feedback 30–45% |
| 03 |
Filter the return so the repeats stay behind High-pass at 400 Hz kills the boxy build-up. Low-pass at 4–5 kHz makes each repeat a little darker than the last, which is what real distance sounds like. HPF 400 Hz · LPF 4–5 kHz |
| 04 |
Draw the send automation, do not ride it live Flat at -inf for the whole track. One rectangle at 0 dB, starting a hair before the word and ending before the next syllable. Send: -inf → 0 dB → -inf |
| 05 |
Pick an open vowel Words ending in ah, oh or ay throw beautifully. Words ending in a hard T or K throw a clicking sound around the room. Target: open vowel endings |
| 06 |
Duck the beat 1 dB under the tail A small dip on the instrumental bus for half a bar makes the throw read as a production decision instead of an accident. Beat: -1 dB for 1/2 bar |
The trick: throw the word before the drop, never the word on it
The instinct is to throw the last word of the pre-chorus, right as the drop lands. It never works, because the drop is the loudest moment in the song and the repeats get buried under the first kick. Move the throw one word earlier. Throw the second-to-last word so the repeats are already decaying through the final beat of the bar, and by the time the drop hits, the tail sits at the level where you can still hear it in the gaps between kick and snare. Same lift, none of the collision. The logic works in reverse at the end of a chorus: throw the last word into the empty first bar of the verse and the transition does the work for you. To check the placement, mute the vocal after the thrown word and listen to the return against the beat. If the repeats vanish under the arrangement, you threw the wrong word.
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
- Throwing every other line. The effect is contrast. Three throws in a song is a record; ten throws is a wash and the ear stops registering them.
- Automating the wet knob. It cuts the tail off mid-decay. Automate the send and the repeats finish on their own.
- No filtering on the return. Full-range repeats stack around 200–400 Hz and make the drop sound smaller, not bigger.
- Feedback set for the solo. 70% feedback sounds huge alone and turns into an unfinished tail running through the next eight bars.
Keep going
FAQ
What delay time should I use for a vocal throw?
1/4 (60000 / BPM ms) for open, dramatic throws and 1/8 dotted (that number divided by 2, times 1.5) when you want it to stay inside the groove.
How much feedback for a throw?
30–45%. That gives four or five audible repeats. Higher values run into the next section and need automation to stop them.
Should the throw delay be in stereo?
Ping-pong works well because the repeats move away from the centre where the dry vocal lives. Just check it in mono before you commit.
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 →