Recording Vocals

The vocal is usually the most important part of a song, and the hardest to fix later. A clean, well-performed take in a quiet, dead-sounding space makes mixing easy. A boomy, roomy or clipped take can't be fully rescued by any plugin. This lesson covers the whole session, from the room to the final comp.

Choose and Treat the Room

The room is part of every vocal you record. Small, bare rooms with hard, parallel walls create short reflections that make a voice sound boxy and hollow. You can't remove that later with reverb or EQ, so stop the reflections before they reach the mic.

Your goal is a dry (or dead) sound: just the voice, with as little room as possible. You'll add the space you want in the mix (see Reverb & Delay).

Cheap ways to get a dry sound

A closet full of clothes: soft clothes absorb reflections very well. Sing facing into the clothes.
Blankets and duvets: hang thick ones behind and to the sides of the singer, on stands or a clothes rack. A duvet over two mic stands makes a decent vocal booth.
Sofas, beds, bookshelves and curtains: a furnished room sounds much better than an empty one.
Reflection filter: a curved panel behind the mic. It helps, but it only blocks reflections from behind the mic. The wall behind the singer matters just as much.
Avoid: bathrooms, kitchens, small empty rooms and the middle of a room with bare walls on every side.

The clap test

Stand where you'll sing and clap once, sharply. A ringing, metallic "zing" means flutter echoes between hard walls. Add soft material until the clap sounds short and dull. If your clap sounds dead, your vocals will too, and that's what you want.

Mic Choice & Placement

A large-diaphragm condenser in cardioid is the usual studio vocal mic. In a noisy or untreated room, a dynamic mic (such as an SM58 or SM7B) can actually sound better because it picks up less of the room. See Microphones for the differences.

Standard vocal setup

Mic: Cardioid condenser, in a shock mount Height: Capsule level with the singer's mouth or nose Pop filter: 5–10 cm in front of the mic Singer: 15–30 cm from the mic (about a hand-span to a ruler's length) Angle: Straight on, or tilted slightly down toward the mouth Behind singer: Blankets, a wardrobe or absorption panels Behind mic: Reflection filter or soft wall

Closer than about 15 cm adds bass (proximity effect) and makes level jumps worse. Further than 30 cm adds more room sound. Mark the singer's spot with tape so the distance stays the same between takes.

Quiet, intimate vocals (R&B, pop ballads): work closer, around 10–15 cm, for warmth and detail.
Loud, belted vocals (rock, gospel): back off to 30 cm or more, or have the singer lean back on loud notes.
Rap: usually close and consistent, 10–20 cm, with a pop filter. Many rappers record standing to keep their breath and energy up.

Set the Input Gain

Input gain is the preamp knob on your interface. It sets how loud the signal is when it reaches the converter. Set it before you record anything, using the loudest part of the song.

Ask the singer to perform the loudest section (usually the chorus) at full performance volume, not a gentle warm-up.
Watch the track meter in your DAW. Raise the gain until the loudest peaks land around -12 to -6 dBFS.
Record at 24-bit. At 24-bit there's so much room below 0 dBFS that leaving 6–12 dB of headroom costs you nothing in noise (see Digital Audio).
If the interface has a clip light, it should never come on. Clipping on the way in is permanent.

Common mistake

Setting gain on a quiet warm-up and then clipping when the singer really goes for it in the chorus. Singers get louder as they warm up and as the energy rises. Leave extra headroom, and check again after a couple of takes.

Monitoring & Latency

The singer needs to hear the track and their own voice while recording. Always use closed-back headphones. Open-back headphones and speakers leak into the mic, and that bleed ends up in your vocal.

Latency

Latency is the delay between singing into the mic and hearing it in the headphones. It comes from the computer's audio buffer. Even a small delay feels distracting, and a large one ruins timing and pitch.

Set the buffer size to 64 or 128 samples while recording. At 48 kHz, 128 samples is about 2.7 ms each way. Total round-trip latency under roughly 10 ms is usually comfortable.
Turn off heavy plugins (big reverbs, linear-phase EQs, lookahead limiters) on the session while tracking, or use your DAW's low-latency mode [Ableton: Reduced Latency When Monitoring; FL Studio: lower buffer in Audio Settings; Logic: Low Latency Mode].
Or use direct monitoring on your interface: the voice goes straight to the headphones with zero delay, before the computer.
Raise the buffer back up (512–1024 samples) for mixing.

A good headphone mix

Give the singer a little reverb in their headphones (on a send, not printed to the recording). It helps them pitch and feel comfortable. If they sing flat, try turning their voice up a little in the cans. If they sing sharp or strain, try turning it down. Some singers like one earcup off to hear their natural voice.

Takes, Comping & Punch-Ins

Recording takes

Record several full passes of each section. Most DAWs can stack these as takes in one lane while looping [Ableton: Take Lanes; FL Studio: Loop recording into Playlist takes; Logic: Take Folders].

Comping

Comping (short for "compiling") means building one perfect vocal from the best parts of several takes. You might take the first line from take 2, the second from take 5 and a single word from take 3.

Record 3–6 full takes of a section.
Listen line by line and mark your favorites. Judge emotion first, then pitch and timing.
Select the best phrase from each take.
Cut at natural gaps or on hard consonants ("t", "k", "p"), and add short crossfades (5–20 ms) so the edits don't click.
Listen to the whole comp in context. Make sure the energy and tone match from line to line.

Punch-in

A punch-in records over just one part of an existing take, such as one wrong word. Set punch-in and punch-out points around the problem, start playback a bar or two earlier (the pre-roll) so the singer can find the groove, and the DAW only records inside the punch range.

Doubles, Harmonies & Ad-Libs

Doubles: the singer performs the same part again, as closely as possible. Layered under the lead (or panned left and right), doubles make a vocal thicker and more polished. Very common in pop and rock choruses.
Harmonies: parts sung at a different pitch, often a third or fifth above or below the lead.
Ad-libs: extra phrases, shouts and runs between lines. Essential in hip-hop and R&B. Record them on their own track after the main vocal is done.
Whisper or octave layers: a whispered double or an octave-down layer adds texture without cluttering the lead.

Record doubles at the same mic distance and gain so they blend. You'll learn how to mix them in Mixing Vocals.

Coaching the Performer

A great performance beats a great mic every time. If you're recording someone else (or yourself), the session's mood matters.

Prepare: lyrics printed or on a tablet, water at room temperature, the session already set up before the singer arrives.
Warm up: 10–15 minutes of humming and scales before serious takes.
Record everything, including the warm-up runs. Some of the best takes happen when nobody's under pressure.
Give specific, positive notes: "Great take. This time, hit 'tonight' a bit harder" works better than "that was flat".
Talk about the story: ask what the line means to them. Emotion fixes more problems than technique.
Take breaks: voices tire. Plan the hardest parts (high choruses) early in the session, before fatigue sets in.

Common Problems & Fixes

Problem	Sounds like	Fix while recording	Fix in the mix
Plosives	Low thumps on "p" and "b"	Pop filter, angle the mic slightly off-axis, move back a few cm	High-pass filter, clip gain down on the thump, or a de-plosive plugin
Sibilance	Harsh, spitty "s", "sh" and "t" sounds, usually around 5–10 kHz	Point the mic slightly off-axis (at the chin or nose), try a darker mic	A de-esser or a narrow dynamic EQ cut
Room sound	Boxy, hollow, distant	More absorption behind the singer, sing closer, try a dynamic mic	Very limited. Room-removal plugins help a bit, but prevention is far better
Boominess	Too much low end, muddy	Back off from the mic, engage the mic's low-cut	High-pass around 80–120 Hz, cut around 200–300 Hz
Clipping	Crackly distortion on loud notes	Lower the input gain and re-record	Usually can't be fixed. Re-record
Headphone bleed	Faint backing track in the vocal	Closed-back headphones, lower headphone volume, no click in the cans if not needed	Hard to remove. Gate or edit gaps between phrases

Sibilance lives in a narrow, high band. Hear what that region sounds like with the demo below.

Try it in your DAW

Set up your mic with a pop filter about 5–10 cm in front and stand 20 cm back. Put a blanket or duvet behind where you'll stand.
Set the project to 24-bit, 48 kHz and the buffer to 128 samples. Create a mono audio track and arm it.
Sing your loudest chorus line and set the gain so peaks reach about -10 dBFS.
Loop an 8-bar section and record at least four takes in a row [Ableton: Take Lanes; FL Studio: Loop record into playlist; Logic: Cycle mode with Take Folders].
Comp the best lines into one take, cutting on consonants with short crossfades.
Punch in one word you're not happy with, using a 2-bar pre-roll.
Record a double of the chorus and an ad-lib track. Pan the double slightly or tuck it 6–10 dB under the lead.

A good starting distance between singer and mic is about cm.

Building one vocal from the best parts of several takes is called .

Harsh "s" sounds are called , and are tamed with a .

1. Where should vocal peaks land when recording at 24-bit?

Right at 0 dBFS
Around -12 to -6 dBFS
Around -40 dBFS
It doesn't matter

That leaves headroom for loud moments, and at 24-bit the noise floor is so low that you lose nothing by recording a bit quieter.

2. What is the best cheap way to reduce room sound at home?

Record in the bathroom
Add reverb in the mix
Sing into a closet full of clothes or surround yourself with thick blankets
Turn the gain up

Soft materials absorb reflections before they reach the mic. Bathrooms and bare rooms are the worst choices.

3. The singer hears an annoying delay on their voice in the headphones. What should you change first?

Use open-back headphones
Raise the sample rate to 192 kHz
Add more reverb
Lower the buffer size or use direct monitoring

The buffer causes latency. A small buffer (64–128 samples) or direct monitoring removes the delay.

4. What is a punch-in?

A compressor setting
Recording over only a short section of an existing take
Turning on phantom power
A type of microphone

You set punch points around the mistake. The DAW plays the take and only records between those points.

5. What's the main purpose of a vocal double?

To make the vocal sound thicker and fuller
To fix plosives
To reduce latency
To replace the lead vocal

A second performance of the same part, layered under or around the lead, thickens the vocal. It's a pop and rock chorus staple.