Home Vocal Recording and Mixing

How to set up a headphone mix for recording vocals in REAPER

By StudioRescue Editorial

Published October 9, 2026

The headphone mix is the part of recording that nobody teaches and everybody gets wrong at least once. A take can be pitchy, flat, over-sung or weirdly timid purely because of what the singer was hearing, and the fix is a thirty second adjustment rather than anything in the vocal chain. Sound On Sound’s cue monitoring guide puts it plainly: the headphone level “can make quite a difference, both to the quality of performance and to the singer’s ability to pitch accurately.”

The reason it goes wrong in a home setup specifically is that there are two completely separate paths your voice can take to your ears, each with its own volume control, and most people do not know which one they are on. Start there.

The ten second test: which path are you on

Put your headphones on, record-arm the vocal track, and talk into the mic. Now move the track’s volume fader up and down in REAPER while you keep talking.

• Your voice gets louder and quieter. You are monitoring through REAPER. The faders in REAPER are your cue mix controls.
• Nothing changes. You are on your interface’s direct monitoring, which sends the mic signal straight to the headphone socket on the box. Focusrite describes this on the Scarlett range as sending the input “straight to the headphone and line outputs on the device,” with “near-zero latency.” REAPER never sees it, so REAPER cannot change it. The control is the interface’s blend or mix knob, or its mixer software.

This single distinction explains most headphone mix confusion. On a REAPER forum thread titled “Can you change monitoring levels, but not actual input levels when recording?”, a user reported moving the faders and hearing no change at all, and the answer was that record monitoring was off and they were hearing the interface directly. The same poster had been cranking preamp gain to hear more of himself and, in his words, had been recording “way too hot” as a result. That is the damage worth avoiding: a monitoring problem solved with the gain knob becomes a recording problem you cannot undo.

Build the cue mix in this order

Order matters more than any specific number, because each step changes what the next one should be.

What the balance should be, and what to change when pitch is off

There is no correct ratio, and anyone who gives you one in decibels is guessing. Sound On Sound’s position is the honest one: the cue mix “should be one the performer wants to hear, not necessarily the ideal balance.” When you are recording yourself you are both people, so you get to experiment.

Where the published guidance is useful is in what to try when the pitch is consistently off in one direction. Sweetwater’s engineering guide suggests that for a singer who is consistently flat you “try turning their vocal down a bit in the headphone mix” so they push harder, and for a singer who is consistently sharp you raise the vocal so they do not have to push. Treat that as an experiment to run rather than a law. It is practitioner guidance, not a measured result, and it will not help if the pitch problem is random rather than in one direction.

One genuine disagreement between good sources is worth knowing about. Sound On Sound says temporary reverb in the cue “usually helps inspire a more confident performance and can help with pitching.” Sweetwater advises starting with little or no vocal effects because reverbs and delays “can make pitch issues harder to detect.” Both are right about different risks: reverb makes most people sing with more confidence, and it also makes pitch errors harder for anyone to hear. A reasonable resolution is to use the smallest amount that makes you perform well, and to check a questionable take with the monitoring reverb bypassed before you keep it. The routing that keeps that reverb out of the recorded file is in how to hear reverb while recording a dry vocal in REAPER.

Keeping the cue mix out of your actual mix

If you monitor through REAPER, the obvious move is to shove the vocal fader up and the beat fader down. It works, and it costs you something: those are the same faders you will mix with later, so you start the mix from a balance you chose for singing rather than for listening.

You have three options, in increasing order of effort.

1. Just note the fader positions and reset them before you mix. Fine for a quick session, easy to forget.
2. Use the routing matrix. The REAPER user guide’s headphone monitoring section describes assigning output aliases, then using the routing matrix to send particular tracks to a second output. It also notes that you can right click a grid cell to get the controls for that send, including volume, which it calls useful “when a musician who wants to particularly hear only some tracks when recording.” That gives you a headphone balance that is independent of your fader positions.
3. Build a monitor bus. Create a bus, untick its Master Send, add a hardware output send to the headphone outputs, and feed it from the vocal and the backing with sends. This needs an interface with more than one output pair, and it is the version worth building if you ever record other people.

If you want to set the balance up without committing anything, right click the record arm button and choose Record disable (input monitoring only). The user guide describes this as proceeding exactly as if you were recording, so you hear what would have been captured without filling the project with junk takes.

The one ear off question

Sliding one earcup off is standard practice, and both of the sources above treat it as normal rather than a hack. Sweetwater suggests sliding the earpiece back while keeping it firmly against the head, and trying both sides, since some singers stay in tune with the right one off and others with the left. The reason it helps is simply that you hear your own voice in the room as well as in the phones.

The cost is bleed. An open earcup pointed at a condenser mic puts the beat and the click into your vocal track. Sound On Sound is direct about this: you need a way to mute the unused phone, “otherwise you’re inviting spill problems.” At home, the practical version is to keep the loose cup pressed against your head rather than dangling, keep the overall level down, and point the open side away from the mic. If it is already baked into a take, see how to remove click track and headphone bleed from a vocal take.

How loud is too loud

You cannot read your headphone level in decibels from a fader number, and no article can tell you what yours is. What is worth knowing is the scale. NIOSH’s recommended exposure limit for occupational noise is 85 A-weighted decibels averaged over an eight hour shift, and for each 3 dBA increase it recommends halving the exposure time. Tracking vocals for an afternoon at a level you would call loud is not a trivial exposure.

Two practical rules follow from that without needing a meter. Set the level once, early, with your own voice, and resist raising it after every take. And if you find yourself turning up because something is hard to hear, change the balance instead, because a balance problem solved with volume only ever gets louder.

What a good headphone mix will not fix

• It will not fix latency. If your voice arrives late, no balance change helps. That is block size and monitoring path, not cue mix.
• It will not fix a badly recorded signal. Monitoring is downstream of the preamp. If the gain was wrong, the file is still wrong.
• It will not make you sing in tune. It removes one common cause of pitch problems. The rest is the performance.
• It will not rescue a take that is already recorded. Everything here is a setup change for the next take.
• It is not a mix. A cue mix that helps you perform will usually sound wrong as a listening mix, and that is expected.

If it is still not working

1. Faders do nothing. You are on interface direct monitoring. Use the interface’s blend knob or its mixer software, or turn Direct Monitor off and enable REAPER’s record monitoring instead.
2. You hear yourself twice. Both paths are live. Turn one off.
3. Your voice is only in one ear. On some smaller interfaces the direct monitor signal is mono by design. Focusrite documents this for the Scarlett Solo and earlier 2i2 generations. Monitor through the software instead if you want it centred.
4. You cannot hear previously recorded parts on the same track. Check the record arm context menu for the option to monitor track media when recording.
5. Everything is loud and nothing is clear. Pull the whole headphone level down, then rebuild with your voice first. Clarity is almost always a balance problem wearing a volume costume.
6. The backing itself is the problem. Heavily modulated or pitch-shifted parts can throw your sense of pitch. Sweetwater suggests leaving them out of the cue mix where you can.