5 Vocal Recording Mistakes That Ruin Your Tracks

Every week, I open sessions from artists and producers all over the world. Some through my studio clients, some through RoastYourMix. And I can tell you with absolute certainty — the vocal recording is where most mixes are already lost before I even load a single plugin.

It doesn't matter how good your beat is. It doesn't matter if you spent three months on sound design or if your 808 rattles the walls. If the vocal sounds like it was recorded inside a shipping container with a USB mic taped to a broom handle, the mix is going to fight me every step of the way. And no amount of EQ, compression, or reverb wizardry will make it sit right.

I've mixed hundreds of songs at this point. I also record vocals and instruments in my own studio — Apollo x6, treated room, the whole setup. So I've seen this from both sides. I know what a properly recorded vocal looks like when it hits my DAW, and I know the sinking feeling when I solo a lead vocal and immediately hear three problems stacked on top of each other.

Here are the five vocal recording mistakes I run into most often — and they're all completely avoidable.

---

This one is so common it's almost universal among home recordings. The singer either eats the microphone or stands so far back they might as well be in the next room.

When the singer is too close — less than about four inches — everything falls apart. The proximity effect kicks in hard. The low end bloats. The vocal sounds boomy, muddy, and weirdly intimate in a way that doesn't sit in a mix. Every breath becomes a hurricane. Every lip smack, every saliva click, every little mouth noise gets amplified to the point where I'm spending twenty minutes just cleaning up artifacts before I can even start mixing.

And the plosives. God, the plosives. Those low-frequency bursts from P's and B's that slam the waveform into a flat wall of distortion. I'll get to pop filters in a minute, but distance is the first line of defense, and when a singer is two inches from the capsule, no pop filter in the world is saving you completely.

When the singer is too far — past twelve inches or so — you get the opposite problem. The vocal sounds thin, distant, and washed out. But worse than that, you're now recording more room than voice. Every reflection off every wall is printing onto that track. And once room sound is baked into a recording, it's baked in forever. I can gate it, I can try to clean it with spectral editing, but the damage is done. The vocal will never have that upfront, present quality that a good mix needs.

For most pop, rock, hip-hop, and R&B vocals, six to eight inches is the sweet spot. That gives you enough proximity for warmth and presence without the low-end buildup and noise problems. I tell every artist I work with the same thing — hold your fist up between your mouth and the mic. That's roughly the right distance. Fist width. Simple.

If the singer is dynamic — moves around, gets louder and softer — you may need to coach them to stay planted. A piece of tape on the floor helps. It sounds ridiculous, but it works. Consistency in distance means consistency in tone, and consistency in tone means I can process the vocal with one chain instead of automating everything bar by bar.

If you're recording a singer who belts, pull the mic back to ten or even twelve inches for the loud sections. Or better yet, have them turn their head slightly off-axis when they hit the big notes. This is old-school vocal recording technique and it works beautifully. The alternative is sending me a track where the verse sounds great and the chorus is a clipped, distorted mess — and I see that more than I'd like to admit.

---

I cannot stress this enough. Your room is the most important piece of gear in your recording chain. Not your microphone. Not your preamp. Not your interface. The room.

I regularly receive vocals recorded on genuinely good microphones — AT4050s, Rode NT1s, sometimes even U87 clones — and they sound terrible. Not because the mic is bad, but because the room is a 10x10 bedroom with bare drywall, a hardwood floor, and a window right behind the singer. The mic is doing exactly what it's supposed to do: capturing everything in front of it. And everything in front of it sounds like a bathroom.

What I hear when I open those sessions is flutter echo, comb filtering, and a boxy, hollow quality in the 300-500 Hz range that no EQ can fix without destroying the vocal's body. I can cut 400 Hz all day long, but then the vocal sounds thin and nasal. The problem isn't frequency balance — it's phase cancellation from reflections arriving at the capsule milliseconds after the direct sound. That's physics. You can't un-physics a recording in post.

You don't need a professional vocal booth. You need absorption at the first reflection points. That means something soft and dense behind the singer, something soft on the wall behind the mic, and ideally something overhead. Here's what actually works on a budget:

• Heavy moving blankets hung on a frame or mic stands behind the singer. Not thin bedsheets — actual dense, quilted blankets. The thicker the better.
• A mattress stood up on its side behind the singer. Ugly? Yes. Effective? Absolutely.
• Bookshelves filled with books act as surprisingly good diffusers.
• A closet full of clothes is legitimately one of the best vocal recording spaces in a home. The clothes act as absorption, the small space reduces reflections, and you get a surprisingly dry, controlled sound.

Reflection filters — those curved things that mount behind the mic — are better than nothing, but they're not a solution. They only treat reflections coming from behind the microphone. They do nothing about the walls to the sides, the ceiling, the floor, or the wall behind the singer. I've gotten plenty of vocals recorded with a reflection filter that still sound roomy because the singer was facing a bare wall six feet away and all those reflections were bouncing straight into the front of the capsule. Treat the room. Don't put a band-aid on the mic.

Also — corners are the worst place to record. Low-frequency buildup in corners is brutal. If you're recording in a bedroom, put the singer in the middle of the room if possible, not backed into a corner.

---

This should be non-negotiable, but I still get sessions without one. I can see it instantly in the waveform — those big, ugly, asymmetrical spikes on every P, B, and sometimes T and K. Once a plosive clips the preamp or the converter, that information is gone. It's distorted. It's a burst of low-frequency energy that doesn't belong there, and depending on how bad it is, even the best de-plosive plugins can only partially recover it.

A pop filter costs fifteen dollars. Fifteen. There is no excuse.

Use a nylon mesh pop filter positioned about two to three inches in front of the microphone capsule. Not touching the mic. Not six inches away. Two to three inches. This gives the plosive energy space to dissipate before it hits the capsule.

If you don't have a pop filter and you need to record right now, here's the trick: angle the microphone slightly off-axis — maybe 15 to 20 degrees — so the singer isn't blasting air directly into the capsule. You lose a tiny bit of high-frequency detail, but you avoid the worst of the plosive damage. It's a compromise, but it's better than ruined takes.

Metal mesh pop filters look nicer but they're generally less effective at stopping plosives than the cheap nylon ones. The nylon mesh breaks up the air blast more effectively. I've tested this myself in my studio. The metal ones let more low-frequency energy through. Use the ugly nylon one. Your mixer will thank you.

And here's something a lot of people don't think about — [sibilance](/fix/sibilance) is related to mic positioning too. If you're getting harsh S sounds, try aiming the capsule at the singer's upper lip or chin rather than directly at their mouth. A slight vertical angle change can reduce sibilance significantly at the source, which is always preferable to heavy de-essing in the mix.

---

I open a session. I solo the vocal. The waveform is either a thick sausage smashed against the top and bottom of the track, or it's a thin little line barely visible against the noise floor. Both are problems. Both are avoidable.

When gain is set too high, you get clipping. Digital clipping. Not the warm, forgiving saturation of an analog console — hard, brittle, ugly distortion that sounds like broken glass. And it doesn't just ruin the loud moments. It means the compressor in the signal chain (if there is one) is being slammed constantly, the converter is running out of headroom, and the entire recording has a strained, harsh quality even in the quieter moments.

When gain is set too low, you're recording so quietly that the signal-to-noise ratio is garbage. When I pull that vocal up to a workable level in the mix, I'm also pulling up the noise floor — the hiss, the hum, the room tone, everything. Now I need a noise gate or a spectral de-noiser just to get a clean track, and those tools always have side effects. They always take something away from the vocal's natural character.

Aim for peaks around -12 to -6 dBFS. That gives you plenty of headroom for loud moments and keeps you well above the noise floor. Have the singer do their loudest part — the chorus belt, the scream, whatever their peak is — and set your gain so that hits around -6 dBFS. Then check the quiet parts and make sure they're still well above the noise.

In Cubase, I can see the input meter on the channel strip in real time. Every DAW has this. Use it. Watch it during rehearsal takes. Adjust before you hit record.

Do not ride the gain during a take. Set it before the performance and leave it alone. If you're adjusting gain mid-take, you're introducing level inconsistencies that will haunt me when I'm trying to compress the vocal evenly. If the singer has massive dynamic range — whispers to screams — it's better to record at a conservative level and accept slightly more noise in the quiet parts than to clip the loud parts. I can deal with a bit of noise. I cannot un-clip a distorted transient.

24-bit recording gives you a massive dynamic range — around 144 dB. There is absolutely no reason to record hot anymore. The old advice of "getting the levels as high as possible" comes from the 16-bit era. We are long past that. Record conservatively. Leave headroom. Your mix engineer is literally begging you.

---

This is the one people never think about until it's too late. The singer is wearing headphones. The backing track is playing. The headphones are loud — because the singer wants to feel the energy of the beat. And that backing track is leaking out of the headphone cups and printing onto the vocal mic.

Now I have a vocal track with a quiet, tinny, phase-y ghost of the instrumental underneath it. When I solo the vocal, I can hear the beat. When I mute the vocal, the beat doesn't fully disappear. And when I process the vocal — compress it, EQ it, add effects — I'm processing that bleed too. It creates a hollow, comb-filtered mess that makes the vocal sound like it was recorded underwater.

• Use closed-back headphones. Open-backs and semi-open-backs leak badly. Sony MDR-7506, Audio-Technica ATH-M50x, Beyerdynamic DT 770 — any of these are fine. Just make sure they're closed-back.
• Keep the headphone volume as low as the singer can tolerate. I know they want it loud. Push back on this. Explain why.
• Use only one ear. A lot of singers prefer one cup on and one cup off so they can hear themselves naturally. This is actually a great technique for pitch and performance, but make sure the off-ear cup isn't pointing at the mic. Turn it away.
• Check for bleed before committing to a full session. Record ten seconds of silence with the headphones on and the track playing. Solo that recording. If you can hear the beat, you have a bleed problem. Fix it before you record a single real take.

Bleed is permanent. I cannot remove it. Phase cancellation tricks sometimes help with very specific, simple bleed patterns, but in practice, if there's audible headphone bleed on a vocal, the best option is usually to re-record. I've had sessions where the vocal performance was genuinely incredible — emotional, nailed every note — but the bleed was so bad that the mix suffered no matter what I did. That's a heartbreaking conversation to have with an artist. Avoid it by checking before you record.

---

Recording mistakes don't stop at the technical side. There are two things that can make or break a vocal session even when the recording quality is perfect.

First: comping. If you're sending me a session for mixing, send me the comped vocal — your best take assembled from multiple passes. But also send me the alternate takes on muted tracks. I can't tell you how many times an artist has sent me a single comped vocal where one phrase sounds off, one word is pitchy, or one breath is awkward — and I have no alternates to pull from. Now I'm stuck. I either live with it, or I ask for a re-record that delays the entire project.

Label your takes. "Lead Vocal - Comp" and "Lead Vocal - Alt 1, Alt 2, Alt 3." It takes thirty seconds and it can save days.

Second: communication. Tell me what you want the vocal to feel like. Don't just send stems and disappear. A one-sentence note — "I want the vocal dry and upfront" or "heavy reverb, dreamy vibe" or "reference this song for the vocal sound" — saves me from guessing and potentially going in a direction you hate.

I've mixed tracks where I spent hours crafting a lush, reverb-heavy vocal treatment only to get feedback that they wanted it bone dry. That's not a mixing problem — that's a communication problem. Five minutes of notes up front saves hours of revisions later.

Checklist: Before You Send Vocals for Mixing

• Recorded at 24-bit, 44.1 kHz or higher — never 16-bit
• Peaks between -12 and -6 dBFS — no clipping anywhere
• Clean, dry recordings — no reverb or effects printed on the vocal (send those as separate reference tracks if you want)
• Properly labeled tracks — Lead Vocal, Background Vocals, Ad-libs, Doubles, Harmonies
• Alternate takes included on muted tracks
• A note about the vocal vibe you're going for, or a reference track
• No headphone bleed audible on any vocal track
• Breaths and silence between phrases left intact — don't strip silence or gate before sending. Let the mixer decide what stays and what goes
• Tuning preferences stated clearly — do you want natural pitch correction, hard-tuned, or untouched? Don't assume the mixer knows

---

The difference between an average mix and a great one is almost never the plugins or the mixer's skill. It's the raw material. Give your mixing engineer a well-recorded, properly organized vocal session, and the mix practically builds itself. Send a mess, and even the best mixer in the world is just polishing problems.

I built RoastYourMix because I got tired of hearing the same preventable mistakes over and over. If you're not sure whether your recordings are up to standard — whether your vocals are clean, your room is treated well enough, your levels are right — submit your track and get honest, detailed feedback before you waste time and money on a mix that was doomed from the start. No sugarcoating, no generic advice. Just a straight answer about what's working and what needs to be fixed.