Recording a Voice You Won't Have to Rescue

Every hour you spend fixing a bad recording could have been ten minutes spent setting up a good one. Noise reduction, EQ and compression can polish a decent voice track. They cannot turn a voice recorded in a tiled bathroom from six feet away into something people want to hear. So the work starts before you press record.

One note on tools. SoundStudio is an editor: it opens files, it does not record them. Record with whatever you have (the voice memo app on your phone, the software that came with your USB interface, Audacity, a handheld recorder), save a WAV, then bring it into SoundStudio with Import or by dragging the file onto the page.

Dynamic or condenser

Most mics you will meet are either dynamic or condenser.

A dynamic mic is the kind you see on a stage or in a radio studio. It is not very sensitive, which sounds like a flaw and is usually a gift. It hears the voice six inches away and mostly ignores the refrigerator, the street and the echo off the walls. Plenty of USB podcast mics are dynamics.

A condenser mic is more sensitive. It catches breath, detail and air around a voice, which is lovely in a treated studio. In an ordinary bedroom it also catches the furnace, the dog two rooms away and the slap of your voice off the closet door. Most need 48 V phantom power from your interface, and many USB "podcast" mics are condensers too.

Distance and angle

Audient's podcast guide puts it plainly: with a dynamic, talk into the mic from about 3 to 5 inches; with a condenser, back off to roughly 5 to 8 inches. A fist with the thumb out is about 5 inches, which makes a handy ruler.

Distance sets the ratio of voice to room. Halve the distance and your voice gets much louder at the mic while the room stays the same, so the room drops back. It also brings proximity effect: directional mics add bass as you get closer. A little gives that warm late-night radio sound. Too much makes you sound like you are talking from inside a barrel. If your voice gets boomy, back off an inch.

Pick a distance and keep it. Every time you lean back to read your notes, your voice gets thinner and quieter and the room comes up. Hold the script at eye level next to the mic so your head stays put.

Plosives

Say "Peter picked a pepper" with your hand in front of your mouth. Those puffs of air are plosives, and when they hit a mic capsule they make a low thump. Two habits fix most of them:

• Talk slightly past the mic. Angle it about 30 degrees off to the side of your mouth so the air blast goes by instead of into it.
• Use a pop filter. A mesh or foam screen a couple of inches in front of the mic breaks up the blast. A cheap one works. So does a thin sock over a wire hanger, if you don't mind looking like a crafts project.

Treat the room with what you own

Hard surfaces bounce sound. Your mic hears the voice, then the same voice a few milliseconds later off the wall, the desk and the window. That is what makes a home recording sound "roomy" or hollow. Soft things soak it up.

• Record in the room with the most soft stuff: carpet, curtains, a bed, a couch, bookshelves. A walk-in closet full of clothes is a classic home vocal booth for a reason.
• Hang heavy moving blankets or comforters behind you and to the sides. The wall behind the mic matters less than the wall behind your head, which reflects straight back into the front of the mic.
• Throw a blanket over the desk if it is a bare hard surface right under the mic.
• Kill the noise makers: HVAC, fans, the fridge if it is close, the phone's notifications. Put the laptop on the far side of the mic, or move it to the floor.

Egg-crate foam stuck to the wall does very little at voice frequencies. A thick duvet does much more and costs nothing.

Gain staging

Gain staging means setting the input level so the signal is strong enough to sit well above the noise and never so hot that it clips. The input gain knob is on your interface, or in the software for a USB mic.

Read a paragraph at the volume you will really use, including the excited part. Watch the meter. Aim for normal speech peaking around -12 dBFS, with the loudest laugh or emphasis still landing below -6. If you hit 0 dBFS even once, the file clips there and the top of the wave is cut flat. That damage is permanent. (The SoundStudio theory page explains why at Clipping and distortion, and decibels at #t-db.)

Beginners usually set it too hot because a louder meter feels better. Resist. At 24-bit there is enormous resolution below the peak, so a voice peaking at -12 has lost nothing. You will set the final loudness at the very end of the job, in lesson 7.

48 kHz, 24-bit, WAV

Set your recorder to WAV, 48 kHz sample rate, 24-bit depth. Here is why:

• WAV is uncompressed. MP3 and the AAC files phones like to make throw data away, and every later save throws away more.
• 24-bit gives you that big margin under the peak, so a safe level costs you nothing. Apple's podcast audio spec, as of October 2026, recommends 24-bit for WAV and FLAC masters.
• 48 kHz is the standard rate for video. A voice recorded at 48 kHz drops into a video edit with no conversion. 44.1 kHz is fine for audio-only work too. Just pick one per project and keep every file on it.

The Digital audio section of the theory page explains sample rate and bit depth if you want the why behind the numbers.

Record room tone

Before or after each session, stop talking and record 10 to 20 seconds of the room with nobody making a sound. Do not skip this. That clip of room tone is what noise reduction learns from in lesson 2, and it is what you paste into gaps when you edit in lesson 3.