Record clean, consistent audio first, then use Audacity to remove background noise, shape the tone with EQ, control volume with compression, and export in the format your project needs. That order matters. Audacity can repair a lot, but it cannot turn a harsh, echo-filled recording into studio audio without artifacts.
TLDR: For a reliable Audacity voice over workflow, record in a quiet room at around -12 dB to -6 dB peak level, apply light noise reduction, use EQ to cut rumble and improve clarity, then add compression and export as WAV or MP3. For example, a podcaster recording a 10-minute script may cut editing time by 30% by using a saved effects chain instead of processing every clip manually. Keep every effect subtle; heavy processing is where most amateur voice-overs start to sound thin and metallic.
Set Up Audacity for Voice Over Recording
Start by installing the latest version of Audacity from the official site. Then connect a decent USB or XLR microphone. Built-in laptop microphones usually sound boxy and noisy, no matter how carefully you edit.
Before recording, set the correct input device from the microphone dropdown. Choose mono for standard voice-over work. Spoken voice does not need stereo unless you are recording a special performance setup.
- Project rate: Use 48,000 Hz for video work and 44,100 Hz for audio-only projects.
- Recording channel: Choose 1 mono recording channel.
- Input level: Aim for peaks between -12 dB and -6 dB.
- Room: Record away from fans, traffic, hard walls, and computer noise.
Honestly, it feels like Audacity hides some basic setup choices in plain sight. Spend two minutes checking the device toolbar before you record. It can save you from re-recording an entire script.
Prepare the Room Before You Press Record
Editing starts before Audacity opens. A quiet, dry room gives you more control later. Hang thick curtains, record near clothes, or place soft materials around the microphone area. You do not need a full studio, but you do need to control echo.
Use headphones while recording. This prevents speaker bleed and helps you hear clicks, hum, or mouth noise. Keep the microphone about 6 to 8 inches from your mouth. Speak slightly off-axis to reduce plosives from letters like P and B.
Record a few seconds of silence at the start. This gives Audacity a clean noise sample for noise reduction later. Do not skip this. That “empty” audio is useful.
Record the Voice Over Properly
Click the red Record button and read at a steady pace. If you make a mistake, pause, clap once, and repeat the sentence. The clap creates a visible spike in the waveform, making mistakes easier to find during editing.
Do not record too loud. Clipped audio sounds cracked and harsh. Once clipping occurs, repair options are limited. If the waveform hits the top or bottom of the track, lower the gain and record again.
After recording, save the project as an Audacity project file. This keeps your session editable. Exported audio files are for delivery, not for preserving your full editing session.
Clean Up Mistakes and Timing
First, edit the performance. Remove false starts, long pauses, coughs, chair noise, and repeated lines. Use the Selection Tool to highlight unwanted audio, then press Delete.
Leave natural pauses. A voice-over that is cut too tightly sounds rushed and artificial. For most narration, short pauses between sentences help listeners absorb the message.
If breaths are distracting, reduce them instead of deleting every one. Select the breath and use Effect > Volume and Compression > Amplify with a negative value, such as -6 dB to -12 dB. Removing all breaths can make speech feel unnatural.
Use Noise Reduction Without Wrecking the Voice
Noise reduction is one of Audacity’s most useful tools, but it is also easy to overdo. Heavy settings create a watery, robotic sound. The catch is that a tiny bit of hiss is often better than damaged speech.
- Select the room tone you recorded before speaking.
- Go to Effect > Noise Removal and Repair > Noise Reduction.
- Click Get Noise Profile.
- Select the full voice track.
- Open Noise Reduction again and apply conservative settings.
A safe starting point is:
- Noise reduction: 6 to 9 dB
- Sensitivity: 3 to 6
- Frequency smoothing: 3 to 6 bands
Preview before applying. If the voice turns thin, reduce the amount. If the background is still distracting, fix the room next time rather than forcing the software to do too much.
Apply EQ for Clarity and Warmth
EQ shapes the tone of the voice. In Audacity, use Effect > EQ and Filters > Filter Curve EQ. The goal is not to make the voice sound dramatic. The goal is to make it clear, balanced, and easy to understand.
Start with a high-pass filter around 70 Hz to 100 Hz. This removes low rumble from air conditioners, desk bumps, and traffic. For deeper voices, stay closer to 70 Hz. For lighter voices, 90 Hz or 100 Hz may work better.
Then make small adjustments:
- Cut 200 Hz to 400 Hz if the voice sounds muddy or boxed in.
- Boost 3 kHz to 5 kHz slightly for speech clarity.
- Reduce 6 kHz to 9 kHz if the voice sounds sharp or sibilant.
- Avoid large boosts. Moves of 1 to 3 dB are often enough.
Use your ears, not just numbers. Every microphone and voice is different.
Use Compression to Control Volume
Compression evens out loud and quiet parts. It helps the voice sit at a steady level, which is useful for podcasts, training videos, audiobooks, and ads.
Open Effect > Volume and Compression > Compressor. A good starting setup is:
- Threshold: -18 dB to -12 dB
- Ratio: 2:1 or 3:1
- Attack: 5 ms to 10 ms
- Release: 100 ms to 250 ms
- Make-up gain: Use only if needed
Compression should sound controlled, not crushed. If every word becomes the same volume, the delivery loses life. After compression, use Normalize to set the final peak level, often around -3 dB for general use.
Optional: De-Essing and Limiting
If “S” sounds are harsh, use a de-esser plugin if available, or reduce the painful range with EQ. Sibilance often sits between 5 kHz and 9 kHz. Be careful. Too much reduction makes speech sound dull or lispy.
A limiter can catch sudden peaks after compression. Use it lightly. Set the ceiling to around -1 dB if you need a safe final level for online platforms.
Export the Final Voice Over
When the edit is finished, listen from start to end using headphones. Check for clicks, repeated lines, clipped words, and sudden volume changes. Expect to waste time here if you rushed the first recording pass.
Then go to File > Export Audio. Choose the format based on use:
- WAV: Best for video editors, clients, archives, and broadcast work.
- MP3: Good for web delivery, previews, and smaller file sizes.
- FLAC: Useful for lossless storage with smaller files than WAV.
For MP3, use 192 kbps or higher for spoken voice. For professional delivery, send WAV, 24-bit, 48 kHz unless the client requests something else.
A Simple Audacity Voice Over Chain
For most clean recordings, this order works well:
- Edit mistakes and remove unwanted sections.
- Apply light noise reduction.
- Use EQ to cut rumble and improve clarity.
- Apply compression for stable volume.
- Normalize or limit final peaks.
- Export in the required format.
Audacity is capable, but it rewards restraint. Record well, make small edits, and save presets once you find settings that fit your voice. The best voice-over does not sound heavily processed. It sounds clean, direct, and easy to trust.
