Best way to clean up podcast audio: noise, loudness, pops

Why amateur podcasts sound unprofessional

It's rarely the microphone. A $50 USB mic plugged into a quiet room beats a $500 condenser mic in a noisy coffee shop. The difference between amateur and professional podcasts is almost never the hardware — it's the post-production. Specifically: background noise that wasn't removed, inconsistent loudness that forces listeners to constantly adjust volume, and plosives or pops that hurt the ears.

Most podcasters record, upload, and call it done. Professional podcasters spend 10-15 minutes cleaning audio between recording and publishing. That time investment makes the difference between something listeners respect and something they abandon after 30 seconds.

The order matters: noise removal first, then loudness, then compression

This is the critical mistake: doing these steps out of order ruins your measurement. If you normalize loudness before removing noise, the noise floor gets baked into your reference level, and your loudness pass becomes inaccurate. Same with compression — apply it before noise removal and the compressor will enhance the noise you're trying to remove.

Step 1: Remove background noise — Analyze a 1-2 second section of silence (room tone, air conditioning, computer fan hum). Extract that noise profile. Apply noise reduction across the entire track using that profile. This removes the constant background hiss/hum without affecting speech. Takes 2 minutes.

Step 2: Normalize loudness — After noise removal, measure the overall loudness using LUFS (loudness units relative to full scale). Podcasts should aim for -16 LUFS (Spotify standard) or -14 LUFS (YouTube standard). Apply gain to hit that target. This ensures listeners don't have to adjust volume between your episodes and other podcasts. Takes 1 minute.

Step 3: Compress (optional) — Light compression smooths out dynamic range — making quiet parts audible and loud parts not startling. Use a compressor with a ratio of 2:1 or 4:1, attack time under 50ms, and release over 100ms. This is optional for clean recordings but essential if you have variable loudness. Takes 2 minutes.

Noise removal vs. noise gate: why removal is better

A noise gate is a blunt instrument: anything below a volume threshold gets muted. Loud background noise gets muted; quiet speech gets muted. The result is choppy audio with dead silence whenever you pause.

Noise removal is smarter: it analyzes a quiet section to learn the exact noise signature (frequency content, harmonic structure), then removes only that pattern throughout the track. Speech remains clear; background hiss disappears. The result is natural-sounding audio without holes or choppiness.

Use noise removal for ambient room noise, air conditioning, computer fan hum — the constant background drone that's present even when you're speaking. Use a noise gate only if you have brief loud sounds you want to completely eliminate, which is rare.

Loudness targets: -16 LUFS, -3dB peak, and what they mean

LUFS (Loudness Units relative to Full Scale) is the standard measure. It accounts for how human ears perceive loudness, so -16 LUFS sounds the same loudness across all speakers and platforms.

-16 LUFS: Podcast standard. Quiet enough not to startle, loud enough to be clearly heard in a car. Spotify normalizes all content to -14 LUFS, so hitting -16 gives a tiny bit of headroom.

-3dB peak: The loudest instantaneous peak in your audio should be 3 decibels below maximum volume. This prevents clipping (distortion from the signal hitting the recording ceiling) and gives headroom for streaming platforms' loudness normalization.

-1dB peak: Overly conservative. Gives lots of headroom but leaves loudness on the table. Your podcast will sound quieter than other podcasts unless listeners turn up volume.

In practice: normalize to -16 LUFS, check that peaks are under -3dB, done. If you're unsure, use a free loudness meter (like LUFS Meter for Windows or Nuage for Mac).

Fixing plosives and pops without re-recording

Plosives are harsh bursts when you say "P," "B," or "T" sounds. They're recorded as loud spikes that hurt the ears. A pop filter at the microphone prevents most of them. But if you didn't use one, you can fix them in post.

Approach 1: Use a de-esser or EQ to reduce high-frequency content (3-8 kHz) where plosives live. This softens them without sounding obviously edited.

Approach 2: Manually edit: find the plosive spike in a waveform editor, use a pencil tool to reduce the peak amplitude, blend the edges so it doesn't sound cut. Takes 10-20 seconds per plosive.

Approach 3: Re-record just the bad words. Most DAWs let you punch-in a new recording over the old one. Takes 30 seconds per plosive.

For a podcast with heavy plosives: EQ pass first (1-2 minutes), then manual touch-ups on the worst offenders (5 minutes).

The complete audio cleanup workflow

Total time: 10-15 minutes per episode.

1. Import audio into a DAW (Audacity, Adobe Audition, Reaper, or a browser tool like WavyVid).
2. Remove background noise: Select 1-2 seconds of silence, analyze, apply noise reduction to the full track.
3. Normalize loudness: Measure LUFS, apply gain to hit -16 LUFS target.
4. Check peaks: Verify nothing exceeds -3dB. Reduce if it does.
5. Apply light compression (optional): 2:1 ratio, 50ms attack, 100ms+ release.
6. Fix remaining plosives (optional): EQ or manual editing on the worst offenders.
7. Export: WAV for archive, MP3 for upload.

Free tools for podcast audio cleanup

Audacity: Free, open-source, cross-platform. Has noise removal, loudness normalization, EQ, and compression. Clunky UI but powerful.

WavyVid: Browser-based, no download. Removes background noise, normalizes loudness, handles compression. Simpler interface than Audacity. Your audio never leaves your device.

Descript: Free tier includes transcription and basic noise removal. Paid version adds more tools. Good if you also need captions.

Adobe Audition (paid): Industry standard. Overkill for beginners but powerful if you plan to keep podcasting.

The pragmatic microphone upgrade path

If you've cleaned your audio and it still sounds amateurish: the problem might be the microphone or room.

Upgrade path:

• First: Use what you have and clean it well (this post).
• Second: Add a pop filter ($15) and move to a quiet room.
• Third: Upgrade to a better USB microphone ($50-150: Shure MV7, Audio-Technica AT2020USB).
• Fourth: Invest in a proper microphone, preamp, and acoustic treatment ($500+).

Most podcasters plateau at step two or three and sound professional. Step four is for long-term commitment.

One more time: the order is noise, then loudness, then compression

This is worth repeating because it's the most common mistake. Remove background noise first, normalize loudness second, apply compression last. Any other order ruins your measurements and wastes 10 minutes. Get this right and your podcast will sound professional from your first episode.