Drop in a podcast episode, interview or voice-over and get a cleaner, louder, more consistent voice in one click: noise reduction, rumble cut, de-esser, presence EQ, compression and loudness normalization to −16 LUFS. Compare before and after, then download WAV or MP3.
Last updated:
55%
50%
+3 dB
Peak ceiling −1 dBFS
Your before/after report appears here.
Load a recording or the sample voice. The chain runs high-pass → noise reduction → de-esser → EQ → compressor → loudness normalize → peak limiter.
–
0:00 / 0:00
Measurement
Before
After
Integrated loudness
–
–
Sample peak
–
–
Noise floor (quietest 10%)
–
–
Loudness range (approx.)
–
–
Advertisement
Advertisement
How to use the Podcast Audio Enhancer
Add your recording by dropping it on the box or clicking to choose (or try the noisy sample voice).
Pick a preset: Balanced voice suits most podcasts; use Noisy room for fans and hum, Phone or video call for thin remote audio.
Fine-tune any stage: noise reduction strength, rumble cut frequency, de-esser amount, presence boost, compression and loudness target.
Listen with Play and switch between After and Before to compare; check the report table for loudness and noise floor.
Download the enhanced audio as MP3 or WAV.
The voice processing chain, step by step
Stage
Default setting
What it fixes
High-pass filter
80 Hz, 24 dB/octave
Desk bumps, traffic rumble, mic stand thumps and plosive pops below the lowest voice fundamentals (about 85 Hz for deep male voices)
Noise reduction
55%
Steady hiss, computer fans, air conditioning and hum, using a noise fingerprint from the quietest 10% of the file
De-esser
50%, above 5.5 kHz
Piercing S, SH and T sounds that become harsh after EQ and compression
Mud cut
−2.5 dB at 300 Hz
Boxy, "in a closet" sound from small rooms and close-up dynamic mics
Presence boost
+3 dB at 3.5 kHz
Dull or muffled speech; improves intelligibility on phone speakers and earbuds
Compressor
3:1, −24 dBFS threshold, 5 ms attack, 150 ms release
Uneven levels between loud and quiet words or between hosts
Loudness + limiter
−16 LUFS, −1 dBFS ceiling
Episodes that are too quiet or louder than the rest of the listener's queue
Before compressing, the tool sets the voice to about −20 LUFS so the compressor thresholds work the same for quiet and loud recordings. Loudness is measured with the ITU-R BS.1770-4 method: K-weighting filter, 400 ms blocks with 75% overlap, an absolute gate at −70 LUFS and a relative gate 10 LU below the ungated level.
Reading the before/after report
Integrated loudness should land on your target (for example −16.0 LUFS). A very low starting value, such as −30 LUFS, is normal for raw microphone recordings.
Sample peak should be at or below −1 dBFS so MP3 or AAC encoding does not clip.
Noise floor is the level of the quietest 10% of 50 ms windows, usually the pauses between sentences. Below about −60 dBFS sounds clean; above −45 dBFS the hiss is clearly audible on headphones. Because the loudness stage raises everything, compare the gap between voice and noise rather than the raw number.
Loudness range (an approximation of EBU Tech 3342 LRA) shows how much the level varies. Spoken-word podcasts are comfortable around 4–8 LU; the compressor brings it down.
Tips for better podcast audio
Record each speaker on a separate track if you can, enhance each one, then mix them; the noise profile and compression fit each voice better.
Leave a second or two of room tone at the start: it gives the noise reduction a clean fingerprint.
Keep peaks around −12 to −6 dBFS while recording; clipping cannot be repaired afterwards.
If sibilance is still sharp, raise the de-esser rather than lowering the presence boost, so clarity stays intact.
Frequently asked questions
What loudness should a podcast be?
The widely used recommendation (Apple Podcasts, AES TD1004) is about −16 LUFS integrated for stereo files and −19 LUFS for mono files, with true peaks no higher than −1 dBTP. Spotify normalizes playback to around −14 LUFS. Choose −16 LUFS unless your host or network asks for something else.
What does each stage of the chain do?
The high-pass filter removes rumble and handling noise below the voice. Noise reduction lowers steady hiss, hum and fan noise. The de-esser turns down harsh S and SH sounds. The presence EQ adds clarity around 3.5 kHz and the mud cut thins out boxy 300 Hz resonance. The compressor evens out loud and quiet words, and loudness normalization sets the final level, with a limiter keeping peaks below −1 dBFS.
How does the noise reduction work?
It analyzes the recording in 46 ms slices, takes the quietest 10% of slices as the noise fingerprint, and then turns down each frequency bin in proportion to how much of it is noise (spectral subtraction with smoothing). It works best on steady noise such as hiss, air conditioning, fans and electrical hum, not on sudden sounds like dogs or traffic.
My voice sounds watery or robotic. What should I change?
That is the typical artifact of too much noise reduction. Lower the strength to 25–40% or try the Light touch preset. Recording closer to the microphone in a quieter room always beats heavy processing afterwards.
Should I export mono or stereo?
A single voice is usually best as mono: half the file size at the same bitrate and identical sound on all devices. If you choose Mix to mono, use the −19 LUFS target so it plays back as loud as a −16 LUFS stereo file.
Is my audio uploaded anywhere?
No. The file is decoded and processed by JavaScript in your browser tab. Nothing is sent to a server, which also means long episodes depend on your device's memory and speed.
Why does MP3 export not work?
The MP3 encoder is a small library that loads the first time you export to MP3. If it cannot be downloaded (offline or behind a strict firewall), choose WAV instead.
Are all the tools on FreeToolCabin tested for quality and accuracy?
Yes! Every tool on our site undergoes automated Q/A testing in real browser environments to ensure high precision, correct output, and zero errors. If you happen to spot a bug or have a suggestion, please contact us at [email protected] and our team will review it right away.