Online Podcast Mastering: Noise, Loudness, and LUFS for Apple and Spotify

Online Podcast Mastering

A podcast is, first and foremost, speech. Listeners forgive music a lot of "spectral creativity," but they won't forgive constant AC hum, loudness jumps between guests, or listener fatigue from harsh sibilance. Podcast mastering comes down to dialogue clarity and stable loudness that meets aggregator requirements.

Common problems

Room tone and echo. Recording in an untreated room adds early reflections and noise. Through headphones, listeners immediately hear this as a "cheap studio."

Uneven speaker levels. One guest sits closer to the mic, another farther away — listeners keep reaching for the volume knob.

Wrong LUFS target. Apple Podcasts aims for around −16 LUFS, while Spotify for podcasts is often cited around −14 LUFS. An episode that's too quiet gets lost in the feed next to competitors.

What Magic Master can do

Plans with the extended processing chain include a denoiser and de-esser, multiband compression evens out levels, EQ removes boxiness, and a final limiter keeps True Peak in the safe zone. You choose the target LUFS based on your platform.

Pre-publish checklist

• Minimize noise during recording, or run it through denoising.
• Level the loudness between the host and guests.
• Check sibilance on "s," "sh," and "ch" sounds.
• Set the target LUFS for your aggregator.
• Export as MP3 320 kbps or AAC, depending on your hosting requirements.

How to get started

1. Export your episode as WAV, MP3, or FLAC.
2. Upload it to the app and choose a preset built for speech (like Podcast, if it's available in the style list).
3. Turn on PRO modules for denoising and de-essing if needed.
4. Download the result and check loudness in the built-in analyzer.

Conclusion

A good podcast is about intelligibility and comfort over the long haul. Online mastering lowers the barrier to entry — you don't need a full DAW to hit release-ready loudness and tone. Basic mastering is free (3 episodes a day); PRO modules like denoising and de-essing run on tokens, where 1 token = 1 mastering. Try it at magicmaster.pro.