Why is my audio too quiet?

Why platforms turn your audio down

Instagram, TikTok, YouTube and Spotify all normalise: they measure how loud a clip actually sounds and adjust it toward a common level, so nobody wins by simply being louder. That level is about −14 LUFS.

Which means the loudness war is over and you cannot compete in it. Mastering something to −8 achieves nothing except losing the dynamics — the platform turns it down anyway. What still matters is being consistent, and not being so quiet that when the platform turns you up, it turns the room up with you.

LUFS is not the same as peak

A peak meter measures the tallest sample. LUFS measures how loud something sounds to a person, averaged over time and weighted for how ears actually work. A recording can peak at 0 dB and still sound quiet, which is why “the waveform looks fine” is not a defence.

The mistake almost everybody makes

Recording too far from the microphone, then compensating with gain. That raises your voice and the room by exactly the same amount, so the fan, the traffic and the reflections come up with you. Moving closer is free and works better than any plugin.

What the fix does

A high-pass at 80 Hz to drop rumble, gentle compression to close the gap between your loud and quiet sentences, a measured gain move to the platform target, and a limiter so the encoder cannot clip it. It is measured, adjusted and measured again, so the result lands on the number rather than near it.

Deliberately no EQ. Nothing here changes how your voice sounds — it changes how loud and how even it is. A tone curve tuned for one microphone in one room is a correction for that setup, not an improvement everywhere, and a tool that cannot hear your room has no business guessing at one. If the recording itself sounds wrong, the findings above say so in words you can act on.