Audio equalizer online Boost bass, shape mids, tame treble

An equalizer changes how loud each range of frequencies is inside a file you already have. Lowering the low range thins out a boomy recording; lifting the high range opens up a dull one. It rebalances what is in the audio rather than adding anything to it, which is why it can fix tone but cannot recover a detail the microphone never captured.

That depends on the layout you pick. Three bands gives bass, mid and treble at 320 Hz, 1 kHz and 3.2 kHz. Ten gives the octave centres from 31 Hz to 16 kHz. Twenty adds the half-octave steps between them, from 31 Hz to 18 kHz. Every band moves between −12 and +12 dB.

Three, until three is not enough. Most of what people come here to do — take weight off a boomy recording, lift a voice, add low end — is a broad move, and a broad move is easier to judge on three sliders than on twenty. Reach for ten when one region needs a different amount than its neighbours, and twenty when you are chasing something narrow, like a room resonance sitting between two octave centres.

Raise the 320 Hz band in small steps and listen at the loudest part of the file. Boosting adds level, so if the result starts to sound hard, lower the overall volume with the volume tool and try again.

Pull the 320 Hz band down a little at a time and keep checking the loudest section. Boom in speech and in room recordings usually sits around and just above that centre, so a small cut often clears it without leaving the file thin. If the weight is lower than the band reaches, the high-pass filter on the noise page cuts further down.

Lift the 1 kHz band a little and ease the 320 Hz band down. Most muddiness in speech sits in the low mids, and taking a small amount out there usually does more than boosting the top.

Start with a small cut at 320 Hz and a small lift at 1 kHz, then listen to a full sentence with A/B rather than a single word — tone is easier to judge over a phrase. Leave the treble alone unless the recording sounds dull, since lifting 3.2 kHz also brings up sibilance. If speakers sit at different loudness, even that out with the volume tool afterwards instead of with the EQ.

A desktop mixing EQ lets you place a band at any frequency and set how wide each one is, which is what you need to notch out a single resonance. Here the centres are fixed — at the octave points a graphic equalizer uses — and so is the width. For getting a voice or a track to sit right, twenty fixed bands cover it; for surgical work on one exact frequency, they do not.

No. You choose between three layouts — 3, 10 or 20 bands — and within each one the centres and widths are set. The equalizer offers 3, 10, or 20 fixed bands from −12 to +12 dB; the centre frequencies and the width of each band do not change. For a bass-focused workflow the bass booster page covers that intent separately.

Yes, and there are two ways. The A/B button switches between the equalized signal and the original, and holding the B key does the same for as long as you hold it, which is easier when you want to flip back and forth quickly. Both use the same filters the export uses, so what you hear is what gets rendered.

Each one is a tone curve rather than a fixed set of slider positions, so the same preset fills three sliders or twenty. Ten are included: voice clarity, podcast, bass focus, full bass, bright, warm, soft, club and rock, plus flat to start over. They are starting points — pick the nearest one and adjust while listening.

Yes. Three bands fit a phone screen as they are, and the wider layouts scroll sideways so each slider keeps a thumb-sized target instead of being squeezed. The practical limit is memory rather than the interface: a long file that opens on a desktop can fail to decode on an older phone.

The filtering itself is applied to decoded samples. Export is where quality can shift: saving back to MP3 runs another encoding pass, so choose 16-bit PCM WAV or FLAC when you want to avoid it.

Common audio and video containers are accepted: the editor tries native decoding first and falls back to a local FFmpeg build to pull the first compatible audio track. Exports are MP3 from 64 to 320 kbps, 16-bit PCM WAV, M4A/AAC at 192 kbps, and FLAC.

No account, no email and no sign-up, and there is no cap on how many files you run. Size is limited by what the browser tab can hold rather than by a rule we set, so the ceiling moves with the device.

The audio is decoded and equalized locally in this browser tab and is not sent to the application server to perform the edit. The page itself still loads site resources and consented analytics.