How to boost bass without distortion (or turning it to mud)

Two different things get called "distortion" when you boost bass, and they have opposite fixes. If it crackles, spits, or sounds like the audio is tearing, you're clipping, and the fix is less overall level, not less bass. If it sounds thick, boomy and swallowed, that isn't distortion at all. That's too much energy around 200-300 Hz, and the fix is a cut at 230 Hz while you keep the 60 Hz boost. Get the diagnosis right and the rest takes about two minutes.

What boosting bass does to the signal

A bass boost is a filter that adds gain below a corner frequency. The bass booster on this site is a low shelf at 120 Hz with a range of 0 to +12 dB, and the presets map straight to decibels: Light is +3 dB, Medium +6 dB, High +9 dB, Max +12 dB.

Here's the part most people skip. Gain is gain. Adding +6 dB below 120 Hz doubles the amplitude of everything down there (doubling amplitude is +6.02 dB exactly), and in most music the kick and the bass are already the loudest thing in the file. If the track was mastered close to full scale, which nearly every commercial release is, there's no room left above it. Samples that would have gone past the maximum get flattened to the maximum instead. That flat spot is clipping, and it creates harmonics that were never in the recording. Your ear hears crunch, crackle, or a fizzy edge on every kick drum.

So the honest version of "why does bass boost sound bad" is usually this: nothing is wrong with your bass setting, you just ran out of headroom.

Fix one: clipping. Turn the master down first

The counterintuitive move is to make the whole file quieter before you boost.

1. Open the volume booster and drop the master volume to 50 %. That is exactly −6.02 dB. If you only need a little room, 71 % is about −3 dB.
2. Now apply your bass shelf. A +6 dB shelf on a signal you already pulled down 6 dB puts the bass back roughly where it started in absolute terms, but everything above 120 Hz is now 6 dB quieter, so the balance has shifted toward bass with no new clipping.
3. Watch the live clip warning. It names the percentage of the file that would run past full scale at your current setting. You want that at 0 %.
4. Leave the soft limiter on. It's a tanh-knee waveshaper that stays transparent below 0.7 of full scale (about −3 dBFS) and bends asymptotically toward the ceiling above that, so it catches the odd stray peak without touching the rest.

The limiter is a safety net, not a licence. It's a limiter, not a compressor with attack and release, and if you drive 10 dB of bass into it, it will hold the level but the low end goes dull and squashed, because the peaks it's rounding off are the bass. Use it to catch a decibel or two of overshoot. If the clip warning says 8 % of your file is over, the answer is still to turn down.

If you'd rather have the number than guess, hit Smart Boost. It scans every sample and reports sample peak in dBFS, RMS in dBFS, integrated loudness in LUFS (a real ITU-R BS.1770-4 measurement, K-weighted, 400 ms blocks with 75 % overlap and both gates applied), and headroom in dB. Then it tells you the exact percentage that produces zero clipped samples with 1 dB of margin left, and one click applies it. It reports; it doesn't secretly re-master. If you're aiming at −14 LUFS for Spotify or YouTube, or −16 LUFS for Apple Podcasts, you read the measured number and move the slider yourself. Read the headroom figure before you touch the bass and you'll know immediately how much the file can absorb.

Fix two: mud. Cut 230 Hz, don't cut 60 Hz

Different symptom entirely. Nothing is crackling. The track just sounds heavy and closed-in, like it's playing through a blanket. Vocals lose definition. You turn the bass down and now it's thin and still muddy.

That's a balance problem in the low mids. Roughly 180 to 350 Hz is where the body of guitars, the low end of male vocals, snare weight and the upper harmonics of the bass guitar all pile up. A low shelf at 120 Hz doesn't stop dead at its corner, so a big setting is still adding a few dB up around 200 Hz. Push it far enough and you've fattened the exact band that was already crowded.

The fix uses the five-band equalizer, whose bands sit at 60, 230, 910, 3000 and 14000 Hz, each a peaking filter with Q of 1.0 and a range of ±12 dB:

• 60 Hz: leave it up, or push it further. This is where the sub and the kick fundamental live. It's almost never the source of mud.
• 230 Hz: pull it down 3 to 5 dB. This one cut does more for perceived clarity than any boost you can make. The bass gets tighter and reads as louder even though you removed energy.
• 3 kHz: nudge up 2 dB if vocals still feel buried. A little presence there restores the intelligibility the low end was masking.

Worth knowing about the built-in presets: Bass Boost is +10, +5, −2, 0, +1 and Hip-Hop is +8, +4, −4, +2, +3. Both add 4 to 5 dB at 230 Hz. On a lean recording that's exactly right. On something already thick, like a lot of 1970s rock or a phone voice memo recorded close to the mouth, that +4 or +5 is the reason it sounds bad. Load the preset, then drag 230 Hz down to 0 or −3 and listen again.

Fix three: the speaker is the thing distorting

Third case, and no amount of processing will help. Small drivers have a physical excursion limit. A phone speaker is barely more than a centimetre across and can't reproduce 60 Hz at any useful level, so when you feed it a fat low shelf it doesn't play deep bass, it flaps and buzzes. Cheap laptop speakers and small Bluetooth pucks do the same. So do old headphone drivers with a partly detached voice coil, which usually buzz on one side only.

The test is quick: play the same processed file on a different device. If the buzz follows the file, it's in the audio, and it's clipping. If the buzz only happens on one speaker while the file sounds fine on headphones, the file is clean and the hardware is the limit. In that case, roll the bass back to Light (+3 dB) and put the energy somewhere the driver can actually use, like a +2 dB nudge at 910 Hz to bring up the body of the mix. You cannot make a driver move further than it moves.

A five-minute diagnostic you can actually run

1. Load the file and hit Smart Boost before touching anything. Note the headroom in dB. If it says 0.3 dB, this is a loud modern master and you have nothing to work with. Go straight to step 3.
2. Set the bass shelf to Medium (+6 dB) and look at the clip warning. Anything above 0 % means clipping is at least part of your problem.
3. Drop master volume to 50 %. Re-check the clip warning. It should read 0 %. Listen. If the crunch is gone, you were clipping and you're done.
4. Still sounds bad, but clean? That's mud. Use the level-matched A/B compare. It switches between the processed and untouched signal with the processed side automatically trimmed to the original's peak, so both hit your ear at the same loudness and you hear the tone change instead of just "louder". Toggle it a few times. If the processed side sounds thicker and less clear rather than crunchier, cut 230 Hz.
5. Play it on a second device. If the problem only exists on one speaker, stop processing and blame the hardware.

That A/B step is the one people skip and the one that settles the argument. Louder always sounds better for about four seconds, which is exactly long enough to make a bad EQ decision.

Clean bass boost settings that hold up

Starting points, not laws. All assume the soft limiter is on and the clip warning reads 0 %. Where the master sits below 100 %, that's deliberate: it's making room for the EQ boosts stacked above it.

The two rows above 100 % only work on files that measured with real headroom. Run Smart Boost first and let the clip warning veto the number.

Video works the same way, with one thing to know before you start. The video volume tool loads MP4, MOV and WebM, plays them with the boost applied live, and exports the boosted audio as a WAV at 16-bit, 24-bit or 32-bit float. It does not re-encode the video. You get an audio file back, and you mux it onto the picture in whatever editor you already use. Export is WAV only, so there's no MP3 on the way out either.

What this will not fix

If your source is already clipped, you can't un-clip it. Flattened peaks are missing information, and any "declipper" that says otherwise is guessing at the shape of what's gone. Boosting the bass on a clipped recording just makes the existing crunch louder. Nothing here removes noise or separates stems, and none of it changes your device's system volume, because a web page cannot do that. Neither can any app, ours included. If the ceiling you're hitting is the phone's own volume slider at maximum, processing the file louder is the only lever that exists, and it's the one on this site.

Everything runs in your browser on your own machine. Nothing is uploaded, there's no account, and once the page has loaded you can switch Wi-Fi off and it keeps working. For fixing settings away from a desk, the iPhone version of the app has the same five EQ bands and the same preset values, with per-track settings saved in a library, though the limiter and the level-matched A/B are website-only for now.