How-to
Make an MP3, or any audio file, louder
A voice memo recorded far from the microphone, an old MP3 ripped at a low level, or a podcast episode that sounds quiet next to the others: all of them share the same fix, raising the volume without re-recording anything. The volume tool works on MP3, WAV, M4A, AAC, OGG and FLAC and offers two ways to do it: a straightforward dB boost, or loudness normalization for a file whose level wanders.
Step by step
- Open the volume tool and drop your file in (MP3, WAV, M4A, AAC, OGG or FLAC all work). Leave the mode on "gain" for a simple boost.
- Drag the dB slider up (the range runs from -60 to +30 dB). Start around +6 to +9 dB for a quiet recording and listen back before pushing further; too much gain clips the loudest peaks into a harsh crackle instead of a louder sound. Switch to "normalize" instead if the recording's level drifts up and down rather than being uniformly quiet.
- Run it and download the result. The output keeps your original file's format and extension, so an MP3 stays an MP3.
A fixed dB boost, or loudness normalization
Gain mode applies one number, in decibels, to the whole file: +6 dB roughly doubles the perceived loudness everywhere, quiet passages and loud passages alike. That is the right tool for a recording that is uniformly too quiet, like a phone voice memo taken from across the room. Normalize mode runs ffmpeg's `loudnorm` filter instead, which analyzes the file and evens out its overall loudness to a standard target rather than applying a single gain value; it is the better choice when the level itself is inconsistent, for example a podcast where the host sits closer to the microphone in some segments than others.
Why too much gain distorts instead of helping
Every digital audio file has a ceiling: the loudest sample it can represent without clipping. Raising the gain moves every sample up by the same amount, including the peaks that were already near that ceiling. Push the gain far enough and those peaks slam into the ceiling and get flattened, which sounds like a harsh crackle or crunch rather than a cleanly louder recording. A quiet phone recording usually tolerates +6 to +12 dB comfortably; material that already has loud moments (music, a raised voice) has far less headroom before it starts to clip, so it is worth listening to the result before committing to a large boost.
What loudness normalization does and does not guarantee
This tool runs `loudnorm` in a single pass, which analyzes and adjusts the file in one go. Streaming platforms and broadcasters that certify a file against a precise loudness figure, commonly -23 LUFS for broadcast delivery, use a two-pass measurement instead: a first pass measures the file exactly, then a second pass applies the correction based on those measured values. A single pass gets close and is enough to bring a quiet recording up to a comfortable, consistent level next to your other files; it is not a substitute for a certified two-pass loudness measurement if a platform requires one.
The tools used in this guide
- Audio volumeBoost, reduce or normalize the loudness of an audio file in your browser. No upload.
- Audio cutterCut a clip out of an MP3, WAV or M4A right in your browser: fade in/out, ringtone preset. No upload.
- Audio converterConvert audio to MP3, WAV or AAC directly in your browser. No upload.
- Merge audio filesConcatenate multiple audio files (MP3, WAV, M4A, OGG…) into one. Drag to reorder. No upload.
Frequently asked questions
How many dB should I add to make a quiet voice memo audible?
Start around +6 to +9 dB and listen back. A phone recording taken across a room usually needs a boost in that range to sound comfortable; going much higher risks clipping the loudest moments into a harsh crackle. If the file has any louder passages already, back off and try normalize mode instead, since it evens out the level rather than pushing everything up by a fixed amount.
Does normalize make my file meet a broadcast loudness standard?
Not exactly. This tool applies `loudnorm` in a single pass, which brings the file to a comfortable, consistent level but is an approximation. Platforms that certify audio against a precise target such as -23 LUFS use a two-pass measurement, where the file is measured first and corrected afterward based on those exact numbers. Use normalize here to even out a recording for casual listening; use a dedicated two-pass loudness tool if a platform requires certified compliance.