Use the free Audio Normalizer below: your data never leaves your browser — all work happens on your device, and nothing is uploaded to any server.
Why your recordings all sound like different volumes
If you have ever stitched together clips from different days, you know the problem. One voice note is loud and clear, the next is so quiet you have to turn the volume all the way up, and the third suddenly blasts your ears. The recordings are not broken. They were just made at different distances from the microphone, on different days, with different settings. Normalizing is the simplest fix for that mess. It measures the loudest moment in a file and moves the whole file up or down by one amount, so the loudest moment lands exactly where you want it. Everything else moves by the same amount, so the balance inside the recording stays exactly as it was.
That last point matters more than people think. Normalizing is not compression. Compression squeezes loud and quiet parts closer together. Normalizing leaves the gaps between loud and quiet untouched. A whisper stays a whisper compared to a shout; both just get louder or quieter together. That is why it is the safest first step before you share audio. It fixes the overall level without changing the character of the recording, and it never invents detail that was not there.
Peak level, explained in plain language
Every digital audio file has a ceiling, called 0 dBFS. Nothing can go above it. When a sound hits that ceiling, it clips, which means the top of the waveform gets chopped flat and the sound turns harsh. The peak of your file is simply its single loudest sample, measured against that ceiling. If your peak is at minus 12 dBFS, you have 12 decibels of headroom left. Normalizing to minus 1 dBFS raises everything by 11 decibels, so the peak sits just under the ceiling. That is the usual target for a file you plan to share, because it is loud without risking clipping.
There is also RMS level, which is closer to how loud a file feels on average. A recording can have a high peak but a low RMS if it is mostly quiet with one loud clap in it. Our tool shows you both numbers before and after, so you are not guessing. If the RMS is very low even after normalizing, the file itself is mostly quiet, and no amount of normalizing will make it feel like a studio recording. That honesty saves you from chasing a number that cannot fix the real problem.
How to use this normalizer
Upload your file first. WAV, MP3, OGG and M4A all work, because your browser decodes them before anything else happens. Choose a target peak. Minus 1 dBFS is the best default for sharing. Minus 3 or minus 6 give you more safety margin if you plan to edit the file again later, because later edits can push the level up. Press Normalize, wait a second, and read the before and after numbers. Then play the result. If it sounds right, download the WAV. The original file on your computer is never changed, so you can always go back and try a different target.
One practical tip: normalize at the end of your editing, not the beginning. If you normalize first and then cut, paste and add music, the levels drift again. Do all your cutting first, then normalize once as the final step. Podcasters who record interviews in separate files should normalize each voice to the same target before mixing, so no guest sounds distant while another sounds like they are shouting into the microphone.
What normalizing cannot fix
Normalizing cannot repair clipping. If a recording already hit the ceiling while it was made, the flat tops are baked in, and raising or lowering the volume just makes the same distortion louder or quieter. It cannot separate a voice from background noise either. Turning everything up turns the hum of a fan up by exactly the same amount as the voice. And it cannot make a distant, echoey recording sound close and dry. Those problems need different tools, or better yet, a re-recording with the microphone closer to the speaker. Knowing this limit keeps you from wasting an afternoon pushing a slider that was never meant to solve that problem.
Common questions
Will normalizing harm my audio quality? No, as long as you export to WAV and keep the peak below 0 dBFS. The math is a single multiplication applied to every sample, and it is fully reversible in principle. Should I normalize every file I share? For spoken audio, yes, it is good manners, like straightening a photo before you send it. Does the tool upload my file? No. The file is decoded and processed inside your browser, and the only copy that leaves is the WAV you choose to download.
Pingback: Stereo to Mono Converter - Turn Stereo Audio into True Mono WAV Free