What is audio normalization? It is the process of changing a recording's gain so its volume lands on a chosen target level, which makes quiet and loud files comparable. It scales the whole signal by one amount and does not remove noise or repair distortion.
Transcribe a recording
Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.
Open @potok_now_bot in Telegram
Files up to 2 GB · export to TXT, MD, DOCX, PDF, SRT and VTT · speaker labels and timestamps · about 30 languages
Key facts
- Peak normalization sets the loudest sample to a ceiling, while loudness normalization aims at the level a listener perceives.
- Raising gain lifts the noise together with the voice, so the ratio between them stays the same.
- Clipping that happened during recording cannot be undone by normalizing afterwards.
- Most editors and command-line tools can normalize a batch of files in one pass, which suits an archive of mixed recordings.
- Levelling several clips to one target helps when a transcript is built from files recorded on different devices.
What it looks like
// loudness-normalize a quiet interview to a common target
ffmpeg -i quiet_interview.wav -af loudnorm=I=-16:TP=-1.5:LRA=11 level.wav
// peak-normalize: scale so the loudest sample reaches the ceiling
ffmpeg -i clip.wav -af "volume=replaygain=track" clip_level.wav
More
See the full list of languages, the tools and the glossary.
Frequently asked questions
Does normalization improve a transcript?
Only when the original was very quiet. It brings the voice to a workable level, but it cannot add clarity that the recording never had.
Is normalization the same as noise reduction?
No. Normalization applies one gain change to the whole file. Noise reduction tries to take specific unwanted sound out, and it changes the signal in a different way.
What is the difference between peak and loudness targets?
A peak target limits the single highest point. A loudness target measures average perceived volume over time, which suits speech and mixed clips better.
Should I normalize before sending a file to the bot?
Only if the voice is hard to hear on playback. A clear recording at normal volume needs nothing, and over-boosting can add distortion.
Can normalization fix a clipped recording?
No. Once peaks are flattened by clipping, the lost shape is gone. Re-record if the damage affects the words.
Sources
Each source was opened by hand on the date shown. Pages change, so open the source if a detail matters to you.
- Audio normalization (Wikipedia), Wikipedia · checked
- Signal-to-noise ratio (Wikipedia), Wikipedia · checked
Try it on your own recording
Send your voice message, audio or video to @potok_now_bot on Telegram and the text comes back in the chat.
Open @potok_now_bot in Telegram