I have some noisy audio files, which are voice recordings or lectures. There is no music, but the recordings are noisy.

Sometimes, the voice is not clear. Is there any software that can extract the clear voice out of those audio files? It would be better to have automated processing and batch processing supports.

I'm using Windows 7.


3 Answers 3


Regardless of OS, Audacity can do this just fine. So says Wikipedia:

Noise Removal based on sampling the noise to be removed. Surface noise from records, for example, can be removed with hardly any discernable effect on the music.

See this Audacity Wiki entry for Noise Removal.

As for batch, see the manual page.

  • just a short comment. Sometimes the recording is so bad, that it is almost impossible to extract the voice. And human ear is much more capable to filter out the voice, than any SW. I tried Audacity for some school lessons recordings and in many cases got only terribly sounding distorted voice like from cheap chinese radio in metal barrel - in fact not much better, than the source file.
    – Juhele
    Feb 5, 2012 at 17:29
  • A generic noise removal program manipulating the sound will never ever reach the human brain understanding speech. It's not your ear, it's the brain using the redundancy of the human language, the likely context of the speech and so on. Try to understand random words in a made up language on a noisy track.
    – chx
    Feb 5, 2012 at 18:07
  • @chx does not sound fits in low-high filters? Nov 28, 2020 at 23:03

You might want to try the expertise of https://auphonic.com/

Here you have an example of what they can do:



yes, now the AI-based tools are more popular than old software utilities. You can try as well https://www.lalal.ai - it's created for voice extraction for music, but I already tried it and was impressed by noise reduction it done.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .