trying to draw out spoken words behind loud static noise.

hi so my situation is a little bit …you could say “different” then most. lets say i placed my samsung galaxy s10 android phone in a room on the bed and recorded sound threw that cellphones microphone. so it picked up everything and anything . mostly sounds like a bunch of static noise and someone talking in the background and or making grunting noises. but you cant deciphere it to much. how can i clear this up? or take down the static noise so the voices in the back can be clear and heard? please help ive tried noise reduction . not quite sure if i had done it wrong … im new to audacity. if anyone has any tips or video recommendations please throw them my way.also i am working with a windows 10 computer that i have audascity downloaded on to edit my audio file and it is audacity 2.3.0

Noise Reduction and other rescue tools only work if you can already understand the words. We can’t turn trash into clear speech.


its not trash . you can hear what thecoversation is about. ive got so many dang recorded files its ridiculous. but how can i clear up the static noise from the a/c running in the back and ambient noise. so its clear from it. i watched a video and the guy zoomed in on a piece of the file he had selected untill it was dotts . then he moved the dots up or something and it made them louder and down made it quiter or something…???

you can hear what thecoversation is about.

What’s the goal? Are we trying to make a podcast from the voices?

You can try Effect > Noise Reduction.

Drag-select some noise-only. No voices. Effect > Noise Reduction > Profile.

Then select the whole file with the button on the left and Affect > Noise Reduction: 12, 6, 6 > OK.

See how it sounds. If you reduce much higher than 12, the voices will start to sound hollow and weird.

You can try Noise Gate.

That’s a plugin you need to install. I need to look that one up.


I’ve never used Noise Gate, so this is as much as I know.

Noise Gate’s job is to reduce sound below a certain volume. You set it so the voices “poke up” from the background noise. It’s hard to use and you don’t get perfect speech at the end. Just maybe a little better.


Beware of Rorschach audio, (like the inkblots, but with sound).
The stimulus is random but people perceive meaning in it: people hear words where there is only noise.