PLEASE help me!!! Clean up a conversation file!!!

Hello everybody , first of all congratulation for this AMAZING application , i want some help for cleaning up a conversation file!
I recorded a conversation (2 people speaking) , but there is a lot of background noise (ringing phones, doors clapping, people speaking, office noises etc)!
Is there any way to grab only the conversation of the 2 people without the other noises? My problem is that all the other noises are louder than the conversation! Can this app automatic recognize and separate selected voices?
It is a wma file (6MB) , in very good quality, recorded from olympus VN-5500PC DNS!

  • i want to clean up this file and keep it only for personal use!!!

Thanks!!! (and sorry for my English)

<<<Is there any way to grab only the conversation of the 2 people without the other noises?>>>

The short answer is no. If the bad sounds are slightly louder than the conversation, then none of the noise keying tools will work, and if the bad sound changes many times, like a telephone ringing or a car starting up and driving away, then the noise reduction tools will fail. The telephone and car are now performers in your show.

Most of the time you can’t rescue a bad show in Audacity or many other software programs. If you don’t shoot the live performance correctly, you will spend weeks trying to fix it at home – if you can fix it at all. I’m betting you have no show.


The fact is that i can (barely) hear all the conversation but i need to strongly concentrate on that and ignore the other noises!
Damn, i thought the technology was able to recognize certain voices from an audio file … or i need to stop believe what i see in spy movies! :laughing:
So, you are saying that i can’t do anything with Audacity to clean this up?
(maybe you can recommend me another app?)
Thanks for your quick answer!! :slight_smile:

Narrow band equalization may make the speech a little clearer, (but it won’t remove unwanted voices, phones ringing etc), e.g.

Thanks Trebor, i will try that! :wink:

He’s saying that it is limited what you can do in any program. Audacity has most of the tools that are available in other tools, but you are asking how to take the milk out of your Café au lait, and that is tough to do. Trebor’s suggestion is a good starting point. If you could upload a short sample (just a couple of seconds in WAV format) there may be other suggestions.

<<<i need to strongly concentrate on that and ignore the other noises!>>>

I’m perfectly clear what you want to do. Do you want this for surveillance, or theatrical production? If for surveillance, then the posted equalizer filter may help – very slightly.

That business where the hero pulls sharp, clear picture or sound out of the wreck of a surveillance system is pretty much pure Hollywood fantasy. Ask yourself where would the story be if they couldn’t do that? If it’s a critical plot point, as it almost always seems to be, then the story wins over reality. The license plate is four pixels in the video, but somehow they find out not only the license number, but the fifteen character serial number on the state inspection sticker.

Our company makes movie dogs and cats that can speak English. And Scottish, although that’s not as popular.

There was a story in a national magazine recently about trial jurys asking the lawyers and judges why they just didn’t use critical DNA testing to solve the case like they did on “CSI” last night?

Yes, do reduce the show to mono, export as FLAC and post a little bit of it for us to chew on. Do Not use MP3.