Removing background noise to amplify voice recording

I have a recording that is voice recording while in a Diesel truck, and I am trying to hear the voice recording over the engine noise. I have tried to filter by selecting the background noise and removing it, however, it turns the voice in to high-pitch noises.

Is there a plug-in available that identifies and separates the voice speach from the background noise? There is not other music on in the background. If not, then can you recommend someone who can professional analyse various recordings so that the conversations can be heard?

I appologize for being such an amateur but I am trying to learn.

Any advise would be appreciated.


Use the “Low pass filter” at about 5000 Hz, then use the “High Pass filter” at about 300 Hz.
It won’t make a lot of improvement, but that’s about as good as it gets. You’re trying to do the audio equivalent of trying to un-mix a Martini.

Yeah that did not work. It cuts the noise but the voice sounds are too low. do you know of an expert that can make this happen? It is only about 3 minutes and I don’t need super high quality, I just need to understand what is being said. Let me know if you have any other suggestions.

I called a forensic guy and he wants a $1,000 retainer. Not sure 3 minutes of editing is justified. Again, any suggestions would be greatly appreciated!

Can you recommend any voice accentuating software or sollutions?


CSI is fiction. Even with state of the art equipment the results are unlikely to be much better.

We’ll add “Diesel” to the list of things we can’t fix. The tools all work by recognizing sounds and being able to separate them from the voices in the show. If the show is almost 100% trash, then the tools have no other alternative than to remove 100% of the show – including voices.

The tools also don’t do well when the interference is moving. Give us a nice air conditioner that stays the same pitch and rhythm for two hours. Diesel Trucks (and airplanes) produce noise that moves around. There’s no single thing to grab as a reference, and that’s the end of the job.

People think that because they can understand voices in a very complex mix, that software should be able to do that, too. It can’t. Everything is waves and complex tones. There’s nothing natural that separates voices from anything else, there’s no hook.

Many instruments are designed to mimic a human voice. “Vox Humana” stop on a pipe organ and slide trombone solos. You can tell which one is speaking English, but the software is lost.

Take them up on it.


The “show” isn’t in stereo is it? If it is, you may be able to use one of the voice management tools that can pick out a particular direction left to right in the stereo field. Play the work and if you have two bouncing sound meters, see if they’re different over the course of the show. If they bounce in unison, then that tool set will not work.


This is a block from my web site. I think one of these tools claims to be able to do Vocal Isolation:

Voice Removal or Isolation
–Voice Trap
–Extra Boy