Noise reduction in real time

As I have quite loud fan noises and am able to remove them completely with close to non quality loss in my voice with audacity and the process hardly taking any time at all.

I was wondering. Has it been done before or can someone make it to be possible to record the microphone and apply noise reduction in real time to your microphone and stream that audio to virtual cable driver?

Best noise reduction results I have found by using “Noise reduction (db): 10” and “Sensitivity: 5” and “Frequency smoothing (bands): 0” and “Noise: Reduce”
First I get the Noise profile. Then I apply the noise reduction to where I took the noise profile and where I want the noise to be gone. Then I get new noise profile of the reduced sample and repeat that “reducing process” about 6 times.
In the end I get astonishingly good results compared to just using higher “Noise reduction (db)” or changing any other parameters.

I don’t see a way to attach a file so I cannot give my testing file where I just give example of the reduced noise and with out reduced noise.

While testing “Sensitivity” I found that going below 5 gives artifacts that can be removed by adding “Frequency smoothing (bands)” and it removed bass from the audio. And going above 5 started adding more bass to the recording and reducing other sounds.

I also never thought that my very small room ever had any echo before… But now that I was able to near perfectly (My voice quality does suffer compared to the original recording but that is expected and acceptable as I don’t have a proper mic to begin with) get rid of the background sounds. I got frustrated by the echo sound I have in here. But that is getting away of topic.

Anyone know the code to know is this possible to do in real time or not?
As there is quite many millions of people who would love to get this type of noise reduction in real time so it doesn’t need to be done seperatly.

As I tried to hint before (but I know someone will reply by saying stupid things) this IS NOT processor demanding task. Unless your computer is from the year 98.

Sorry, Audacity is an audio file editor. It doesn’t apply effects in real time.

Most [u]DAWs[/u] can do real-time effects and you can probably find a similar VST effect plug-in that you can use. But, you may not want to run a full-DAW.

Of course, the best thing is to do what the pros do and set-up a quiet studio. :wink:

If you’re in Windows, you can get Windows processing to help you in real time. Normally, this processing gets in the way of home recording because it doesn’t like music very much and we tell people how to turn it off, but you may be the poster child for using it. It’s designed to accept speaking voices and reject everything else (such as room or fan noises).

Follow these instructions backwards.


I found a VST plugin that came with all sorts of filters and it works with OBS But I don’t know how I could use that plugin outside of OBS. (I don’t know what the VST even stands for)
I would love to have all the filters from the plugin that I use with OBS to be applied to my microphone across all applications.
That way I can talk with friends over the net with out having my fan noises in the background.