I have an audio file to clean up, but I'm not quite satisfied with the results and would like to ask some feedback, maybe some of you can get a better result and share your knowledge with me?
It's an old interview and I'll put up a clip of the original and my try on it. It's high frequency noise and I tried to remove it with Audition in spectral view, selecting the freqs between 2,5k to 7k (that's all there is), taking the noise sample and then cranked the noise reduction to a somewhat satisfying result, to still keep the voices natural. I'm wondering, can I clean it better to retain more natural human voice and remove more noise?
Your help is invaluable!
Here are the samples:
+ Reply to Thread
Results 1 to 3 of 3
what about this:
Audacity, filters: "Noise Reduction"(NR 24db, Sensitivity 6db, 12 bands)->Amplify(8 or 10db)->Equalization(only maintaining audio from 100hz to 4khz)
But I think that you used the NR in audition in a wrong way using the spectral selection, the NR is better to do it directly instead of selecting some audio band and only do it in that band.
The file uploades is actually encoded in opus at 32kbps and ziped as the site don't allow to upload opus files.
It might be helpful if you could upload a longer clip, one that has a few seconds of silence (noise only).
Might be able to get a better noise print and result
EDIT the provided file was sufficient, I have an older version of iZotope RX so I gave it a go.
I'm not an expert (hobby only) but the denoise plugin was able to help.
Last edited by davexnet; 16th Apr 2017 at 14:05.