We would like to make quite similar frequencies better decernible for further AI-analysis. Because of that, we would like to stretch the audio in the frequency domain. This means that instead of having two frequency bands of 30Hz and 40Hz, is it possible to “stretch” them to have maybe 28Hz and 45Hz?

Thanks a lot and if that wasn’t clear enough, please don’t hesitate to ask! 🙂

We tried cleaning the data with ffmpeg, but it appears that for our task we need a clearer representation of the frequencies in the data