Hacker News Reader

Comment by smusamashah | Hacker News Reader

Is there a tool that can remove very noisy audio recording of a song using actual song as a reference?

I found a very old audio cassette from my childhood with me and some other kids talking while a song is playing in background. I tried subtracting the song using Audacity but for that to work reference song and recording must align "perfectly" which is very very hard. Not just the timing (which i found can be a problem with cassettes) loudness/frequency distribution must also align perfectly.

Found Smartsubtract https://oxfordwaveresearch.com/products/smartsubtract/ which seems to do exactly the same but it's not available for download.

Is there any (AI even?) tool that might do that? I tried an online AI tool which claimed it can extract voices but it returned back silence. I want to try OpenVino but not sure it will be useful with faint spoken words in a noisy environment with a song.

View on HN

I don't know if there's an available tool to do it, but "assuming these sources are aligned in time, remove this reference B from that recording A" would be quite a nice undergrad problem. You'd do something like cross-correlate A with B, multiply B by the correlation coefficient and then subtract the result from A. In the frequency domain, because that makes things a little easier.

The next question on the problem would be "Give at least three reasons why this doesn't perfectly remove the reference sound," of course.

I would try openvino, it might get you part of the way there.

Other things to do would be to fix any tape warble or flutter, normalize the volume, do simple things like high pass filtering anything below 75hz (as most voices don't make audible volume at those frequencies, especially children's voices.

Then I would get a spectrum analyzer plugin and see if there are any spots that are clearly music vs children speaking and zap them out.

(Audition is pretty good software for this, you might still be able to find the download and mass unlock serial key that Adobe released for version 3.0 somewhere on the internet, of course, this is only for people who bought it as a perpetual license back in the day and need to activate it now that the activation servers have gone offline, so no pirating it!)

I'm not gonna say it will be perfect but you might do well enough to be able to hear what everyone is saying and it not sound too bizarre.

I tried Photosounder which works on FFT of the sounds. I expected it have many powers of photoshop to crop/copy/subtract spectrum which I couldn't do with it. It has option of layers but its not as intuitive as photoshop like tools.

Also looking at frequency spectrum and removing sounds from there I learned that music and spoken both contain a bunch of different frequency. To completely eleminate music, i have to remove all of it which is not easy to see or trace manually.

From what you are saying, given audition is an adobe tool, it should be sufficient.

Try iZotope RX

Thanks for suggestion. Some quick lookup suggests it should be able to do it. Will give this a try.

Reply on news.ycombinator.com