Copy link to clipboard
Copied
Hi guys,
so I have a recording from a classroom study where participants work in groups of two. There is a recording device (in this case, an iPad) on every table, but the groups are sitting too closely together so the recording picked up more than one group's conversation.There is one group of girls who talk very quietly and I would like to understand what they are saying, but the male voices in the background are just too loud. Can anybody think of a way to remove the (deeper, louder) male voice in the background?
Google just doesn't know this specific situation. I have tried opening the file in the waveform editor and opening the spectral frequency display with the colors, but I can't figure out which are the noises corresponding to the voice I want to remove...
Any ideas? Thanks so much in advance (a simple "it's impossible" would also help, at least then I know I don't need to keep looking...)
Grüße
Caro
'It's impossible' is in fact the correct answer. NR only gets rid of continuous noise, so is no use for this, and anyway, the tools that do this sort of thing are intended to get rid of quieter sounds. Unfortunately the frequency ranges of human speech aren't actually that different - knocking out some of the lower frequencies won't aid recognition, as most of this involves the higher partials. I'm afraid you're rather stuck with this one...
Copy link to clipboard
Copied
'It's impossible' is in fact the correct answer. NR only gets rid of continuous noise, so is no use for this, and anyway, the tools that do this sort of thing are intended to get rid of quieter sounds. Unfortunately the frequency ranges of human speech aren't actually that different - knocking out some of the lower frequencies won't aid recognition, as most of this involves the higher partials. I'm afraid you're rather stuck with this one...
Copy link to clipboard
Copied
Awesome, thank you for the quick reply, that's very helpful. I'll just repeat the study with a better room layout, lesson learned 😉