I have some speech recordings but don't want my voice to be spotted. Is possible to alter them in a non-reversible way? Thx.
delete the files
>>56288365
Erratic wow and flutter effects are difficult to reverse, so on top of whatever modifications you do, those are good for scrambling the audio.
Also apply heavy noise reduction, because barely-audible residual noises can be filtered out to determine the type of location you're in, and throw on a steep low-cut filter starting at like 320hz to eliminate anything resembling low frequencies.
You mean a recording where you want to leave other things in, but also want to remove your voice from (like from the background)?
Look up izotope rx. It's really easy to use and it basically works like photoshop.
>>56288487
Nope, I mean keeping my voice (because those recordings are essentially dialogs) but modifying it. If other voices are altered too is just fine.
>>56288487
Can you remove the dudes voice from this audio and leave only the girl
https://www.youtube.com/watch?v=AqnpwixB7P8
I'm sure you can google how
>>56288614
I'll do it for money, though it won't be in Audacity lol
>>56288365
Something I used to do:
1. Speed up the track 2x (And I don't mean changing the pitch I mean just the tempo. I think audacity has and option for that too)
2. Take the absolute value of the PCM data (There is a Nyquist prompt that supposedly can do this with (abs s), but the Nyquist prompt never worked for me, no idea what's wrong)
3. Lower the pitch 2x (both pitch and tempo, again Audacity can do that)
4. Add some bandpass between 300 and 4000 Hz or something.
If you really have something to hide, that can be useless though. People are recognized by their ductus and vocabulary, not their voices. That's pretty much impossible to remove, so use a speech generator instead.
play with the equalizer - the only way to reverse this is to know what curve you used
save at a lower quality
>>56288667
do it for the money shots it will generate
>>56288365
You could put it through voice recognition and text-to-speech software.
I surely had enough options to try, thx everyone!
>>56288365
Do it with different filters in layers, I suppose.
>>56288365
>Is possible to alter them in a non-reversible way?
Yes, overwrite them with random data