A: Gaussian noise
B: Signal Tones
C: Gaussian noise with ‘tone holes’
How can i create holes in gaussian noise?
I have a track with gaussian noise and a track with a tone signal, i want to take the 'negative" of the tone track. Picture C describes how i imagine it would look.
-
Make two identical copies of gausian noise track
-
Set one phase inverted and combine both tracks output in a bus/another track, they should cancel each other into silence.
-
Send the tone track (preferably inverted) into the bus aswell so now it has 3 inputs,
-
Add in another tone track that is phase inverted/opposite from step 3 and mixes together with step 3s output
Now, where the fun comes in… what happens when you add fx to different parts of the signal chain so these cancellations become imperfectly dynamic and evolving. What happens when the signals cancelling have compressors with different attack and release settings? What happens when you throw a transient designer like elysia nvelope on different stages of this signal chain?
This kind of effect isn’t too unlike a basic guitar pedal comb filter just a lot more unhinged.
i tryed as you said but the result was just silence, what am i doing wrong?
Yeah, you need a gate when adding in the inversed/phase shifted tone so that it only sends signal when their is signal to cancel, but that is a frequency dependent question too…
guess it cant be done
That’s silly. We just haven’t figured out how to get it to work for you.
i feel maybe i should just try to convert the audio into some fft file and then see if i can somehow invert the values in that file with a script
That’s an interesting thought. I’m useless when it comes to scripting but it’s interesting to me.
I’ve got some free time over the next couple hours, I’ll see what I can come up with.
Out of curiosity, what kinda project you working on?
The project is to invert the MFSK-64 protocol, its used to send text over Ham radio (in this case i generate the tones with Fldigi). I want to make a fusion between data and music. But the tones itself would be rather annoying to listen to, so my idea was to send “sound holes” where the tones would be, and then to decode holes instead of tones. This would allow me to use any music i want, and just encode the data into the holes, a song missing a few frequency’s here and there is still listnable.
-
My first instinct is phase inversion. I feel like we can do this with sidechaining again though.
A couple filters to isolate the frequencies you want to remove, and then a gate? Idk I might have to fiddle around in Reaper later and see what I can come up with. It’s been like 15-20 years since I spent much time in front of a DAW. Imma think on this.
I am also trying with Reaper right now, sadly i didnt have much luck with phase inversion.
Maybe i can track the tones somehow and then shift a notch compressor or something like that over the noise. So if for example there is a 300hz tone in track B, it will move the notch compressor to 300hz in track A. No idea how to track individual tones sady
You could maybe EQ/filter down to the tone frequencies and then pass it to a gate… a notch filter (or a couple) should work.
I like this puzzle
