Analog Subtract Audio From White Noise

Thanks Phil. But I am thinking that the varying voltage definition would apply exclusively to white noise from a single source.

By adding together a number of sources, say a few dozen back-based transistors, the frequencies in the composite noise move toward "infinity".

So what I am thinking, as a possible first step, is to use multiple noise sources to ensure a broader noise spectrum at any given moment in time.

Then mix this with an inverted copy of the audio track I want to "remove" from the noise.

The audio will then "cancel" any frequency component within the noise that is similar to the original (non-inverted) audio track.

Comments anyone?

Mike Towner

No, it doesn't.

No, they don't.

You invert the audio, it's basically still the same audio. Polarity is irrelevant for audio (although both channels of stereo sound should match).

There are two conventional meanings of the term 'mix'. For audio, it's simple addition. It could be subtraction, which is fundamentally the same thing. For everyone else, 'mixing' means multiplication, also known as modulation. Neither will remove specific frequencies from the noise.

No it won't.

Will you finally tell us what this is for?

Jeroen Belleman

You are one confused puppy.

.... Phil

Rob wrote.

Maybe he can comprehend what he reads, you pompous asshole.

tion - white noise is simply a randomly varying voltage. IOW, the value se en at some future point in time is unpredictable.

number of sine waves at an infinite number of frequencies - the implication being that each one has an infinitely small amplitude.

Mixing them certainly won't achieve that. You'd need to use the audio to co ntrol a set of notch filters that could then remove whatever they find. Wit h say 1000 filters you'd want to do it in software. Audacity implements at least part of the process.

IIUC I think Phil's key point is that while if you look at white noise over time you see all frequencies, at any one moment you don't.

I'm surprised at the confusion around the point of the OP's idea. Is it not to encode audio into what appears to be just noise to enable clandestine c ommunication. It might even work if you sent the signal a huge number of ti mes using different chunks of noise each time, then you can see at the rece iving end what's missing.

I suspect you'd be better off looking at communications options with more n oise than signal. Such was done very effecively in the 1940s.

NT