Is there a way to find the pitch for noisy electronic output? [closed]

Question

Is there a method available to help differentiate notes on a wide variety of electronic noise sounds (not-quite white/pink noise, heavy buzzing, fuzz, etc.)?

I've taken on an odd project where I am trying to get the pitch of sounds that are not generally particularly musical to try and make something out of it. They are electronically generated from very old computer hardware (old Atari POKEY, if anybody cares). However, the software/devices I have been trying to use fails to pick up notes on so much of the sounds produced, even though I can clearly hear a "note" in the sound. I can hum along and tune that, but I want to chart them to +/- 2 cent accuracy if at all possible, and I am not pitch perfect myself.

Using tuners have not been very successful. I have tried:

FMIT (Free Music Instrument Tuner) for Windows -- has a few options to control the pick up, but many sounds get nothing or jump all over the place

ReaTune for Reaper DAW -- I cannot get this to pick up anything meaningful

Various Snark instrument tuners that I have -- Holding these up to the speakers does nothing, regardless of the volume of the output

Are there methods to help figure out the pitch/frequency that noise is making?

Here are some samples (in WAV, so they are ~1-3MB) that I sustain for a moment, then start changing the pitch to make it obvious:

"White Noise" - I'd be surprised if much can be found here, but I'm asking anyways:

http://junk.zolaerla.com/Media/Stack/Distort0.wav

http://junk.zolaerla.com/Media/Stack/Distort0-Poly9.wav

Buzzy, almost engine-like sounds:

http://junk.zolaerla.com/Media/Stack/Distort4-64.wav

http://junk.zolaerla.com/Media/Stack/Distort4-15-ch0.wav

http://junk.zolaerla.com/Media/Stack/Distort4-15-ch2.wav

Answer 1

Some sounds do have pitch, some don't. In general, pitched sounds will have strong discrete sine components, following harmonic series, while unpitched ones will have continuous spectrum and nonharmonic relations between frequencies, but there is no clear boundary. Pitch belongs to psychoacoustics, not math or physics.

You can try to look at the frequency spectrum of a sound (e.g., using free audacity), and see if there is any dominant frequency in the sound. However, in the end, it's only your ear which can tell you if a sound does have a pitch and what is it.

Here are some examples of synthesizing sounds with a perceivable pitch by applying frequency filters.

1. Note how the sound gradually changes from unpitched to pitched when the filter strength is increased (3:28–3:59):

2. Note how the tone changes between two series of examples with broader and narrower filter frequency range. If you tried to increase the filter width more, the pitch would become gradually less apparent (0:11–0:31):

3. Note how the sound containing multiple frequencies might be perceived as a single sound, or a chord (2:54–3:26):

You can perform similar experiments with audacity or other free software.

Answer 2

Note: The original question asked for detecting a pitch in a pure noise signal. This is what this answer refers to.

Noise is actually a random waveform. In order to perceive a signal as a specific note (pitch) the waveform has to be periodic, which is not random. Therefore noise cannot have a defined pitch by definition. If a sound has a pitch it is not noise.

You can also go the opposite way and use a synthesizer software like amsyth to produce a noise at a specific pitch. You will hear no differences in sound regardless of note you use to produce the sound. An A will sound the same as a C or a Bb or anything else.