Brain Dump

White Noise

Tags
speech-processing

A [see page 12, completely random] stochastic signal.

The value of a signal at time \(t\) gives no information about it at time \(t+n\).

Note: Perfect white noise has a constant see page 13, spectral density. Every frequency of sound is present throughout the speach signal, just at (slightly) different amplitudes.

Gaussian White Noise

Plot the [see page 15, probability] of white noise by amplitude. I.E. amplitudes with a large Y value are more likely than amplitudes with a low Y value. It's a PDF for white noise.

Binary White Noise

White noise with one of 2 possible [see page 17, amplitudes]. This sounds just like Gaussian White Noise because the different spectral graphs of it are [see page 18, similar] to GWN.

Note: we hear frequencies. If the Power Spectral Density is the same, the sound is the same.