Amplitude Quantisation
Quantisation of a signals amplitude at a given point in time.
We define the [see page 4, dynamic range] as the range of numbers that can be used to represent a signals amplitude.
If we use \(n\) bits with \(2^{n-1}\) possible values we have a dynamic range in dB of: \[ 20 log_{10}(2^{n-1}) \]
We often use a dynamic range of \(16 \mathrm{\;bits} = 65535 \mathrm{\; possible\;values} = 96 \mathrm{\;dB}\)
If our quantisation bits are too [see page 4, small] we run into approximation errors (too high, too low) or clipping issues (when a signal is too high to be quantised so we just use the largest value).
These errors make sound noisy.