A bit more in depth about Bit Depth...


There is no difference in “resolution” between 8/16/24 or "32" bit PCM audio.
In fact, the only difference is in the (theoretical) noise floor.

Consider the following...

Each additional bit is added to the LSB (least significant bit),
and is is equal to 1/2 the value of the next highest bit.
It's like adding more places to the right of the decimal point.

A longer word length only represents smaller numbers more accurately.

This additional accuracy means only one thing - that the theoretical noise floor (which is entirely due to rounding errors) is lower,
but the bits above the noise floor are absolutely identical whether the word width is 8, 16, 24, or 32 bits.


If the audible part of the signal (the part above the noise floor) changed in any way
when the signal was represented as a 32 or 24 or 16 or 8 bit word,
this change in the waveshape would produce harmonics not present in the original signal, (aka "aliases”).

But it doesn't.

If your recording was made using microphones, preamps, softsynths or samples,
the inherent noise floor is already many times higher than the noise floor of dithered 16 bit audio,
and properly dithered digital audio has infinite "resolution" because it eliminates all quantization errors in the signal.


©2112 Chuck Zwicky