Jesse Bangs wrote:

> Why do people often think of the voiceless sounds as "hard" and the
> voiced ones as "soft"?  I've always thought of it the other way around.

It completely escapes my understanding how anyone could consider [t] to
be softer than [d], or [f] softer than [v].  Try saying [afata] and
[avada], that should make it clear.  In the first utterance, the
consonants interrupt the flow of the word with percussive unvoicedness,
while the second word glides off the tongue in a single soft mellifluous curve.

As for the physical aspect:  Unvoiced consonants have a lot of
high-frequency spikes, like percussion instruments in music, while
voiced ones have much smoother Fourier signatures, like plucked strings.
  Surely nobody would consider a violin pizzicato harder than a drum solo?

-- Christian Thalmann