[sdiy] Basic waveform question

[Magnus Danielson]

jbv skrev:
> David,
> 
> Perception of sound is a very complex topic and I'm far from being an expert,
> but there are a few things that I know : perception is not so much related to
> the shape of the waveform, but rather to its harmonic content...
> Furthermore, there are plenty of acoustic "oddities", for instance : if a
> waveform
> is lacking its fundamental, but if the 1st and 2nd harmonics have a certain
> amplitude
> ratio, then the ear will "hear" a "ghost" fundamental that isn't in the
> signal...
> Last but not least, continuous fixed tones can be perceived as similar over
> long
> periods of time, as the human ear is built to catch differences rather than
> similar aspects
> of sounds...
> I guess experts could provide tons of such clues... So the conclusion might be
> :
> don't expect too much similarity between what you hear and what you see on the
> scope...

The hard part is not to measure something and see a difference there, 
the hard part is to correlate it to what you hear, to find a useable 
measure which predicts some sensation and to be able to design a control 
to achieve this sensation. Some sensations is fairly easy (frequency, 
frequency stability, amplitude etc) where as others is not as easy, 
especially for a control to go "all the way" in a useful maner.

Waveshape isn't everything. Phase shift the content with some fixed 
phase-shifting and waveshape does change where as overtone spectra 
remains unchanged.

In the same fashion, spectrum analysis as such does not give much hints, 
spectrum over time or over amplitude acceleration (LaPlace/Z-transform) 
gives much more information. Yet again this needs to be interpreted.

Cheers,
Magnus