[sdiy] harmonics & vibrato

[Richard Wentk]

At 11:10 18/12/2002 +0000, Brendan wrote:
>I suppose I wondered if there might be an area under/near the finger that is
>partially damped - somewhere between being fully stopped and freely
>vibrating - and so any modes that have peaks in that area might be reduced
>in amplitude, changing the timbre if the shape of the finger pad changes.
>
>Having said that, only energy at frequencies at the very limits of human
>hearing (and above those limits) have peaks this close to their nodes (in
>the millimetre scale) so as interesting a thought as it may be, it probably
>just doesn't matter! :)

No, I think it does. I've been thinking a lot about how real sounds are 
waaaay more complicated than we think they are.

Like vibrato. Firstly real vibrato isn't a sine wave. It's going to be 
slightly randomised wrt both pitch and amplitude. And it probably won't be 
perfectly symmetric. And it will have a frequency curve that varies. Listen 
to a real cellist and you can hear that their vibrato often starts slow and 
then speeds up a little.

Secondly I'm wondering how the speed of a moving finger compares with the 
speed of the displacement waves in a string that create the sound. If 
they're comparable, you're going to get harmonic non-linearities because 
the nodes won't have time to settle the way they would if a fixed obstacle 
were damping the string.

Bottom line is you'll get interesting timbral variations that are richer 
than just a generic fixed LFO wobble. If I had a frequency shifter to hand 
(which I don't) I'd wonder about creating string vibrato with a slightly 
randomised LFO driving the frequency shift amount as well as VCO frequency.

If anyone does have one to hand, I'd be interested to hear how the results 
compare with simple vibrato.

Richard