[PD] Rainy days

Andy Farnell padawan12 at obiwannabe.co.uk
Sat Jan 12 01:21:37 CET 2008

Something new to share for a bit of fun. It's for an exercise in the synthetic
sound design book and it's about causality in sound. Some bits are old and maybe
you've heard them before on my site, but the final ensemble is fresh. 

The first part is a wind model. It starts with a global (world) variable for 
windspeed and adds local gusts according to buildings nearby. Then objects are 
placed in the scene like telephone poles, trees and rocks. This is a new version
that works by doing rule of thumb calculations on turbulence (critical
Reynolds number from size and texture) to guesstimate the frequency (quasi-
oscillation from vortex shedding). Thin wires whistle with a high sound and big
buildings groan and wail with a much lower sound. 

Okay here's that...


Next there's rain

If you're into graphics then the best way of thinking about this is as a 
"particle system", or rather it's about statistical distributions. Normally 
the rain just falls down in a regular stream and makes the sound of whatever 
it hits (usually ground, leaves or puddles of water). Noise is treated to
get a certain distribution and width of clicks in a Gabor/Stockhausen sense,
some things in the scene can be simple resonantors driven by the droplets.


Now, another model in my box of tricks is a window pane I've built (from
Perrys square lamina description... that's a glass laminar model made with
banded waveguides.)

Here's a few knocks on the virtual window with a virtual stick.


So I sit here looking out of my real window, and I guess artists mimic their
surroundings, so this sound is an insight into where I live. Tonight the rain
is lashing down and the wind is howling in the trees like a typical British 
winter. I thought this would make a good example of causality. What if we 
take the wind environment, put the rain particles into the wind and add a 
window? Here's what it sounds like inside on a stormy day (the wind is 
deliberately modulated in a periodic way to accentuate the effect).


I hope this gives something interesting about potential of procedural 
audio fx with Pd, maybe this isn't the sort of thing one can do so easily 
with traditional sound design. 

best 2 all,

Use the source

More information about the Pd-list mailing list