dunewind.jar
audio sample of three effects mixed together/all playing at once.
This is more experimental than work towards a given application or game.
I have been playing around with some noise sculpting via Java and thought I’d share the jar. Objectivity comes and goes, and I sometimes get into states where I can’t tell if something is intereting or a waste of time. So it would be nice to hear what others think. Among other things, I have found that it is really easy for me to hear what I think I programmed instead of hearing what is really being produced.
The noise sources in each case originated with 10-second “generated content” files from Audacity. Pretty cool that you can pick a type of noise (e.g., White, Pink, Brownian) and have it create a track for you.
In each of these, I am using a tool that plays back a continuous string of “macro-granules”, basically jumping into random places in the file and contatenating short fragments of the audio content with a bit of overlap to smooth the joins.
The tool was rebuilt so that each stereo channel could have its volume modulated independently. Also, a control was put in to allow variable playback speeds but it is not channel independent.
The rumble started with Brownian noise, filtered with 6 db rolloff at 75Hz, both low-pass and high-pass. It seems to break up (distort) with increasing likelihood as one pushes the volume above 70. I’m wondering how film theaters get that amazing rumble. Can’t seem to do it with headphones without cranking the volume of the external amplifier and turning all the other sounds way down in volume. Maybe theaters just have really strong sub-woofers hooked up. I have been experimenting with various filter settings of white and pink noise as well as Brownian, and have not found an “ultimate rumble”.
The “wind-in-ears” also uses Brownian noise, but in this case filtered high-pass at 400Hz and low-pass at 600Hz, 12 db rolloffs. I was playing around with some randomized lfo to create “gusts”. I was hoping to get a sense of direction by varying the time between when the turbulence altered the left and right ear sound sources. The first person offering feedback said it was too subtle for him to hear directionality. I did some bug-fixing and modifications and the setting on the graphic seems to sound pretty good. Try moving the “ear-to-ear” difference from around -15 to 15 and back, with turbulence well over 50, and the other components silent, if the ear-to-ear sense is not audible. I meant for the left-side of the slider to correspond to the head turning left and the wind hitting the right ear first (and vice versa for the other direction). Closing the eyes seems to help. Thoughts on whether or not this works as a possible VR effect?
The “sand” swirling all around is also kind of problematic. I altered the control that is providing “turbulence” to rise much more quickly than it drops (am imagining some sort of craft riding through sand dunes as if they were waves). With each bump, the playback rate drops (making the sound a little more thud-like) then speeds up (making the sound dissipate in a more hiss-like state). I made two of these sand-generating sliders to add to the surround-sound nature. I’m still pondering how to incorporate changes to convey directionality when one turns one’s head.
All in all, still playing around and experimenting with how these noises can be altered in realtime and make useful effects. Am trying to expose useful controls to help search and discover useful settings. I’ve spent so much time coding that I’ve kind of lost touch with sound design, so it is nice to exercise that muscle again and try and figure out how to cook up some decent noise-based effects.