[continue autosave] Learn Sporth
log in learn new browse

by Paul Batchelor

_wav 8192 gen_sine _d var 2 0.2 1 0 _wav osc 5 20 biscale 1 dust dup _d set 1100 1.9 0 wpkorg35 jcrev _d get 0.5 maytrig 0.01 tgate 0.001 port 1000 0.3 0 _wav osc * + _d get 0.5 maytrig 0.004 tgate 0.001 port 2000 0.3 0 _wav osc * + _d get 0.2 maytrig 0.004 tgate 0.001 port 500 0.3 0 _wav osc * + _d get 0.3 maytrig 0.04 tgate 0.005 port -0.05 0.05 _d get 1102 11025 trand randh * + dup _d get 0.1 maytrig 0.4 0.1 0.1 tenv * 0.6 0.3 delay 12 1024 512 pshift 1000 butlp -6 ampdb * +
show code >>

This was a quick patch I whipped up some morning while eating some oatmeal. It is part of my ongoing collection of sporthlings. By and large, this patch was derived by layering simple sounds together.

Tables and Variables

Tables generation is minimal. A sine lookup table is generated for some oscillators, A single variable d is used for the dust signal.

 _wav 8192 gen_sine
 _d var


The dust ugen is a ugen that fires random impulses at a certain density. It is unique in that it works well as both a control signal as well as an audio signal. This patch uses this signal for both purposes.

The first argument to dust is amplitude, which is set to a constant 2.


The second argument is density, which is controlled by a low-frequency oscillator, set at 0.2 Hz. It is scaled via biscale to go between 5 and 20 impulses per second.

 0.2 1 0 _wav osc 5 20 biscale

The rest of dust is written out. The final argument to dust is the mode, which has been set to 1, bipolar mode. The signal is duplicated, and one of the values is set to teh variable d.

 1 dust dup _d set

The dust signal continues onward, being filtered by wpkorg35. It is set to a cutoff frequency of 1100 with a very high resonance (this particular filter has a range of 0-2, and the resonance is 1.9. The sonic output of putting impulses into a highly resonant filter little sinusoidal blip sounds. One of my favorite go-to effects.

 1100 1.9 0 wpkorg35

To add some body, the blips are fed into the chowning reverb, which I have found to be very favorable to sinusoidal signals for some reason.



The next layer in our cake of sound are what I will call "blerps". These are rhythmic blips tuned at whole integer ratios from one another.

The make up of a single "blerp" consists of a sinusoid and an envelope. The envelope is created gate signal (generated via tgate) being sent through a portamento filter. The gate is controlled via the dust signal stored in variable d. To make things less busy, the trigger is fed into a maytrig.

 _d get 0.5 maytrig 0.01 tgate 0.001 port 

The envelope generatored is then multipled with a sinusoidal oscillator tuned at 1000Hz. This signal is then added onto the current signal.

 1000 0.3 0 _wav osc *  + 

The same thing more or less is repeated 2 more times, with a few variations in parameters. The oscillators are tuned to 2000Hz and 500Hz.

 _d get 0.5 maytrig 0.004 tgate 0.001 port 
 2000 0.3 0 _wav osc *  + 
 _d get 0.2 maytrig 0.004 tgate 0.001 port 
 500 0.3 0 _wav osc *  + 


Randh, like dust, can function equally well as a control signal and an audio signal. When the rate is set to audio-rate frequencies, you can get different "colors" of noise, an sound reminiscent to 8-bit game sounds.

For starters, an envelope signal is generated in similar manner to the envelopes created in the previous section.

 _d get 0.3 maytrig 0.04 tgate 0.005 port 

The arguments for ugen randh can now begin. These two arguments tell randh to produce random values between (+/-)0.05.

 -0.05 0.05 

The third argument to randh is rate (or frequency), which we have delegated to a triggerable random number generator trand. The trigger signal is being fed by our dust signal, and is producing random values between 1102Hz and 11025Hz.

 _d get 1102 11025 trand  

Finally, randh is called and multiplied with the corresponding envelope. It is then added to the current signal.

 randh * + 


The effects processing is used in a very unusual way. One could argue that the processing is an instrument as well.

To begin, the signal is duplicated so that a dry version of the signal is stored.


The first thing created is a "throw" signal. The dust signal used from before feeds into a very sparse maytrig. This trigger signal feeds into a tenv, where a very gentle envelope signal is multipled with the signal to be processed. The results is we are occasionally "throwing" the signal into processor. This term is borrowed from what is known by mix engineers as a "reverb throw".

 _d get 0.1 maytrig 0.4 0.1 0.1 tenv * 

The thrown signal is fed into a feedback delay of 300ms with a feedback amount of 60 percent.

 0.6 0.3 delay 

The feedback delay is fed into the pitch shifter pshift, where it is tuned up an octave.

 12 1024 512 pshift 

The pitch shifted signal is put through a butterworth lowpass, and then attenuated by 6dB. The dry and wet signal are then added together.

 1000 butlp -6 ampdb * + 

Compile & play Stop audio