strudel-samples.txt

A presentation on loading and manipulating samples with strudel.

Table of Contents

All the percussion we have used thus far are samples.

We can choose different sample banks to pull samples from:

groove: stack("bd!2 [~ bd]!2".swing(2), "hh*8", "~ ~ cp ~").s()
  .bank("RolandTR808")

Try changing the sample bank:

The bank function works by putting the name of the sample bank with an _ at the front of sample names.

bank: s("bd*2").bank("RolandTR808")

explicit: s("RolandTR808_bd*2").late(1/4)

Find sample banks in the drum machine tab of the sounds window.

Not all sample banks support the same percussion sounds!

Changing the velocity of a sample changes its loudness.

d1: "bd*4".s().bank("RolandTR909")
  .velocity(range(0.6, 1, sine))

Strudel can start and end samples in different spots:

$: n("0*2").s("numbers").late(1/16).begin("0 0.2").cut(0)

$: "<bd sd>*8".s().end("0.1 0.5 0.75 1").slow(2)

Strudel also supports filters:

$: "sd*8".s().bank("RolandTR909")
  .hpf(range(0, 2000, sine).slow(8))
  .hpq("<0 0 10>") // Resonance
  .hpattack("<0 0.1 1 4 -4>".slow(2)) // Envelope attack

Samples can have variations; 2 ways to select the variations:

• Pass a pattern of sample indices into the n function
• Append the index to the sample name with :

samples({
  conga: [
    'https://glfmn.io/samples/menegass-conga-7.wav',
    // CC-BY 4.0: https://freesound.org/people/MrRentAPercussionist/sounds/455509/ 
    'https://glfmn.io/samples/mrrentapercussionist-congas-muted-slap.wav'
  ]
});

$: cat("conga:<0 1>*2".s(), "conga*2".s().n("0 1"))

One can make sample packs in strudel using a special strudel.json file.

{
  "_base": "https://glfmn.io/samples/",
  "conga": [ "menegass-conga-7.wav", "mrrentapercussionist-congas-muted-slap.wav" ],
  "tumbao": [ "stomachache-bongopattern145bpm" ],
  "train": [ "nnus_train_whistle1.wav" ]
}

_base

A URL that gets added to the beginning of every other URL in the file.

Try playing this sample:

samples({
  tumbao: ['https://glfmn.io/samples/stomachache-bongopattern145bpm.wav']
});

$: "tumbao".s()

cut groups can mitigate the chaos. Try adding .cut(1).

We can give more space for the sample to play out with slow:

samples({
  tumbao: ['https://glfmn.io/samples/stomachache-bongopattern145bpm.wav']
});

$: "tumbao".s().cut(1).slow(2)

We can speed up the sample to hear more of it:

samples({
  tumbao: ['https://glfmn.io/samples/stomachache-bongopattern145bpm.wav']
});

$: "tumbao".s().cut(1).slow(2).speed(1.25)

We can even try changing strudels tempo:

samples({
  tumbao: ['https://glfmn.io/samples/stomachache-bongopattern145bpm.wav']
});

setcps(1/2);

$: "tumbao".s().cut(1).slow(2).speed(1.25)

setcps sets cycles-per-second. A lower cps results in a faster tempo.

We can use a formula to convert bpm to cps:

const bpm = 120;
setcps(bpm/60/4);

loopAt speeds up the sample so that it loops perfectly in n cycles.

samples({
  tumbao: ['https://glfmn.io/samples/stomachache-bongopattern145bpm.wav']
});

setcps(125/60/4)

$: "tumbao".s().loopAt(5)

Combining knowledge of sample length and strudel's tempo is key!

fit makes the pattern perfectly fit its event duration:

$: n("<0 [0 1]>").s("numbers").end("<1!2 0.5>/2").fit()

Strudel has lots of fun and gnarly ways of chopping up samples and making them feel new again.

But first, let's get some break beats :)

samples('github:yaxu/clean-breaks');

setcps(120/60/4)

$: s("amen").loopAt(4)

chop seamlessly splits the sample into multiple parts.

samples('github:yaxu/clean-breaks');

setcps(144/60/4)

$: s("amen").chop(8).color("<blue red>").loopAt(4)
  ._punchcard({ labels: 1 })

Once the sample is chopped, we can start to transform it with patterns:

samples('github:yaxu/clean-breaks');

setcps(120/60/4);

$: s("groove").chop(32).loopAt(2)
  .velocity(sine.range(0.5, 1).fast(2))
  .mul(speed("0.75 1.5 0.9 1"))
  ._punchcard({ labels: 1 })

Instead of chopping, we can striate:

$: s("numbers:0 numbers:1 numbers:2 numbers:3").cut(1)
  .color("red white purple brown")
  .every(2, striate(4))
  ._punchcard({ labels: 1 })

Repeats the pattern n times, but only playing the 1st nth of the sample the first time, the 2nd nth the next time, and so on.