Workshop
Welcome to this Tidal Cycles tutorial. This is designed to be used as a worksheet during hands-on beginner/mixed workshops. By Lucy Cheesman, adapted to wiki format by Alex McLean.
Getting started
Once everything is installed, follow the following startup procedure each time.
Launch SuperDirt
SuperDirtshould be started automatically when you run the SuperCollider IDE application. If not, in the editor window of the SuperCollider IDE, type'SuperDirt.start'and run the code by holding downCtrland pressingEnter(while your cursor is on the same line as the code).Launch Tidal Cycles
In your text editor (Pulsar, vim, VS Code, etc), start a new file and save it with a
.tidalextension (e.g.examples.tidal). Tidal will be automatically launch when you type and execute your first command.
Estuary
Even if you haven't installed Tidal on your computer yet, it's still possible to play with it online. Estuary lets you play with Tidal and several other live-coding systems inside your browser, without the need to install anything in your own computer.
Estuary is a perfect place to learn, teach, play with others, and test distinct live-coding languages.
However, note that not all features in Tidal will work on Estuary, only a subset (called Mini-Tidal).
Notes in Haskell
Haskell uses double dashes -- at the beginning of a line to denote a comment. A comment is code that will be ignored by the interpreter. You can use comments to add notes in your code. You can also use comments to ignore a specific line or pattern:
–- I'm a comment
-- this pattern will not play
-- d1 $ s "bd hh sn hh"
-- "fast 2" will be ignored
d1
-- $ fast 2
$ s "hh*8"
Basic patterns
The basic format for making sound in Tidal looks like this:
d1 $ sound "drum"
You can stop making a sound using silence:
d1 $ silence
There are two types of sounds you can use with sound: either they are synths definitions (like superpiano, see Synthesizers), or they are samples. In the latter case, you write the name of the folder that contain the sample set. By default, the first sample is used, but you can pick a different sample from the same set, with :: and a number:
d1 $ sound "drum:1"
Also, it is possible to specify the folder and the sample in two parts:
d1 $ sound "drum" # n 1
Note that s is a synonym of sound, so d1 $ s "drum" # n 1 is the same pattern.
Default sample library
Some of the samples which come with Tidal are listed below. Try some out!
flick sid can metal future gabba sn mouth co gretsch mt arp h cp
cr newnotes bass hc tabla bass0 hh bass1 bass2 oc bass3 ho odx
diphone2 house off ht tink perc bd industrial pluck trump printshort
jazz voodoo birds3 procshort blip drum jvbass psr wobble drumtraks koy
rave bottle kurt latibro rm sax lighter lt arpy feel less stab ul
You can see what other sounds there are in the default library by looking in the Dirt-Samples folder. Find it via the SuperCollider menu: 'File > Open user support directory > downloaded-quarks > Dirt-Samples'. Additionally, you can also add your own custom samples. In the Pulsar editor, you can add a setting that will load a tab with all the Dirt-Samples (see Pulsar).
Make a sequence:
d1 $ sound "bd hh sn hh"
The more steps in the sequence, the faster it goes:
d1 $ sound "bd bd hh bd sn bd hh bd"
This is because of the way Tidal handles time. There is a universal cycle (sort of like a musical 'bar') which is always running. Tidal will play all of the sounds between the speech marks in one cycle, unless we tell it not to (we’ll learn how to do that later). You’ll also notice Tidal will space the sounds out evenly within the cycle Which means we can end up with polyrhythmic structures (more on those later). We can change the length of the cycle using setcps (where cps stands for cycles per second) - this is a bit like bpm (beats per minute).
setcps 0.6
You can use d1, d2, d3...d9 to play multiple sequences at the same time:
d2 $ sound "sn sn:2 sn bd sn"
You can stop all the running patterns with hush (or by pressing Ctrl+.).
You can pause everything by changing the cycle length to a negative number (remember to put negative numbers in brackets).
setcps (-1)
Start it up again with a positive number
setcps 0.6
Or you can solo one channel:
d1 $ sound "arpy cp arpy:2"
d2 $ sound "sn sn:2 bd sn"
solo 2
-- now only the second pattern will be playing
unsolo 2
-- now both will be playing, again
mute 2
-- now only the first pattern will be playing
unmute 2 -- (or unmuteAll)
-- now both will be playing
The Pulsar plugin adds some key shortcuts for this common operations, like Ctrl+1 to toggle mute for the first pattern, or Ctrl+0 to unmute all. You can see the complete list of keybindings inside Pulsar, by going to Edit > Preferences > Packages, selecting tidalcycles, and scrolling down to the Keybindings section.
More variety
Let's add some more variety to our sequences:
Add a silence/rest with ~:
d1 $ sound "bd ~ sn:3 bd sn:5 ~ bd:2 sn:2"
Fit a subsequence into a step with square brackets:
d1 $ sound "bd [bd cp] bd bd"
This can make for flexible time signatures:
d1 $ sound "[bd bd sn:5] [bd sn:3]"
You can put subsequences inside subsequences:
d1 $ sound "[[bd bd] bd sn:5] [bd sn:3]"
Keep going..
d1 $ sound "[[bd [bd bd bd bd]] bd sn:5] [bd sn:3]"
You can repeat a step with *:
d1 $ sound "bd sd*2"
This works with subsequences too:
d1 $ sound "bd [sd cp]*2"
Or you can do the opposite using /:
d1 $ sound "bd sn/2"
d1 $ sound "bd [sn cp]/2"
* works by 'speeding up' a step to play it multiple times. / works by 'slowing it down'.
We can also schedule patterns across cycles using < and >:
d1 $ sound "bd <sd cp arpy>"
d1 $ sound "<bd sn> <sd [cp cp]> <bd [cp cp]>"
The syntax we are using in these examples is called mini-notation, and can be used in many places within Tidal, not only the sound function.
Other common mini-notation symbols are | to choose a random option, , to play two patterns simultaneously, and ! to replicate a pattern.
Choose one of the two samples randomly:
d1 $ sound "[bd:0|bd:1]"
d1 $ sound "[sn|cp]"
Play a snare and a clap at the same time:
d1 $ sound "[sn,cp]"
Play three bass drums and a snare:
d1 $ sound "bd!3 sn"
Note the difference between this and "bd*3 sn": in the first example there are four events, all of them lasting the same time. In the latter, the three bd last for half a cycle, and the sn lasts the other half. "bd!3 sn" is the same as bd bd bd sn.
Effects
Vowel
Tidal has lots of effects we can use to change the way things sound. vowel is a filter which adds a vowel sound -- try a, e, i, o and u:
d1 $ sound "drum drum drum drum" # vowel "a"
We create patterns of effects in much the same way we create patterns of sounds. We call these effect and sound patterns 'control patterns'. So:
d1 $ sound "drum drum drum drum" # vowel "a o e e"
Remember that we can use "<>" to schedule across cycles:
d1 $ sound "drum drum drum drum" # vowel "<a o e e>"
You can add a non-vowel letter to pause the vowel effect:
d1 $ sound "drum drum drum drum" # vowel "a o p p"
Tidal does its best to map patterns across to one another:
d1 $ sound "drum drum drum drum" # vowel "a o e"
The structure comes from the left - try swapping the parameters:
d1 $ vowel "a o ~ i" # sound "drum"
Gain, pitch and panorama
gain changes the volume of different sounds:
d1 $ sound "bd hh sn:1 hh sn:1 hh" # gain "1 0.7 0.5"
speed and note are used for pitching samples. speed affects the speed of playback (e.g. 2 = up an octave):
d1 $ sound "numbers:1 numbers:2 numbers:3 numbers:4" # speed "1 1.5 2 0.5"
Or we can take the pattern from the speed parameter:
d1 $ speed "1 2 4" # sound "jungbass:6"
note pitches the sample up in semitones (e.g. 12 = up an octave):
d1 $ up "0 ~ 12 24" # sound "jungbass:6"
pan allows us to create stereo effects (0 = left, 0.5 = middle, 1 = right):
d1 $ sound "numbers:1 numbers:2 numbers:3 numbers:4" # pan "0 0.5 1"
Distortion, reverb, delay and filters
shape is one of the several function you can use to add distortion (but be careful - it also makes the sound much louder):
d1 $ sound "kurt:4 kurt:4" # shape "0 0.78" # gain "0.7"
Delay is achieved using the combination of up to four functions:
d1 $ sound "cp" # delay 0.8 # delaytime (1/6) # delayfeedback 0.6 # lock 1
Use lock 1 to indicate that the time provided to delaytime is in cycles instead of seconds.
All of them receive patterns:
d1 $ sound "industrial:3*4" # delay "<0 0.4 0.8>" # delaytime "0.2 0.05" # delayfeedback "<0.5 0.9>" # lock 1
To add a reverb effect use the functions dry, room and size:
d1 $ sound "[~ sn]*2" # dry 0.4 # room 0.6 # size 0.8
There are also several frequency filters available: low pass, high pass, dj type filter, among others.
Low pass filter:
d1 $ sound "tabla*4" # n "0 1 2 3" # cutoff 400 # resonance 0.2
High pass filter:
d1 $ sound "tabla*4" # n "0 1 2 3" # hcutoff 600 # hresonance 0.2
cutoff and hcutoff receive the frequency in hertz of the cutoff point. resonance and hresonance go from 0 to 1, but be aware that high resonance values can result in a very loud sound.
djf is a more immediate filter: it receives a number between 0 and 1. With values lesser than 0.5 it is a low pass filter, and with values greater than 0.5 it is a high pass filter.
You can take a look at the Effects section to learn more about effects and to see the complete list of effects.
Transforming patterns
We can start to make much more complex patterns using transformations. Using functions like slow you can start to transcend the cycle. slow stretches the pattern over more cycles:
Slow, fast and hurry
d1 $ sound "arpy arpy:1 arpy:2 arpy:3"
d1 $ slow 2 $ sound "arpy arpy:1 arpy:2 arpy:3"
fast squashes the pattern into less than one cycle. You might also see people writing density - it’s the same thing. Take a look:
fast 0.5 is the same as slow 2!
d1 $ fast 2 $ sound "arpy arpy:1 arpy:2 arpy:3"
d1 $ fast 0.5 $ sound "arpy arpy:1 arpy:2 arpy:3"
hurry is similar to fast, but also applies a speed transformation:
d1 $ sound "arpy arpy arpy:1 arpy:2"
d1 $ hurry 2 $ sound "arpy arpy arpy:1 arpy:2"
d1 $ hurry 0.5 $ sound "arpy arpy arpy:1 arpy:2"
See the Time section in the Reference to learn more about time-changing functions.
Reorganise patterns
Tidal Cycles offers many functions you can use to alter your patterns in different ways. In this section, some of them are introduced, but there are many more. You can check these reference sections to find more: alteration, accumulation and conditions.
You can reverse a pattern with rev:
d1 $ rev $ sound "arpy arpy:1 arpy:2 arpy:3"
Or play it forwards and then backwards with palindrome:
d1 $ palindrome $ sound "arpy arpy:1 arpy:2 arpy:3"
iter starts the pattern at a different point each cycle, shifting it the given number of times until it gets back to where it started:
d1 $ iter 4 $ sound "arpy arpy:1 arpy:2 arpy:3"
every allows us to schedule transformations or effects in different cycles. The following example will play twice as fast every four cycles:
d1 $ every 4 (fast 2) $ sound "arpy arpy:1 arpy:2 arpy:3"
... or you could schedule an effect in the same way, using #:
d1 $ every 4 (# vowel "a o") $ sound "arpy arpy:1 arpy:2 arpy:3"
jux (short for juxtapose) takes a transformation or an effect and plays it in one speaker the original pattern plays in the other speaker:
d1 $ sound "arpy arpy:1 arpy:2 arpy:3"
d1 $ jux (rev) $ sound "arpy arpy:1 arpy:2 arpy:3"
d1 $ jux (hurry 2) $ sound "arpy arpy arpy:1 arpy:2"
chunk applies a transformation or an effect to a different part of the pattern each time. For example with 4 as a parameter, it will step through each quarter of the cycle.
d1 $ chunk 4 (hurry 2) $ sound "arpy arpy:1 arpy:2 arpy:3"
d1 $ chunk 4 (# speed 2) $ sound "alphabet:0 alphabet:1 alphabet:2 alphabet:3"
Even further into transformations
More than one transformation is possible! You can chain them together using .:
d1 $ jux (rev . (slow 1.5)) $ sound "arpy arpy:1 arpy:2 arpy:3"
Remember that (almost) everything is a pattern so we can apply these transformations to our effects too:
d1 $ sound "jvbass [jvbass jvbass] jvbass ~" # note "1 [3 5] 7"
d1 $ sound "jvbass [jvbass jvbass] jvbass ~" # iter 3 (note "1 [3 5] 7")
You can create an LFO on any parameter by using fast or slow, range, and an oscillator such as sine or saw:
d1 $ d1 $ s "bd*8" # pan (slow 4 $ sine)
d1 $ s "moog*16" # n "<0 1 2>" # legato 1 # cutoff (range 200 2400 $ saw) # resonance 0.2
By default, oscillators such as sine, cosine or saw give values from 0 to 1. This is fine for some parameters (like pan), but you can use range to scale these values to whatever range you want.
The previous examples trigger one oscillator value for event. This is fine if there are a lot of events per cycle. However, if there are fewer, longer events, we need to pick several values from the oscillator in order to accomplish a smooth movement of the LFO. You can do this using control busses:
d1 $ s "moog" # n "<0 1 2>" # legato 1 # cutoffbus 1 (segment 32 $ range 200 2400 $ saw) # resonance 0.2
Here we can hear how the sound changes gradually during the cycle. There are busses for many parameters, all of them named like the parameter plus bus. In this last example, segment 32 tells the oscillator to pick 32 values each cycle.
Different kind of patterns
What is pattern, anyway? Let's think about some different kinds of pattern and how Tidal can represent them.
Cyclic / repetitive
We can use n to choose samples from a folder, this allows us to apply patterns there too:
d1 $ n "0 1 2 3" # sound "arpy"
run is a short way of writing out sequential patterns:
d1 $ n (run 4) # sound "arpy"
or we can use:
d1 $ n "0 .. 3" # sound "arpy"
Symmetry
d1 $ slow 2 $ n "0 1 2 3 3 2 1 0" # sound "arpy"
d1 $ palindrome $ n (run 4) # sound "arpy"
Polymetric / polyrhythmic sequences
Play two subsequences at once by using square brackets (sort of like one big subsequence!) separating with a comma:
d1 $ sound "[voodoo voodoo:3, arpy arpy:4 arpy:2]"
If you use curly brackets instead of square you get a different effect. With square brackets both halves of the sequence are fitted into the cycle (polyrhythm). With curly brackets the pulse is set by the left hand pattern. The right hand pattern can then overlap (or underlap!) (polymeter):
d1 $ sound "[voodoo voodoo:3, arpy arpy:4 arpy:2]"
d1 $ sound "{voodoo voodoo:3, arpy arpy:4 arpy:2}"
d1 $ sound "[drum bd hh bd, can can:2 can:3 can:4 can:2]"
d1 $ sound "{drum bd hh bd, can can:2 can:3 can:4 can:2}"
d1 $ sound "[bd sn, can:2 can:3 can:1, arpy arpy:1 arpy:2 arpy:3 arpy:5]"
d1 $ sound "{bd sn, can:2 can:3 can:1, arpy arpy:1 arpy:2 arpy:3 arpy:5}"
Euclidean rhythm/Bjorklund
If you give two numbers in brackets after an element in a pattern, then Tidal will try to distribute the first number of sounds equally across the second number of steps:
d1 $ sound "bd(5,8)"
You can use this notation within a single element of a pattern:
d1 $ sound "bd(3,8) sn*2"
d1 $ sound "bd(3,8) sn(5,8)"
You can also add a third parameter, which ‘rotates’ the pattern so it starts on a different step:
d1 $ sound "bd(5,8,2)"
Randomness
Randomness can help us quickly introduce character and variation into our patterns. sometimes works a bit like every, but instead of happening after a set period, changes have a random chance of appearing:
d1 $ sometimes (# speed "2") $ sound "drum*8"
often (75%) works like sometimes (50%) but happens more often:
d1 $ often (# speed "2") $ sound "drum*8"
irand generates a random integer up to the number specified. (e.g. to play a random sample):
d1 $ sound "arpy(3,8)" # n (irand 16)
rand generates a random decimal between 0 and 1:
d1 $ sound "tink*16" # gain rand
You can use degradeBy to remove random elements. The number indicates how likely a sample is to play:
d1 $ degradeBy 0.2 $ sound "tink*16"
(degrade on its own is the same as degradeBy 0.5)
Or, you can use ? to remove sounds with a 50% likelihood:
d1 $ sound "bd sn:2? bd sn?"
Manipulating Samples
So far we've just used short samples. Longer samples can cause us some problems if we’re not careful. Let’s see what happens with a long sample:
d1 $ sound "bev"
-- wait a bit, then..
hush
As you can hear, Tidal will keep triggering the sample each cycle, even if it’s very long. Even if you stop the pattern playing, you will still need to listen while the samples play out. You can use cut to truncate the sample when the next one is triggered:
d1 $ sound "bev" # cut 1
The number in cut define a group, so you can play with interference across different patterns:
d1 $ sound "bev ~" # cut 1
d2 $ slow 4 $ sound "pebbles ~" # cut 1
legato also truncates samples, but using a fixed length:
d1 $ sound "bev ~ bev ~" # legato 1
We can also chop samples for a granular synthesis effect:
d1 $ chop 32 $ sound "bev"
striate is similar to chop but organises the playback in a different way:
d1 $ slow 4 $ chop 4 $ sound "arpy:1 arpy:2 arpy:3 arpy:4"
d1 $ slow 4 $ striate 4 $ sound "arpy:1 arpy:2 arpy:3 arpy:4"
randslice chops the sample into pieces and then plays back a random one each cycle:
d1 $ randslice 32 $ sound "bev"
We can also use loopAt to fit samples to a set number of cycles:
d1 $ loopAt 8 $ sound "bev"
As always we can add patterns and transformations to these functions, or combine them for interesting effects:
d1 $ loopAt "<8 4 16>" $ chop 64 $ sound "bev*4" # cut 1
d1 $ rev $ loopAt 8 $ chop 128 $ sound "bev"
See more ways to manipulate longer samples at the Sampling reference section.
Superdirt synthesizers
So far we have used only samples, but SuperDirt also comes with many Supercollider synthesizers like superpiano, supersaw or superfm, also known as synths for short.
Each of them has it's own functions and parameters, but in general you can use them in a very similar way to samples:
d1 $ n "0 4 7" # sound "superpiano"
You can also control external synthesizers by MIDI or OSC
Difference between functions n and note
When using synths, both n and note functions are exactly the same: you may have noticed that the above example plays a C note, an E note (which is 4 semitones above C), and a G note (which is 7 semitones above C). This is exactly the same as:
d1 $ note "0 4 7" # sound "superpiano"
When using samples, n refers to the file index in the sample folder, sorted alphabetically (ascending) and counted from 0 (zero). It is possible for each sample to correspond to a note, if you have sampled every single note of an instrument. However when using note, the sample is pitched up or down (and the sample duration is affected accordingly).
So, for example:
d1 $ sound "bd*4" # n "<0 4>" # note "0 12 -7 -12"
This will play the first sample in the bd folder on odd cycles and the fifth sample on even cycles. On each cycle, the sample will be played 4 times: one as is, one pitched an octave above (12 semitones), one a fifth below (7 semitones), and the last one an octave below.
Playing notes
Most of this tutorial is dedicated to rhythm, but Tidal Cycles also offers ways to play notes, scales, chords and arpeggios.
You already know how to play notes: using the note function or, in case you have a per-note sampled instrument, choosing notes with the n function.
You can also write notes based on the Western Music Theory naming convention which uses the first 7 letters of the alphabet (A to G). For example, these two codes are equivalent:
d1 $ note "c a f e" # s "supermandolin"
d1 $ note "0 9 5 4" # s "supermandolin"
Note names are simply translated to numbers in tidal, so you can use either method, or both at the same time!
Note that you can follow any note name with s or f to indicate sharp and flat respectively. Also, note that 0 and c refer to the C note on the fifth octave. You can append the octave number following any note name:
d1 $ note "c4 a3 f6 e5" # s "supermandolin"
It can also be useful to move the octave using |+ or |-. This will play on the third octave:
d1 $ note "c a f e" # s "superpiano" |- note 24
To know more about how to play scales, chords and arpeggios, see the Harmony or the how-tos Build Arpeggios and Play Chords
Where to go from here
Some suggestions:
- Play, try, investigate. Here you have plenty of information to get you started. Look up the reference pages to learn more as you need it.
- Follow Alex's video course for a longer and deeper tutorial, with plenty of examples and video support.
- Join the forum and/or the discord server to ask for help, help others, and learn about how other people is using Tidal Cycles.