Discover adamrenklint/bap Open Source project

Bap is a toolkit for making beats and composing sequences with Javascript and Web Audio for playback in modern browsers. It is inspired by the classic "MPC workflow" and built to make all aspects of beatmaking completely modular and reusable.

Made by Adam Renklint, Berlin april 2015

Install and import

From npm

$ npm install --save bap

From rawgit CDN

<script src="https://cdn.rawgit.com/adamrenklint/bap/v0.8.0/bap.min.js"></script>

Usage

var bap = require('bap'); // or window.bap

// a kit is like an instrument, or program in mpc terms
var kit = bap.kit();
var oscillator = bap.oscillator({
  frequency: 440
});
// a kit connects infinite slots with infinite layers
kit.slot('Q').layer(oscillator);
kit.slot('W').layer(oscillator.with({ frequency: 330 }));
kit.slot('E').layer(bap.sample('foo.wav'));

// a pattern is a loop made up of channels and notes
var pattern = bap.pattern();
pattern.channel(1).add(
  ['1.*.01', '1Q', 48, 70, 0, -50],
  ['1.2.01', '1W', 96, 100, 0, 50],
  ['1.4.01', '1E']
);

// connect the kit, and play
pattern.kit(1, kit).start();

Basic concepts

Bap runs at 96 ticks per beat, with a position signature like MPC: bar.beat.tick
Kits are like instruments (programs in MPC terms) and contains infinite slots, each with infinite layers of samples and oscillators
Patterns are playable collections of channels containing notes, and connect with kits
Notes are defined by six main parameters: position, target, duration, volume, pan and pitch
Only position and target params are required, all others can be null/falsy/undefined
Positions containing expressions are automatically expanded
When a layer is played, it merges the params of the note, channel, layer, slot and kit

Known issues

Creating effect nodes on-the-fly is not performing well in Firefox, resulting in clipping on the initial run of a pattern

Resources

API

All objects are based on ampersand-state
on(name, callback) register event callback
off(name, [callback]) unregister event callback
once(name, callback) register single-run event callback
with(params) return a clone of itself with params
toJSON() return current params as JSON

bap

clock reference to clock singleton
volume number between 0 and 999, master volume setting, defaults to 100
kit(params) returns a new kit
slot(params) returns a new slot
layer(params) returns a new layer
oscillator(params) returns a new oscillator
sample(params) returns a new sample
pattern(params) returns a new pattern
sequence(sequence..., params) returns a new sequence
channel(params) returns a new channel
note(params) returns a new note
reverb(params) returns a new reverb
delay(params) returns a new delay
compressor(params) returns a new compressor
overdrive(params) returns a new overdrive
filter(params) returns a new filter
chorus(params) returns a new chorus
phaser(params) returns a new phaser
pingpong(params) returns a new [ping pong delay](#ping pong delay)
new() returns a new instance of Bap, with its clock and event bus separated from other instances

params

mute boolean, defaults to false
volume number between 0 and 999, defaults to 100
length number, length in seconds, overriden by duration if shorter
duration number, duration in ticks, overriden by length if shorter
attack number, attack in seconds
release number, release in seconds
pitch number between -999 and 999 representing the pitch shift in semitones, defaults to 0
pan number between -100 and 100, defaults to 0

events

started triggered on Note, Channel, Pattern, Kit, Slot and Layer when sound source starts playing
stopped triggered on Note, Channel, Pattern, Kit, Slot and Layer when sound source stops playing

clock

playing boolean, current state of playback, can be changed to start or pause
position string in format bar.beat.tick, can be set to move playback position
bar, beat, tick numbers, equal and bound to position, can be set to move playback position
tempo number, current tempo of playback, read only
step function, called on each step with note and time as arguments, able to cancel step by returning false
sequence pattern or sequence currently playing
start() start playback, if current pattern is set
start(pattern) set current pattern and start playback
pause() stop playback
stop() stop playback and set position to 1.1.01

kit

slot() returns blank slot assigned to next id
slot(id) returns existing or blank slot with id
slot(id, slot) assign slot instance to id
slot(slot) assign slot instance to next id
connect(effect) route output signal to destination via effect or chain
bypass boolean to bypass all effects, effect type string or array of strings to bypass specific effect types, defaults to false

slot

layer() returns a blank layer assigned to next id
layer(id) returns existing or blank layer with id
layer(id, layer) assign layer instance to id
layer(layer) assign layer instance to next id
layer(sampleSrc) returns a new sample layer, assigned to next id
start(time, [params]) start playback of slot at (AudioContext) time
start([params]) start playback of slot immediately
stop(time, [params]) stop playback of slot at (AudioContext) time
stop([params]) stop playback of slot immediately
connect(effect) route output signal to destination via effect or chain
bypass boolean to bypass all effects, effect type string or array of strings to bypass specific effect types, defaults to false

layer

start(time, [params]) start playback of slot at (AudioContext) time
start([params]) start playback of slot immediately
connect(effect) route output signal to destination via effect or chain
bypass boolean to bypass all effects, effect type string or array of strings to bypass specific effect types, defaults to false

oscillator

frequency number, frequency of oscillation in hertz, defaults to 0
note string, note identifier like C3 or a4 - if set, overrides frequency
shape string, shape of waveform, defaults to sine, other values are square, sawtooth, triangle and custom

sample

src string, url used to load sample buffer
offset number, starting point offset in seconds, defaults to 0
channel string, defines how to handle stereo buffers: left or right uses a single channel, merge and diff combines or differentates between channels, default is null and does nothing
reverse boolean, reverse buffer or slice of buffer
loop number, loop length in seconds, defaults to 0 i.e. not looping
slice(pieces) returns a kit with the sample sliced into even-sized sections
bitcrush number between 0 and 16, resamples waveform to defined bit depth, defaults to 0, i.e. no resampling
bitcrushFrequency number between 20 and 22050, normalization frequency at which to apply the bitcrusher effect, defaults to 6500
bitcrushMix number between 0 and 100, ratio of wet bitcrushed signal to mix with dry signal, defaults to 50
trimToZeroCrossingPoint boolean, automatically trim sample start and end to zero crossing point to avoid clipping, defaults to true

pattern

playing boolean, current state of playback, can be changed to start or pause
tempo number, playback tempo in bpm, defaults to 120
bars number, length of pattern in bars, defaults to 1
beatsPerBar number, amount of beats per bar, defaults to 4
loop boolean, define if pattern should loop, defaults to true
volume number between 0 and 999, master volume for pattern, defaults to 100
transform function to be called after expanding position expressions into notes, called after channel.transform
channel() returns a blank channel assigned to next id
channel(id) returns existing or blank channel with id
channel(id, channel) assign channel instance to id
channel(channel) assign channel instance to next id
start() start playback of pattern
pause() stop playback
stop() stop playback and set position to 1.1.01
kit(id, kit) connect kit to id
kit(id) return kit connected to id
then(sequence, ...) return new sequence with passed sequences and patterns after pattern
after(sequence, ...) return new sequence with passed sequences and patterns before pattern
and(sequence, ...) return new sequence with passed sequences and patterns layered with pattern

sequence

constructor(sequence, ..., [params]) the sequence constructor optionally takes any number of sequences and patterns as argument before the usual params
playing boolean, current state of playback, can be changed to start or pause
loop boolean, define if sequence should loop, defaults to false
sequences an array of sequences, patterns or arrays of sequences and patterns
bars number, length in bars, read-only
then(sequence, ...) return new sequence with passed sequences and patterns after sequence
after(sequence, ...) return new sequence with passed sequences and patterns before sequence
and(sequence, ...) return new sequence with passed sequences and patterns layered with sequence

channel

transform function to be called after expanding position expressions into notes, called after note.transform, can return false to not execute pattern.transform
add(note, note, ...) schedule note(s) to be played within context of channel
connect(effect) route output signal to destination via effect or chain
bypass boolean to bypass all effects, effect type string or array of strings to bypass specific effect types, defaults to false

note

transform function to be called after expanding position expressions into notes, called before channel.transform, can return false to not execute channel.transform
start([time]) start playback of note at (AudioContext) time or immediately
stop([time]) stop playback of note at (AudioContext) time or immediately
connect(effect) route output signal to destination via effect or chain
bypass boolean to bypass all effects, effect type string or array of strings to bypass specific effect types, defaults to false

reverb

wet number between 0 and 999, amount of wet signal, defaults to 30
dry number between 0 and 999, amount of dry signal, defaults to 100
time number, impulse time in seconds, defaults to 1
decay number, drop off time in seconds, defaults to 3
filter string, type of filter (highpass, lowpass, bandpass, lowshelf, highshelf, peaking, notch, allpass), defaults to highpass
cutoff number, frequency where filter is applied, defaults to 2000
reverse boolean, defaults to false
bypass boolean, defaults to false

delay

wet number between 0 and 999, amount of wet signal, defaults to 50
dry number between 0 and 999, amount of dry signal, defaults to 100
sync boolean, sync delay with current tempo, defaults to false
time number between 0.001 and 4, delay time in seconds (or beats if sync is true), defaults to 0.3
feedback number between 0 and 999, amount of regeneration from processed signal, defaults to 50
filter string, type of filter (highpass, lowpass, bandpass, lowshelf, highshelf, peaking, notch, allpass), defaults to highpass
cutoff number, frequency where filter is applied, defaults to 2000
bypass boolean, defaults to false

compressor

threshold number between -100 and 0, decibel value above which the compression will start taking effect, defaults to -12
knee number between 0 and 40, decibel value representing the range above the threshold where the curve smoothly transitions to the compressed portion, defaults to 30
ratio number between 0 and 20, amount of change in dB needed in input for 1 dB change in the output, defaults to 12
attack number, seconds required to reduce the gain by 10 dB, defaults to 0
release number, seconds required to increase the gain by 10 dB, defaults to 0.25
gain number between 0 and 999, amount of gain to processed signal, defaults to 100
bypass boolean, defaults to false

overdrive

wet number between 0 and 999, amount of wet signal, defaults to 50
dry number between 0 and 999, amount of dry signal, defaults to 50
preBand number between 0 and 100, amount of preband filtering, defaults to 50
color number between 20 and 22050, frequency cutoff for preband filtering, defaults to 800
postCut number between 20 and 22050, frequency cutoff for post filter, defaults to 3000
gain number between 0 and 999, amount of gain to processed signal, defaults to 100
bypass boolean, defaults to false

filter

based on BiquadFilterNode
wet number between 0 and 999, amount of wet signal, defaults to 50
dry number between 0 and 999, amount of dry signal, defaults to 50
shape string, type of filter (highpass, lowpass, bandpass, lowshelf, highshelf, peaking, notch, allpass), defaults to highpass
frequency number between 20 and 22050, frequency at which to apply effect, defaults to 440
q number between 0.001 and 100, controls the frequency band width or peak at cutoff, defaults to 1
value number between -999 and 999, amount of gain to affected frequency band, defaults to 0
gain number between 0 and 999, amount of gain to processed signal, defaults to 100
bypass boolean, defaults to false

chorus

wet number between 0 and 999, amount of wet signal, defaults to 50
dry number between 0 and 999, amount of dry signal, defaults to 50
rate number between 0.01 and 99, defaults to 1.5
feedback number between 0 and 999, defaults to 0.2
delay number between 0 and 1, defaults to 0.005
gain number between 0 and 999, amount of gain to processed signal, defaults to 100
bypass boolean, defaults to false

phaser

wet number between 0 and 999, amount of wet signal, defaults to 50
dry number between 0 and 999, amount of dry signal, defaults to 50
rate number between 0.01 and 99, defaults to 1.5
depth number between 0 and 1, defaults to 0.3
feedback number between 0 and 999, defaults to 0.2
stereoPhase number between 0 and 180, defaults to 45
modulationFrequency number between 500 and 1500, defaults to 750
gain number between 0 and 999, amount of gain to processed signal, defaults to 100
bypass boolean, defaults to false

ping pong delay

wet number between 0 and 999, amount of wet signal, defaults to 50
dry number between 0 and 999, amount of dry signal, defaults to 50
feedback number between 0 and 1, defaults to 0.2
left number between 0.001 and 10, left channel delay in seconds, defaults to 0.15
right number between 0.001 and 10, right channel delay in seconds, defaults to 0.2
gain number between 0 and 999, amount of gain to processed signal, defaults to 100
bypass boolean, defaults to false

Feedback and issues

Report bugs with Github Issues
Ask questions on Gitter
Send shoutouts on Twitter
Or just email me

Develop

npm install install all dependencies
npm start run examples development server
npm test run tests
npm run test:watch run and watch tests
npm run coverage generate coverage report with Istanbul
npm publish run tests, publish to npm, build minified version, tag and deploy examples

Contribute

Check the roadmap and open issues
Fork the repo
Add tests and/or examples
Submit a pull request

Props

Roger Linn for making MPC60
Marley Marl for innovating the art of sampling
DJ Premier, J Dilla, Pete Rock, Damu the Fudgemunk, Black Milk, Large Professor and Apollo Brown for making dope beats to be inspired from
Matt McKegg for building Bopper and Ditty, which are used in this project since version 0.1

License

Generated with redok @ Thursday July 23rd, 2020 - 11:44:40 PM

adamrenklint/bap

adamrenklint

Reviews

Repository Details