def play(params):
""" Usage:
shine.py [length] [volume]
"""
length = params.get('length', dsp.stf(dsp.rand(0.1, 1)))
volume = params.get('volume', 100.0)
volume = volume / 100.0 # TODO: move into param filter
octave = params.get('octave', 2) + 1 # Add one to compensate for an old error for now
note = params.get('note', ['c'])
note = note[0]
quality = params.get('quality', tune.major)
glitch = params.get('glitch', False)
superglitch = params.get('superglitch', False)
pinecone = params.get('pinecone', False)
glitchpad = params.get('glitch-padding', 0)
glitchenv = params.get('glitch-envelope', False)
env = params.get('envelope', False)
ratios = params.get('ratios', tune.terry)
pad = params.get('padding', False)
bend = params.get('bend', False)
bpm = params.get('bpm', 75.0)
width = params.get('width', False)
wform = params.get('waveform', False)
instrument = params.get('instrument', 'r')
scale = params.get('scale', [1,6,5,4,8])
shuffle = params.get('shuffle', False) # Reorganize input scale
reps = params.get('repeats', len(scale))
alias = params.get('alias', False)
phase = params.get('phase', False)
pi = params.get('pi', False)
wild = params.get('wii', False)
root = params.get('root', 27.5)
trigger_id = params.get('trigger_id', 0)
tune.a0 = float(root)
try:
# Available input samples
if instrument == 'r':
instrument = 'rhodes'
tone = dsp.read('sounds/synthrhodes.wav').data
elif instrument == 's':
instrument = 'synthrhodes'
tone = dsp.read('sounds/220rhodes.wav').data
elif instrument == 'c':
instrument = 'clarinet'
tone = dsp.read('sounds/clarinet.wav').data
elif instrument == 'v':
instrument = 'vibes'
tone = dsp.read('sounds/glock220.wav').data
elif instrument == 't':
instrument = 'tape triangle'
tone = dsp.read('sounds/tape220.wav').data
elif instrument == 'g':
instrument = 'glade'
tone = dsp.read('sounds/glade.wav').data
elif instrument == 'p':
instrument = 'paperclips'
tone = dsp.read('sounds/paperclips.wav').data
elif instrument == 'i':
instrument = 'input'
tone = dsp.capture(dsp.stf(1))
except:
instrument = None
tone = None
out = ''
# Shuffle the order of pitches
if shuffle is not False:
scale = dsp.randshuffle(scale)
# Translate the list of scale degrees into a list of frequencies
freqs = tune.fromdegrees(scale, octave, note, quality, ratios)
freqs = [ freq / 4.0 for freq in freqs ]
# Format is: [ [ path, offset, id, value ] ]
# Offset for video
osc_messages = [ ['/dac', float(dsp.fts(length)), 1, tune.fts(osc_freq)] for osc_freq in freqs ]
# Phase randomly chooses note lengths from a
# set of ratios derived from the current bpm
if phase is not False:
ldivs = [0.5, 0.75, 2, 3, 4]
ldiv = dsp.randchoose(ldivs)
length = dsp.bpm2ms(bpm) / ldiv
length = dsp.mstf(length)
reps = ldiv if ldiv > 1 else 4
# Construct a sequence of notes
for i in range(reps):
# Get the freqency
freq = freqs[i % len(freqs)]
# Transpose the input sample or
# synthesize tone
if wform is False and tone is not None:
# Determine the pitch shift required
# to arrive at target frequency based on
# the pitch of the original samples.
if instrument == 'clarinet':
diff = freq / 293.7
elif instrument == 'vibes':
diff = freq / 740.0
else:
diff = freq / 440.0
clang = dsp.transpose(tone, diff)
elif wform == 'super':
clang = dsp.tone(length, freq, 'phasor', 0.5)
clang = [ dsp.drift(clang, dsp.rand(0, 0.02)) for s in range(7) ]
clang = dsp.mix(clang)
elif wform is False and tone is None:
clang = dsp.tone(length, freq, 'sine2pi', 0.75)
clang = dsp.amp(clang, 0.6)
else:
clang = dsp.tone(length, freq, wform, 0.75)
clang = dsp.amp(clang, 0.6)
# Stupidly copy the note enough or
# trim it to meet the target length
clang = dsp.fill(clang, length)
# Give synth tones simple env (can override)
if wform is not False and env is False:
clang = dsp.env(clang, 'phasor')
# Apply an optional amplitude envelope
if env is not False:
clang = dsp.env(clang, env)
# Add optional padding between notes
if pad != False:
clang = dsp.pad(clang, 0, pad)
# Add to the final note sequence
out += clang
# Add optional aliasing (crude bitcrushing)
if alias is not False:
out = dsp.alias(out)
# Cut sound into chunks of variable length (between 5 & 300 ms)
# Pan each chunk to a random position
# Apply a sine amplitude envelope to each chunk
# Finally, add variable silence between each chunk and shuffle the
# order of the chunks before joining.
if glitch is not False:
out = dsp.vsplit(out, dsp.mstf(5), dsp.mstf(300))
out = [dsp.pan(o, dsp.rand()) for o in out]
out = [dsp.env(o, 'sine') for o in out]
out = [dsp.pad(o, 0, dsp.mstf(dsp.rand(0, glitchpad))) for o in out]
out = ''.join(dsp.randshuffle(out))
# Detune between 1.01 and 0.99 times original speed
# as a sine curve whose length equals the total output length
if bend is not False:
out = dsp.split(out, 441)
freqs = dsp.wavetable('sine', len(out), 1.01, 0.99)
out = [ dsp.transpose(out[i], freqs[i]) for i in range(len(out)) ]
out = ''.join(out)
if wild is not False:
#out = dsp.vsplit(out, 400, 10000)
out = dsp.split(out, 3000)
out = [ dsp.amp(dsp.amp(o, dsp.rand(10, 50)), 0.5) for o in out ]
#out = [ o * dsp.randint(1, 5) for o in out ]
for index, o in enumerate(out):
if dsp.randint(0, 1) == 0:
out[index] = dsp.env(dsp.cut(o, 0, dsp.flen(o) / 4), 'gauss') * 4
if dsp.randint(0, 6) == 0:
out[index] = dsp.transpose(o, 8)
out = [ dsp.env(o, 'gauss') for o in out ]
freqs = dsp.wavetable('sine', len(out), 1.02, 0.98)
out = [ dsp.transpose(out[i], freqs[i]) for i in range(len(out)) ]
out = ''.join(out)
if pinecone == True:
out = dsp.pine(out, int(length * dsp.rand(0.5, 8.0)), dsp.randchoose(freqs) * dsp.rand(0.5, 4.0))
# Adjust output amplitude as needed and return audio + OSC
if pi:
return (dsp.amp(out, volume), {'osc': osc_messages})
else:
return dsp.amp(out, volume)
def play(params=None):
params = params or {}
length = params.get('length', dsp.stf(dsp.rand(5, 12)))
volume = params.get('volume', 20.0)
volume = volume / 100.0 # TODO move into param filter
octave = params.get('octave', 1)
notes = params.get('note', ['c', 'g'])
hertz = params.get('hertz', False)
quality = params.get('quality', tune.major)
glitch = params.get('glitch', False)
waveform = params.get('waveform', 'sine')
ratios = params.get('ratios', tune.terry)
alias = params.get('alias', False)
wild = params.get('wii', False)
bend = params.get('bend', False)
bbend = params.get('bbend', False)
env = params.get('envelope', 'gauss')
harmonics = params.get('harmonic', [1,2])
reps = params.get('repeats', 1)
root = params.get('root', 27.5)
pinecone = params.get('pinecone', False)
if bbend == True:
bend = True
tune.a0 = float(root)
# These are amplitude envelopes for each partial,
# randomly selected for each. Not to be confused with
# the master 'env' param which is the amplit
wtypes = ['sine', 'phasor', 'line', 'saw']
layers = []
if hertz is not False:
notes = hertz
for note in notes:
tones = []
for i in range(dsp.randint(2,4)):
if hertz is not False:
freq = float(note)
if octave > 1:
freq *= octave
else:
freq = tune.ntf(note, octave)
snds = [ dsp.tone(length, freq * h, waveform, 0.05) for h in harmonics ]
snds = [ dsp.env(s, dsp.randchoose(wtypes), highval=dsp.rand(0.2, 0.4)) for s in snds ]
snds = [ dsp.pan(s, dsp.rand()) for s in snds ]
if bend is not False:
def bendit(out=''):
if bbend == True:
bendtable = dsp.randchoose(['impulse', 'sine', 'line', 'phasor', 'cos', 'impulse'])
lowbend = dsp.rand(0.8, 0.99)
highbend = dsp.rand(1.0, 1.25)
else:
bendtable = 'sine'
lowbend = 0.99
highbend = 1.01
out = dsp.split(out, 441)
freqs = dsp.wavetable(bendtable, len(out))
freqs = [ math.fabs(f) * (highbend - lowbend) + lowbend for f in freqs ]
out = [ dsp.transpose(out[i], freqs[i]) for i in range(len(out)) ]
return ''.join(out)
snds = [ bendit(snd) for snd in snds ]
tones += [ dsp.mix(snds) ]
layer = dsp.mix(tones)
if wild != False:
layer = dsp.vsplit(layer, 41, 4410)
layer = [ dsp.amp(dsp.amp(l, dsp.rand(10, 20)), 0.5) for l in layer ]
layer = ''.join(layer)
if pinecone != False:
layer = dsp.pine(layer, length, freq, 4)
layer = dsp.env(layer, env)
layers += [ layer ]
out = dsp.mix(layers) * reps
# Format is: [ path, offset, id, value ]
if hertz is not False:
osc_message = ['/dac', 0.0, 0, tune.fts(notes[0])]
else:
osc_message = ['/dac', 0.0, 0, tune.nts(notes[0], octave - 1)]
#return (dsp.amp(out, volume), {'osc': [ osc_message ]})
return dsp.amp(out, volume)
