juno.py modified to respect voices=.. (for tulip juno6 app).

amy_headers.py

@@ -363,10 +363,8 @@ def nothing(message):
  # Do juno
  for i in range(128):
  amy.log_patch()
- p = juno.JunoPatch()
- j = p.from_patch_number(i)
- j.base_oscs = list()
- v = j.get_new_voices(1)
+ j = juno.JunoPatch()
+ j.set_patch(i)
  f.write("\t/* %d: Juno %s */ \"%s\",\n" % (i, j.name, amy.retrieve_patch())) 
  num_oscs.append(5)
  # Do dx7

juno.py

@@ -151,20 +151,6 @@ def get_juno_patch(patch_number):
  return name, bytes(sysex)
 
 
-class NoteObj:
- """Object to receive note_on(note, velocity) messages."""
-
- def __init__(self, osc):
- self.osc = osc
-
- def note_on(self, note, velocity, time=None):
- self.note = note
- amy.send(osc=self.osc, note=note, vel=velocity, time=time)
-
- def note_off(self, time=None):
- amy.send(osc=self.osc, vel=0, time=time)
-
-
 class JunoPatch:
  """Encapsulates information in a Juno Patch."""
  name = ""
@@ -210,12 +196,6 @@ class JunoPatch:
  BITS1 = ['stop_16', 'stop_8', 'stop_4', 'pulse', 'saw']
  BITS2 = ['pwm_manual', 'vcf_neg', 'vca_gate']
 
- # Attributes for voice management.
- next_osc = 0
- # How many AMY oscs are used per voice?
- oscs_per_voice = 5
- # List of base_oscs for allocated voices.
- base_oscs = []
  # Patch number we're based on, if any.
  patch_number = None
  # Name, if any
@@ -228,6 +208,8 @@ class JunoPatch:
  sub_freq = '440'
  # List of the 5 basic oscs that need cloning.
  oscs_to_clone = set()
+ # Amy voices
+ amy_voices = list()
 
  @staticmethod
  def from_patch_number(patch_number):
@@ -263,8 +245,8 @@ def _init_from_sysex(self, sysexbytes):
  # Bits 3 & 4 also have flipped endianness & sense.
  setattr(self, 'hpf', [3, 2, 1, 0][int(sysexbytes[17]) >> 3])
 
- def __init__(self, base_osc=0):
- self.next_osc = base_osc
+ def set_voices(self, amy_voices):
+ self.amy_voices = amy_voices
 
  def _breakpoint_string(self):
  """Format a breakpoint string from the ADSR parameters reaching a peak."""
@@ -275,7 +257,7 @@ def _breakpoint_string(self):
 
  def init_AMY(self):
  """Output AMY commands to set up patches on all the allocated voices.
- Send amy.send(osc=base_osc, note=50, vel=1) afterwards."""
+ Send amy.send(osc=0, note=50, vel=1) afterwards."""
  #amy.reset()
  # base_osc is pulse/PWM
  # base_osc + 1 is SAW
@@ -284,7 +266,7 @@ def init_AMY(self):
  # base_osc + 4 is LFO
  # env0 is VCA
  # env1 is VCF
- # These are offsets from the base_oscs[].
+ # These are the canonical oscs (to add to amy.send(voices=..)).
  self.pwm_osc = 0
  self.saw_osc = 1
  self.sub_osc = 2
@@ -303,25 +285,8 @@ def init_AMY(self):
  self.update_vcf()
  self.update_env()
  self.update_cho()
- self.init_voices()
-
- def init_voices(self):
- """Having set up the base set of oscs, copy to the other voices."""
- # Assume base_oscs[0] is configured, setup the remainder.
- if self.base_oscs:
- base_osc = self.base_oscs[0]
- self.clone_oscs()
- for other_base_osc in self.base_oscs[1:]:
- for osc in self.voice_oscs + [self.lfo_osc]:
- proto_osc = base_osc + osc
- target_osc = other_base_osc + osc
- amy.send(osc=target_osc, clone_osc=proto_osc)
- # chained_osc is cleared by clone_osc, set them up.
- for i in range(len(self.voice_oscs) - 1):
- # The first 3 voice oscs chain to the next one
- amy.send(osc=other_base_osc + self.voice_oscs[i],
- chained_osc=other_base_osc + self.voice_oscs[i + 1])
-
+ self.clone_oscs()
+
  def _amp_coef_string(self, level):
  return '0,0,%s,1,0,0' % ffmt(max(.001, to_level(level) * to_level(self.vca_level)))
 
@@ -331,27 +296,24 @@ def _freq_coef_string(self, base_freq):
 
  def clone_oscs(self):
  """Clone modifications on osc 0 to remaining oscs, then fixup."""
- if not self.base_oscs:
- return
- base_osc = self.base_oscs[0]
- clone_osc = base_osc + self.pwm_osc
+ clone_osc = self.pwm_osc
  # Make the clones have their filters turned off.
- amy.send(osc=clone_osc, filter_type=amy.FILTER_NONE)
- amy.send(osc=base_osc + self.saw_osc, clone_osc=clone_osc)
- amy.send(osc=base_osc + self.saw_osc, wave=amy.SAW_UP,
+ self.amy_send(osc=clone_osc, filter_type=amy.FILTER_NONE)
+ self.amy_send(osc=self.saw_osc, clone_osc=clone_osc)
+ self.amy_send(osc=self.saw_osc, wave=amy.SAW_UP,
  amp=self._amp_coef_string(float(self.saw)),
- chained_osc=base_osc + self.sub_osc)
- amy.send(osc=base_osc + self.sub_osc, clone_osc=clone_osc)
- amy.send(osc=base_osc + self.sub_osc, wave=amy.PULSE,
+ chained_osc=self.sub_osc)
+ self.amy_send(osc=self.sub_osc, clone_osc=clone_osc)
+ self.amy_send(osc=self.sub_osc, wave=amy.PULSE,
  amp=self._amp_coef_string(self.dco_sub),
  freq=self.sub_freq,
- chained_osc=base_osc + self.nse_osc)
- amy.send(osc=base_osc + self.nse_osc, clone_osc=clone_osc)
- amy.send(osc=base_osc + self.nse_osc, wave=amy.NOISE,
+ chained_osc=self.nse_osc)
+ self.amy_send(osc=self.nse_osc, clone_osc=clone_osc)
+ self.amy_send(osc=self.nse_osc, wave=amy.NOISE,
  amp=self._amp_coef_string(self.dco_noise)) # No chained_osc.
  # Re-enable the filter on PWM osc. Also, enable its chaining.
- amy.send(osc=clone_osc, filter_type=amy.FILTER_LPF24,
- chained_osc=base_osc + self.saw_osc)
+ self.amy_send(osc=clone_osc, filter_type=amy.FILTER_LPF24,
+ chained_osc=self.saw_osc)
 
  def update_lfo(self):
  lfo_args = {'freq': to_lfo_freq(self.lfo_rate),
@@ -433,8 +395,7 @@ def update_cho(self):
  # We choose juno 60-style I+II. Juno 6-style would be freq=8 depth=0.25
  chorus_args['chorus_freq'] = 1
  chorus_args['chorus_depth'] = 0.08
- #self.amy_send(osc=amy.CHORUS_OSC, **chorus_args)
- # *Don't* repeat for all the notes, these ones are global.
+ # *Don't* send to oscs, these ones are global.
  amy.send(**chorus_args)
 
  # Setters for each Juno UI control
@@ -449,62 +410,26 @@ def set_param(self, param, val):
  getattr(self, 'update_' + group)()
  if self.recloning_needed:
  self.clone_oscs()
- self.update_voices()
 
  def send_deferred_params(self):
  for group, params in self.post_set_fn.items():
  if self.dirty_params.intersection(params):
  getattr(self, 'update_' + group)()
  self.clone_oscs()
- self.update_voices()
  self.dirty_params = set()
  self.defer_param_updates = False
 
- def amy_send(self, osc, **kwargs):
- if self.base_oscs:
- base_osc = self.base_oscs[0]
- # Adjust relative args.
- offset_args = dict(kwargs)
- for relative_arg in ['mod_source', 'chained_osc']:
- if relative_arg in offset_args:
- offset_args[relative_arg] += base_osc
- # Apply configuration in full to first voice.
- amy.send(osc=base_osc + osc, **kwargs)
- self.oscs_to_clone.add(osc)
-
- def update_voices(self):
- # Assume base_oscs[0] is configured, setup the remaining base_osc voices.
- if self.base_oscs:
- base_osc = self.base_oscs[0]
- for other_base_osc in self.base_oscs[1:]:
- if self.pwm_osc in self.oscs_to_clone:
- self.oscs_to_clone.add(self.saw_osc)
- self.oscs_to_clone.add(self.sub_osc)
- self.oscs_to_clone.add(self.nse_osc)
- for osc in self.oscs_to_clone:
- proto_osc = base_osc + osc
- target_osc = other_base_osc + osc
- amy.send(osc=target_osc, clone_osc=proto_osc)
- # First 3 oscs must be chained to successors.
- if osc in [self.pwm_osc, self.saw_osc, self.sub_osc]:
- amy.send(osc=target_osc, chained_osc=target_osc + 1)
- self.oscs_to_clone = set()
-
- def get_new_voices(self, num_voices):
- """Setup a bunch of secondary voices."""
- note_objs = []
- for voice_num in range(num_voices):
- base_osc = self.next_osc
- self.next_osc += self.oscs_per_voice
- self.base_oscs.append(base_osc)
- note_objs.append(NoteObj(base_osc))
- self.init_AMY()
- return note_objs
-
  def set_patch(self, patch):
  self._init_from_patch_number(patch)
- print("New patch", patch, ":", self.name)
+ #print("New patch", patch, ":", self.name)
  self.init_AMY()
 
  def set_pitch_bend(self, value):
+ # Global, not osc-specific.
  amy.send(pitch_bend=value)
+
+ def amy_send(self, osc, **kwargs):
+ voices_str = None
+ if self.amy_voices:
+ voices_str = ','.join([str(v) for v in self.amy_voices])
+ amy.send(voices=voices_str, osc=osc, **kwargs)