"""
Sound Bank by Yui Kinomoto @arlez80
"""
const drum_track_bank = 128
const SoundFont = preload( "SoundFont.gd" )
# デフォルト
var default_mix_rate_table = [819,868,920,974,1032,1094,1159,1228,1301,1378,1460,1547,1639,1736,1840,1949,2065,2188,2318,2456,2602,2756,2920,3094,3278,3473,3679,3898,4130,4375,4635,4911,5203,5513,5840,6188,6556,6945,7358,7796,8259,8751,9271,9822,10406,11025,11681,12375,13111,13891,14717,15592,16519,17501,18542,19644,20812,22050,23361,24750,26222,27781,29433,31183,33038,35002,37084,39289,41625,44100,46722,49501,52444,55563,58866,62367,66075,70004,74167,78577,83250,88200,93445,99001,104888,111125,117733,124734,132151,140009,148334,157155,166499,176400,186889,198002,209776,222250,235466,249467,264301,280018,296668,314309,332999,352800,373779,396005,419552,444500,470932,498935,528603,560035,593337,628618,665998,705600,747557,792009,839105,889000,941863,997869,1057205,1120070,1186673,1257236]
var default_ads_state = [
{ "time": 0, "volume": 1.0 },
{ "time": 0.2, "volume": 0.95 },
]
var default_release_state = [
{ "time": 0, "volume": 0.95 },
{ "time": 0.01, "volume": 0.0 },
]
# 音色テーブル
var presets = {}
"""
楽器
"""
func create_preset( ):
var instruments = []
for i in range( 0, 128 ):
instruments.append( null )
return {
"name": "",
"number": 0,
"instruments": instruments,
"bags": [],
}
"""
ノート
"""
func create_instrument( ):
return {
"mix_rate": 44100,
"stream": null,
"ads_state": default_ads_state,
"release_state": default_release_state,
"preset": null,
# "assine_group": 0, # reserved
}
"""
再生周波数計算
"""
func calc_mix_rate( rate, center_key, target_key ):
return round( rate * pow( 2.0, ( target_key - center_key ) / 12.0 ) )
"""
追加
"""
func set_preset_sample( program_number, base_sample, base_mix_rate ):
var mix_rate_table = default_mix_rate_table
if base_mix_rate != 44100:
print( "not implemented" )
breakpoint
var preset = self.create_preset( )
preset.name = "#%03d" % program_number
preset.number = program_number
for i in range(0,128):
var inst = self.create_instrument( )
inst.mix_rate = mix_rate_table[i]
inst.stream = base_sample
inst.preset = preset
preset.instruments[i] = inst
self.set_preset( program_number, preset )
"""
追加
"""
func set_preset( program_number, preset ):
self.presets[program_number] = preset
"""
指定した楽器を取得
"""
func get_preset( program_number, bank = 0 ):
var pc = program_number | ( bank << 7 )
if not self.presets.has( pc ):
if drum_track_bank == bank:
pc = drum_track_bank
else:
pc = program_number
if not self.presets.has( pc ):
pc = self.presets.keys( )[0]
var preset = self.presets[pc]
return preset
"""
サウンドフォント読み込み
"""
func read_soundfont( sf, need_program_numbers = null ):
var sf_insts = self._read_soundfont_pdta_inst( sf )
var bag_index = 0
var gen_index = 0
for phdr_index in range( 0, len( sf.pdta.phdr )-1 ):
var phdr = sf.pdta.phdr[phdr_index]
var preset = self.create_preset( )
var program_number = phdr.preset | ( phdr.bank << 7 )
preset.name = phdr.name
preset.number = program_number
var bag_next = sf.pdta.phdr[phdr_index+1].preset_bag_index
var bag_count = bag_index
while bag_count < bag_next:
var gen_next = sf.pdta.pbag[bag_count+1].gen_ndx
var gen_count = gen_index
var bag = {
"preset": preset,
"coarse_tune": 0,
"fine_tune": 0,
"key_range": null,
"instrument": null,
"pan": 0.5,
}
while gen_count < gen_next:
var gen = sf.pdta.pgen[gen_count]
match gen.gen_oper:
SoundFont.gen_oper_coarse_tune:
bag.coarse_tune = gen.amount
SoundFont.gen_oper_fine_tune:
bag.fine_tune = gen.amount
SoundFont.gen_oper_key_range:
bag.key_range = {
"high": gen.uamount >> 8,
"low": gen.uamount & 0xFF,
}
SoundFont.gen_oper_pan:
bag.pan = ( gen.amount + 500 ) / 1000.0
SoundFont.gen_oper_instrument:
bag.instrument = sf_insts[gen.uamount]
gen_count += 1
if bag.instrument != null:
preset.bags.append( bag )
gen_index = gen_next
bag_count += 1
bag_index = bag_next
# 追加するか？
if need_program_numbers != null:
if not( program_number in need_program_numbers ) and not( phdr.preset in need_program_numbers ):
continue
# 追加
self._read_soundfont_preset_compose_sample( sf, preset )
self.set_preset( program_number, preset )
func _read_soundfont_preset_compose_sample( sf, preset ):
"""
var samples = []
var zero_4bytes = PoolIntArray( [ 0, 0 ] )
for bag in preset.bags[0].instrument.bags:
var sample = PoolIntArray( )
var start = bag.sample.start + bag.sample_start_offset
var end = bag.sample.end + bag.sample_end_offset
for i in range( 0, ( end - start ) / 2 ):
sample.append_array( zero_4bytes )
samples.append( sample )
"""
for pbag_index in range( 0, preset.bags.size( ) ):
var pbag = preset.bags[pbag_index]
for ibag_index in range( 0, pbag.instrument.bags.size( ) ):
var ibag = pbag.instrument.bags[ibag_index]
var start = ibag.sample.start + ibag.sample_start_offset
var end = ibag.sample.end + ibag.sample_end_offset
var start_loop = ibag.sample.start_loop + ibag.sample_start_loop_offset
var end_loop = ibag.sample.end_loop + ibag.sample_end_loop_offset
var mix_rate = ibag.sample.sample_rate * pow( 2.0, ( pbag.coarse_tune + ibag.coarse_tune ) / 12.0 ) * pow( 2.0, ( pbag.fine_tune + ibag.sample.pitch_correction + ibag.fine_tune ) / 1200.0 )
var ass = AudioStreamSample.new( )
ass.data = sf.sdta.smpl.subarray( start * 2, end * 2 )
ass.format = AudioStreamSample.FORMAT_16_BITS
ass.mix_rate = mix_rate
ass.stereo = false #bag.sample.sample_type != SoundFont.sample_link_mono_sample
ass.loop_begin = start_loop - start
ass.loop_end = end_loop - start
if ibag.sample_modes == SoundFont.sample_mode_no_loop or ibag.sample_modes == SoundFont.sample_mode_unused_no_loop:
ass.loop_mode = AudioStreamSample.LOOP_DISABLED
else:
ass.loop_mode = AudioStreamSample.LOOP_FORWARD
var key_range = ibag.key_range
if pbag.key_range != null:
key_range = pbag.key_range
for key_number in range( key_range.low, key_range.high + 1 ):
#if preset.number == drum_track_bank << 7:
# if 36 <= key_number and key_number <= 40:
# print( key_number, " # ", ibag.sample.name );
if preset.instruments[key_number] != null:
continue
var instrument = self.create_instrument( )
instrument.preset = preset
if ibag.original_key == 255:
instrument.mix_rate = ibag.sample.sample_rate
else:
instrument.mix_rate = self.calc_mix_rate( mix_rate, ibag.original_key, key_number )
instrument.stream = ass
var a = ibag.adsr.attack_vol_env_time
var d = ibag.adsr.decay_vol_env_time
var s = ibag.adsr.sustain_vol_env_level
var r = ibag.adsr.release_vol_env_time
instrument.ads_state = [
{ "time": 0, "volume": 0.0 },
{ "time": a, "volume": 1.0 },
{ "time": a+d, "volume": s },
]
instrument.release_state = [
{ "time": 0, "volume": s },
{ "time": r, "volume": 0.0 },
]
preset.instruments[key_number] = instrument
func _read_soundfont_pdta_inst( sf ):
var sf_insts = []
var bag_index = 0
var gen_index = 0
for inst_index in range(0, len( sf.pdta.inst ) - 1 ):
var inst = sf.pdta.inst[inst_index]
var sf_inst = {"name": inst.name, "bags": [] }
var bag_next = sf.pdta.inst[inst_index+1].inst_bag_ndx
var bag_count = bag_index
var global_bag = {}
while bag_count < bag_next:
var bag = {
"sample": null,
"sample_id": -1,
"sample_start_offset": 0,
"sample_end_offset": 0,
"sample_start_loop_offset": 0,
"sample_end_loop_offset": 0,
"coarse_tune": 0,
"fine_tune": 0,
"original_key": 255,
"keynum": 0,
"sample_modes": 0,
"key_range": { "high": 127, "low": 0 },
"vel_range": { "high": 127, "low": 0 },
"adsr": {
"attack_vol_env_time": 0.001,
"decay_vol_env_time": 0.001,
"sustain_vol_env_level": 1.0,
"release_vol_env_time": 0.001,
},
}
var gen_next = sf.pdta.ibag[bag_count+1].gen_ndx
var gen_count = gen_index
while gen_count < gen_next:
var gen = sf.pdta.igen[gen_count]
match gen.gen_oper:
SoundFont.gen_oper_key_range:
bag.key_range.high = gen.uamount >> 8
bag.key_range.low = gen.uamount & 0xFF
SoundFont.gen_oper_vel_range:
bag.vel_range.high = gen.uamount >> 8
bag.vel_range.low = gen.uamount & 0xFF
SoundFont.gen_oper_overriding_root_key:
bag.original_key = gen.amount
SoundFont.gen_oper_start_addrs_offset:
bag.sample_start_offset += gen.amount
SoundFont.gen_oper_end_addrs_offset:
bag.sample_end_offset += gen.amount
SoundFont.gen_oper_start_addrs_coarse_offset:
bag.sample_start_offset += gen.amount * 32768
SoundFont.gen_oper_end_addrs_coarse_offset:
bag.sample_end_offset += gen.amount * 32768
SoundFont.gen_oper_startloop_addrs_offset:
bag.sample_start_loop_offset += gen.amount
SoundFont.gen_oper_endloop_addrs_offset:
bag.sample_end_loop_offset += gen.amount
SoundFont.gen_oper_startloop_addrs_coarse_offset:
bag.sample_start_loop_offset += gen.amount * 32768
SoundFont.gen_oper_endloop_addrs_coarse_offset:
bag.sample_end_loop_offset += gen.amount * 32768
SoundFont.gen_oper_coarse_tune:
bag.coarse_tune = gen.amount
SoundFont.gen_oper_fine_tune:
bag.fine_tune = gen.amount
SoundFont.gen_oper_keynum:
bag.keynum = gen.amount
SoundFont.gen_oper_attack_vol_env:
bag.adsr.attack_vol_env_time = pow( 2.0, gen.amount / 1200.0 )
SoundFont.gen_oper_decay_vol_env:
bag.adsr.decay_vol_env_time = pow( 2.0, gen.amount / 1200.0 )
SoundFont.gen_oper_release_vol_env:
bag.adsr.release_vol_env_time = pow( 2.0, gen.amount / 1200.0 )
SoundFont.gen_oper_sustain_vol_env:
var s = min( max( 0, gen.amount ), 1440 )
bag.adsr.sustain_vol_env_level = ( 1440.0 - s ) / 1440.0
SoundFont.gen_oper_sample_modes:
bag.sample_modes = gen.uamount
SoundFont.gen_oper_sample_id:
bag.sample_id = gen.uamount
bag.sample = sf.pdta.shdr[gen.amount]
if bag.original_key == 255:
bag.original_key = bag.sample.original_key
#_:
# print( gen.gen_oper )
gen_count += 1
# global zoneでない場合
if bag.sample != null:
sf_inst.bags.append( bag )
else:
global_bag = bag
gen_index = gen_next
bag_count += 1
sf_insts.append( sf_inst )
bag_index = bag_next
return sf_insts