Skip to content

Commit

Permalink
Add multisample support to custom instruments in voice packs. Patch p…
Browse files Browse the repository at this point in the history 
…ermanent bank size based on new bank0
  • Loading branch information
@rrealmuto
rrealmuto committed Dec 29, 2024
1 parent 2a584d1 commit 050b8b7
Show file tree
Hide file tree
Showing 4 changed files with 85 additions and 42 deletions.
2 changes: 1 addition & 1 deletion ASM/src/hacks.asm
View file
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ Gameplay_InitSkybox:
.area 0x1C
lui at, hi(AUDIO_THREAD_INFO_MEM_START)
lw a0, lo(AUDIO_THREAD_INFO_MEM_START)(at)
jal 0x800B8654
jal 0x800B8654 ; AudioLoad_Init
lw a1, lo(AUDIO_THREAD_INFO_MEM_SIZE)(at)
lw ra, 0x0014(sp)
jr ra
Expand Down
4 changes: 4 additions & 0 deletions Audiobank.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -148,6 +148,9 @@ def __init__(self):
self.parents: list = []
self.bank_offset = -1 # offset of the sample within the bank. -1 indicates the sample hasn't been placed yet
self.original_offset = -1
self.codec: int = 0 # ADPCM is the only codec that seems to work
self.medium: int = 0
self.tag: bool = False
self.book: AdpcmBook = None
self.loop: AdpcmLoop = None
self.data: bytearray = None
Expand Down Expand Up @@ -443,6 +446,7 @@ def __init__(self, inst_id: int, bankdata: bytearray, audiotable_file: bytearray
self.lowNoteSample: Sample = Sample.from_rom_data(bankdata, audiotable_file, audiotable_index, self.lowNoteSampleOffset, audiotable_id, self, sampleCache, adpcmbookCache) if self.lowNoteSampleOffset != 0 else None
self.normalNoteSample: Sample = Sample.from_rom_data(bankdata, audiotable_file, audiotable_index, self.normalNoteSampleOffset, audiotable_id, self, sampleCache, adpcmbookCache) if self.normalNoteSampleOffset != 0 else None
self.highNoteSample: Sample = Sample.from_rom_data(bankdata, audiotable_file, audiotable_index, self.highNoteSampleOffset, audiotable_id, self, sampleCache, adpcmbookCache) if self.highNoteSampleOffset != 0 else None
self.tag: bool = False

def get_bytes(self):
bytes = bytearray(1)
Expand Down
12 changes: 12 additions & 0 deletions Rom.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -248,6 +248,18 @@ def rebuild_audio_data(self, audiobank_index_addr: int):
# Read audio banks back out and see if they match
test = []

# Update the size of bank 0 if necessary
# Read the sizes from the ROM
# 8010a1b8 and 8010a1bc contain uint32_t with the size of the init pool and the size of the permanent pool. The init pool contains the permanent pool so need to increase them both
initPoolAddress = 0x8010a1b8 - 0x800110A0 + 0xA87000
permanentPoolAddress = initPoolAddress + 4
initPoolSize = self.read_int32(initPoolAddress)
permanentPoolSize = self.read_int32(permanentPoolAddress)
initPoolSize = initPoolSize - 0x3AA0 + len(self.audiobanks[0].placed_data)
permanentPoolSize = permanentPoolSize - 0x3AA0 + len(self.audiobanks[0].placed_data)
self.write_int32(initPoolAddress, initPoolSize)
self.write_int32(permanentPoolAddress, permanentPoolSize)

# Read Audiotable index
audiotable_index_header: bytearray = self.read_bytes(AUDIOTABLE_INDEX_ADDR, 0x10)
audiotable_index_length = int.from_bytes(audiotable_index_header[0:2], 'big')
Expand Down
109 changes: 68 additions & 41 deletions Voices.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -433,7 +433,7 @@ def process_pak_sfx_by_id(pak_sfx_id: int, sfx_id_map, pak_sounds, age, settings
# Add the ones we have
for name, decompressed in pak_opts:
_file = io.BytesIO(decompressed)
soundData, numSampleFrames, sampleRate = process_sound_file(name, _file, age, settings, trim=True)
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(name, _file, age, settings, trim=True)
_file.close()
to_add.append((name, 0, rom_targets[i], soundData, numSampleFrames, sampleRate, None))
i += 1
Expand All @@ -446,7 +446,7 @@ def process_pak_sfx_by_id(pak_sfx_id: int, sfx_id_map, pak_sounds, age, settings
for i in range(0, len(rom_targets)):
name, decompressed = pak_opts[i]
_file = io.BytesIO(decompressed)
soundData, numSampleFrames, sampleRate = process_sound_file(name, _file, age, settings, trim=True)
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(name, _file, age, settings, trim=True)
_file.close()
to_add.append((name, 0, rom_targets[i], soundData, numSampleFrames, sampleRate, None))
pass
Expand All @@ -460,7 +460,7 @@ def process_pak_sfx_by_id(pak_sfx_id: int, sfx_id_map, pak_sounds, age, settings
patch = mapping['patch']
name, decompressed = pak_opts[0]
_file = io.BytesIO(decompressed)
soundData, numSampleFrames, sampleRate = process_sound_file(name, _file, age, settings, trim=True)
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(name, _file, age, settings, trim=True)
to_add.append((name, 0, rom_targets[0], soundData, numSampleFrames, sampleRate, patch))
_file.close()
else:
Expand Down Expand Up @@ -551,44 +551,71 @@ def patch_voice_pack(rom: Rom, age: VOICE_PACK_AGE, voice_pack: str, settings: S
sample_file = voice_map["direct_bank"][bank_str][index_str]
with zf.open(sample_file) as f:
# Read and process the file
soundData, numSampleFrames, sampleRate = process_sound_file(sample_file, f, age, settings)
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(sample_file, f, age, settings)
sfxs.append((sample_file, bank, index, soundData, numSampleFrames, sampleRate, None))
if "direct_bank_inst" in voice_map.keys():
for bank_str in voice_map["direct_bank_inst"].keys():
bank = int(bank_str, 16)
bank_index = int(bank_str, 16)
bank = rom.audiobanks[bank_index]
for index_str in voice_map["direct_bank_inst"][bank_str].keys():
index = int(index_str, 16)
sample_file = voice_map["direct_bank_inst"][bank_str][index_str]
with zf.open(sample_file) as f:
# Read the .aifc file
# Need to get the loop predictors out of it
soundData, numSampleFrames, sampleRate, book, loop = process_aifc_file(f)
inst_patch.append((sample_file,bank, index, soundData, numSampleFrames, sampleRate, book, loop))
instrument = bank.instruments[index]
instrument.tag = True
instrument_json: dict = voice_map["direct_bank_inst"][bank_str][index_str]
if "lowNote" in instrument_json.keys():
with zf.open(instrument_json["lowNote"]) as f:
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(f.name, f, age, settings)
# Pad the data to 16 bytes
soundData += bytearray((16 - (len(soundData)%16))%16)
tuning = sampleRate / 32000
tuning = tuning * instrument_json["lowTuning"]
lowSample = Sample()
lowSample.tag = True
instrument.lowNoteTuning = float(tuning)
lowSample.loop = loop
lowSample.book = book
lowSample.data = soundData
# Update sample data length = length
lowSample.size = len(soundData)
instrument.lowNoteSample = lowSample
if "normalNote" in instrument_json.keys():
with zf.open(instrument_json["normalNote"]) as f:
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(f.name, f, age, settings)
# Pad the data to 16 bytes
soundData += bytearray((16 - (len(soundData)%16))%16)
tuning = sampleRate / 32000
tuning = tuning * instrument_json["normalTuning"]
normalSample = Sample()
normalSample.tag = True
instrument.normalNoteTuning = float(tuning)
normalSample.loop = loop
normalSample.book = book
normalSample.data = soundData
# Update sample data length = length
normalSample.size = len(soundData)
instrument.normalNoteSample = normalSample
if "highNote" in instrument_json.keys():
with zf.open(instrument_json["normalNote"]) as f:
soundData, numSampleFrames, sampleRate, book, loop = process_sound_file(f.name, f, age, settings)
# Pad the data to 16 bytes
soundData += bytearray((16 - (len(soundData)%16))%16)
tuning = sampleRate / 32000
tuning = tuning * instrument_json["highTuning"]
highSample = Sample()
highSample.tag = True
instrument.highNoteSampleTuning = float(tuning)
highSample.loop = loop
highSample.book = book
highSample.data = soundData
# Update sample data length = length
highSample.size = len(soundData)
instrument.highNoteSample = highSample
if "normalRangeLow" in instrument_json.keys():
instrument.normalRangeLo = instrument_json["normalRangeLow"]
if "normalRangeHigh" in instrument_json.keys():
instrument.normalRangeHi = instrument_json["normalRangeHigh"]
zf.close()

sfx_data_start = len(rom.audiotable)

for _, bank_index, inst_id, soundData, numSampleFrames, sampleRate, book, loop in inst_patch:
# Calculate the tuning as sampling rate / 32000.
tuning = sampleRate / 32000

# Pad the data to 16 bytes
soundData += bytearray((16 - (len(soundData)%16))%16)

bank = rom.audiobanks[bank_index]

inst: Instrument = bank.instruments[inst_id]

# Update the sfx tuning
inst.normalNoteTuning = float(tuning)

# Update loop end as numSampleFrames
inst.normalNoteSample.loop = loop
inst.normalNoteSample.book = book
inst.normalNoteSample.data = soundData
# Update sample data length = length
inst.normalNoteSample.size = len(soundData)

# Patch each sfx that we have
for _, bank_index, sfx_id, soundData, numSampleFrames, sampleRate, patch in sfxs:
# Calculate the tuning as sampling rate / 32000.
Expand Down Expand Up @@ -633,19 +660,19 @@ def process_sound_file(file_name: str, file: BinaryIO, age: VOICE_PACK_AGE, sett
# Check if this is a file format that sf supports
filename, ext = os.path.splitext(file_name)
if ext.strip('.').upper() in sf.available_formats():
soundData, numSampleFrames, sampleRate = process_soundfile_file(file, age, settings, trim)
soundData, numSampleFrames, sampleRate, book, loop = process_soundfile_file(file, age, settings, trim)
elif ext == ".aifc":
soundData, numSampleFrames, sampleRate, book, loop = process_aifc_file(file)
elif ext == ".bin":
soundData, numSampleFrames, sampleRate = process_bin_file(file)
soundData, numSampleFrames, sampleRate, book, loop = process_bin_file(file)
else:
raise Exception(f"Unsupported file format {ext} in custom voice pack.")

return soundData, numSampleFrames, sampleRate
return soundData, numSampleFrames, sampleRate, book, loop


# Read an audio file using the soundfile python library
def process_soundfile_file(f: BinaryIO, age: VOICE_PACK_AGE, settings: Settings, trim=False) -> tuple[bytes, int, int]:
def process_soundfile_file(f: BinaryIO, age: VOICE_PACK_AGE, settings: Settings, trim=False) -> tuple[bytes, int, int, AdpcmBook, AdpcmLoop]:
data, sampleRate = sf.read(f)
if data.ndim == 2 and data.shape[1] == 2:
# Convert stereo to mono by averaging the two channels
Expand All @@ -661,15 +688,15 @@ def process_soundfile_file(f: BinaryIO, age: VOICE_PACK_AGE, settings: Settings,
frames = data.tobytes()
numSampleFrames = len(data)
soundData = adpcm_encode(frames, len(data)) # Encode the raw samples
return soundData, numSampleFrames, sampleRate
return soundData, numSampleFrames, sampleRate, None, None

# Used for patching SFX AIFC files that have already been stripped into raw binary ready to patch into the ROM
# Assume a vanilla sampling rate of 20000
def process_bin_file(f: BinaryIO) -> tuple[bytes, int, int]:
def process_bin_file(f: BinaryIO) -> tuple[bytes, int, int, AdpcmBook, AdpcmLoop]:
soundData = f.read()
numSampleFrames = int(len(soundData) * 16 / 9)
sampleRate = 20000
return (soundData, numSampleFrames, sampleRate)
return (soundData, numSampleFrames, sampleRate, None, None)

# Pretty basic aifc file parser. Extracts the already encoded .aifc data metadata from the file
def process_aifc_file(f: BinaryIO) -> tuple[bytes, int, int]:
Expand Down

0 comments on commit 050b8b7

Please sign in to comment.