neonbjb/tortoise-tts
1
0
Fork 0
mirror of https://github.com/neonbjb/tortoise-tts.git synced 2026-03-18 03:04:39 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity

Clarifies input and adds docstrings

Browse source
This commit is contained in:
TheMaakarov 2023-06-04 00:56:21 +02:00
parent 3d003c608d
commit c871a8f7f7
1 changed files with 83 additions and 25 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

108
scripts/recording_tools.py
View file

@ -8,21 +8,34 @@ import soundfile as sf
DURATION = 10.0
def countdown(count: int):
"""Create a countdown for specified number of seconds.
def countdown(count):
"""Create a countdown for specified number of seconds."""
for s_count in range(count):
print(f'{(count-s_count)/1000}', end='\r')
Args:
count (int): The amount of milliseconds to wait for.
"""
for elapsed_ms in range(count):
print(f'{(count-elapsed_ms)/1000}', end='\r')
time.sleep(0.001)
def record_until_keypress(output_folder, samplerate=22050, channels=1):
"""Record audio in chunks until interrupted, then chop into 10s fragments."""
def record_until_keypress(
output_folder: str,
samplerate: int=22050,
channels: int=1):
"""Record audio in chunks until interrupted, then chop into 10s fragments.
Args:
output_folder (str): Path to the voice folder where the chunks will be saved.
samplerate (int, optional): Target sample rate. Defaults to 22050.
channels (int, optional): The number of channels (1=mono, 2=stereo). Defaults to 1.
"""
recordings = []
print("·You are recording! Press Ctrl+C to stop.")
i=0
try:
while True:
recording = sd.rec(int(samplerate * DURATION), samplerate=samplerate, channels=channels, blocking=True)
recording =sd.rec(
int(samplerate * DURATION), samplerate=samplerate, channels=channels, blocking=True)
i+=1
print(f"Chunk #{i} was recorded")
recordings.append(recording)
@ -36,11 +49,24 @@ def record_until_keypress(output_folder, samplerate=22050, channels=1):
except Exception:
print(f"Error saving chunk #{i+1}.")
def record_audio(file_path, num_samples=1, samplerate=22050, channels=1, timeout=5000):
"""Record audio with the specified parameters."""
def record_audio(
file_path: str,
num_samples: int=3,
samplerate: int=22050,
channels: int=1,
timeout: int=5):
"""Record audio with the specified parameters.
Args:
file_path (str): Path to the voice folder where the samples will be saved.
num_samples (int, optional): The amount of samples to save. Defaults to 3.
samplerate (int, optional): Target sample rate. Defaults to 22050.
channels (int, optional): The number of channels (1=mono, 2=stereo). Defaults to 1.
timeout (int, optional): The seconds to wait between samples. Defaults to 5.
"""
for i in range(num_samples):
print(f"Preparing to record sample {i+1} in 5 seconds...")
countdown(timeout)
print(f"Preparing to record sample {i+1} in {timeout} seconds...")
countdown(timeout * 1000)
print("Recording...")
recording = sd.rec(int(samplerate * DURATION), samplerate=samplerate,
@ -51,17 +77,27 @@ def record_audio(file_path, num_samples=1, samplerate=22050, channels=1, timeout
print(f"Recording of sample {i+1} finished and saved as '{fname}'.")
def chop_audio(input_path, output_folder, convert=False, samplerate=22050):
"""Chop an audio file into chunks with specified duration."""
if convert:
def chop_audio(input_path: str,
output_folder: str,
no_conversion: bool=True,
samplerate: int=22050):
"""Chop an audio file into chunks with specified duration.
Args:
input_path (str): Path to the original voice sample.
output_folder (str): Path to the voice folder where the sample's chunks will be saved.
no_conversion (bool, optional): Ignore the conversion to WAV format. Defaults to True.
samplerate (int, optional): Target sample rate. Defaults to 22050.
"""
if no_conversion:
intermediate_path = input_path
else:
print("Loading file...")
audio = AudioSegment.from_file(input_path)
print("Converting to WAV...")
intermediate_path = "intermediate.wav"
audio.export(intermediate_path, format="wav")
else:
intermediate_path = input_path
print("Loading WAV file...")
data, og_samplerate = sf.read(intermediate_path)
@ -81,18 +117,40 @@ def chop_audio(input_path, output_folder, convert=False, samplerate=22050):
if __name__ == "__main__":
parser = argparse.ArgumentParser(description='Process some audio.')
parser.add_argument('command', choices=['record', 'chop', 'keypress'], help='Command to execute')
parser.add_argument('--file_path', help='Path to the file to process')
parser.add_argument('--output_folder', help='Folder to save the chunks')
parser.add_argument('--convert', action='store_true', default=False, help='Convert file to WAV format')
parser.add_argument('--num_samples', type=int, default=1, help='Number of samples to record')
parser.add_argument('--rec_timeout', type=int, default=5000, help='Milliseconds between recordings')
parser.add_argument(
'command',
choices=['record', 'chop', 'keypress'],
help='Command to execute')
parser.add_argument(
'--file_path',
help='Path to the file to process')
parser.add_argument(
'--output_folder',
help='Folder to save the chunks')
parser.add_argument(
'--no-convert',
action='store_false', default=True,
help='Convert file to WAV format')
parser.add_argument(
'--num_samples',
type=int, default=1,
help='Number of samples to record')
parser.add_argument(
'--rec_timeout',
type=int, default=5,
help='Seconds between recordings')
args = parser.parse_args()
if args.command == 'record':
record_audio(args.file_path, num_samples=args.num_samples, timeout=args.rec_timeout)
record_audio(
args.file_path,
num_samples=args.num_samples,
timeout=args.rec_timeout)
elif args.command == 'chop':
chop_audio(args.file_path, args.output_folder, convert=args.convert)
chop_audio(
args.file_path,
args.output_folder,
no_conversion=args.no_convert)
elif args.command == 'keypress':
record_until_keypress(args.file_path)
record_until_keypress(args.file_path)