mirror of
https://github.com/storytold/storyteller-ml.git
synced 2026-10-12 23:48:57 +00:00
Michael Chung
94 lines
3.6 KiB
Python
94 lines
3.6 KiB
Python
import pathlib
|
|
import numpy as np
|
|
input_path = pathlib.Path.cwd() / pathlib.Path("..") / pathlib.Path("Vall-E-mount") / pathlib.Path("input")
|
|
models_path = pathlib.Path.cwd() / pathlib.Path("..") / pathlib.Path("Vall-E-mount") / pathlib.Path("models")
|
|
presets_path = pathlib.Path.cwd() / pathlib.Path("..") / pathlib.Path("Vall-E-mount") / pathlib.Path("presets")
|
|
prompts_path = pathlib.Path.cwd() / pathlib.Path("..") / pathlib.Path("Vall-E-mount") / pathlib.Path("prompts")
|
|
temp_path = pathlib.Path.cwd() / pathlib.Path("..") / pathlib.Path("Vall-E-mount") / pathlib.Path("temp")
|
|
import soundfile as sf
|
|
import torchaudio
|
|
import torchaudio.transforms as T
|
|
import torch
|
|
|
|
|
|
def stitch_wav_tensors_with_crossfade(wav_tensor1,wav_tensor2,sample_rate:int):
|
|
# Load two audio signals as NumPy tensors
|
|
|
|
# Compute the length of the cross-fade window in samples
|
|
fade_len = int(0.01 * sample_rate)
|
|
|
|
# Create a cross-fade window
|
|
fade_window = np.hanning(2*fade_len)
|
|
|
|
# Concatenate the signals with cross-fade
|
|
audio_out = np.concatenate((wav_tensor1[:-fade_len],
|
|
wav_tensor1[-fade_len:] * fade_window[:fade_len] + wav_tensor2[:fade_len] * fade_window[fade_len:],
|
|
wav_tensor2[fade_len:]))
|
|
|
|
sf.write("audio_out.wav", audio_out, sample_rate)
|
|
return audio_out
|
|
|
|
|
|
def trim_audio_range(file_path, start_seconds, end_seconds, output_path="range_trimmed_audio.wav"):
|
|
"""
|
|
Trims an audio file to a specified range and saves the trimmed audio.
|
|
|
|
Args:
|
|
- file_path (str): Path to the audio file to be trimmed.
|
|
- start_seconds (float): Start time of the desired audio segment in seconds.
|
|
- end_seconds (float): End time of the desired audio segment in seconds.
|
|
- output_path (str): Path to save the trimmed audio. Defaults to "range_trimmed_audio.wav".
|
|
|
|
Returns:
|
|
- None
|
|
"""
|
|
# Load the audio file
|
|
waveform, sample_rate = torchaudio.load(file_path)
|
|
|
|
# Convert the range to sample indices
|
|
start_sample = int(start_seconds * sample_rate)
|
|
end_sample = int(end_seconds * sample_rate)
|
|
|
|
# Trim the audio to the specified range
|
|
trimmed_waveform = waveform[:, start_sample:end_sample]
|
|
|
|
# Save the trimmed audio
|
|
torchaudio.save(output_path, trimmed_waveform, sample_rate)
|
|
|
|
|
|
def cut_audio_range(file_path, start_seconds, end_seconds, output_path="cut_audio.wav"):
|
|
"""
|
|
Cuts out a specific segment from an audio file and saves the result.
|
|
|
|
Args:
|
|
- file_path (str): Path to the audio file from which the segment will be cut.
|
|
- start_seconds (float): Start time of the segment to be cut out in seconds.
|
|
- end_seconds (float): End time of the segment to be cut out in seconds.
|
|
- output_path (str): Path to save the resulting audio. Defaults to "cut_audio.wav".
|
|
|
|
Returns:
|
|
- None
|
|
"""
|
|
# Load the audio file
|
|
waveform, sample_rate = torchaudio.load(file_path)
|
|
|
|
# Convert the range to sample indices
|
|
start_sample = int(start_seconds * sample_rate)
|
|
end_sample = int(end_seconds * sample_rate)
|
|
|
|
# Cut out the specified range
|
|
cut_waveform = torch.cat([waveform[:, :start_sample], waveform[:, end_sample:]], dim=1)
|
|
|
|
# Save the resulting audio
|
|
torchaudio.save(output_path, cut_waveform, sample_rate)
|
|
|
|
def test():
|
|
try:
|
|
path = pathlib.Path("/home/tensor/code/TTSDockerContainer/VALL-E-X/../Vall-E-mount/temp/audio_concat_a521f31b-baf3-49a3-aca6-924b5191aad3.wav")
|
|
wav, sr = torchaudio.load(path)
|
|
t = wav[:,int(5 * sr):int(15 * sr)]
|
|
torchaudio.save(pathlib.Path("./testfull.wav"),t,sr)
|
|
except Exception as e:
|
|
print(e)
|
|
|
|
test() |