OPENGIT
/
Retrieval-based-Voice-Conversion-WebUI
mirror of https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Add files via upload

This commit is contained in:
RVC-Boss 2023-05-28 23:40:54 +08:00 committed by GitHub
parent e8d92c3e91
commit 7bd25c4623
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 46 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
MDXNet.py
View File

@ -141,7 +141,7 @@ class Predictor:
# del self.model # del self.model
progress_bar.close() progress_bar.close()
return _sources return _sources
def prediction(self, m,vocal_root,others_root): def prediction(self, m,vocal_root,others_root,format):
os.makedirs(vocal_root,exist_ok=True) os.makedirs(vocal_root,exist_ok=True)
os.makedirs(others_root,exist_ok=True) os.makedirs(others_root,exist_ok=True)
basename = os.path.basename(m) basename = os.path.basename(m)
@ -151,8 +151,8 @@ class Predictor:
mix = mix.T mix = mix.T
sources = self.demix(mix.T) sources = self.demix(mix.T)
opt=sources[0].T opt=sources[0].T
sf.write("%s/%s_main_vocal.wav"%(vocal_root,basename), mix-opt, rate) sf.write("%s/%s_main_vocal.%s"%(vocal_root,basename,format), mix-opt, rate)
sf.write("%s/%s_others.wav"%(others_root,basename), opt , rate) sf.write("%s/%s_others.%s"%(others_root,basename,format), opt , rate)
class MDXNetDereverb(): class MDXNetDereverb():
def __init__(self,chunks): def __init__(self,chunks):
@ -167,8 +167,8 @@ class MDXNetDereverb():
self.denoise=True self.denoise=True
self.pred=Predictor(self) self.pred=Predictor(self)
def _path_audio_(self,input,vocal_root,others_root): def _path_audio_(self,input,vocal_root,others_root,format):
self.pred.prediction(input,vocal_root,others_root) self.pred.prediction(input,vocal_root,others_root,format)
if __name__ == '__main__': if __name__ == '__main__':
dereverb=MDXNetDereverb(15) dereverb=MDXNetDereverb(15)

35
infer-web.py
View File

@ -77,7 +77,7 @@ from infer_pack.models import (
SynthesizerTrnMs768NSFsid, SynthesizerTrnMs768NSFsid,
SynthesizerTrnMs768NSFsid_nono, SynthesizerTrnMs768NSFsid_nono,
) )
from scipy.io import wavfile import soundfile as sf
from fairseq import checkpoint_utils from fairseq import checkpoint_utils
import gradio as gr import gradio as gr
import logging import logging
@ -235,7 +235,8 @@ def vc_multi(
filter_radius, filter_radius,
resample_sr, resample_sr,
rms_mix_rate, rms_mix_rate,
protect protect,
format1
): ):
try: try:
dir_path = ( dir_path = (
@ -271,8 +272,8 @@ def vc_multi(
if "Success" in info: if "Success" in info:
try: try:
tgt_sr, audio_opt = opt tgt_sr, audio_opt = opt
wavfile.write( sf.write(
"%s/%s" % (opt_root, os.path.basename(path)), tgt_sr, audio_opt "%s/%s.%s" % (opt_root, os.path.basename(path),format1), audio_opt,tgt_sr
) )
except: except:
info += traceback.format_exc() info += traceback.format_exc()
@ -283,7 +284,7 @@ def vc_multi(
yield traceback.format_exc() yield traceback.format_exc()
def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg): def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg,format0):
infos = [] infos = []
try: try:
inp_root = inp_root.strip(" ").strip('"').strip("\n").strip('"').strip(" ") inp_root = inp_root.strip(" ").strip('"').strip("\n").strip('"').strip(" ")
@ -318,7 +319,7 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg):
and info["streams"][0]["sample_rate"] == "44100" and info["streams"][0]["sample_rate"] == "44100"
): ):
need_reformat = 0 need_reformat = 0
pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal) pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal,format0)
done = 1 done = 1
except: except:
need_reformat = 1 need_reformat = 1
@ -332,7 +333,7 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg):
inp_path = tmp_path inp_path = tmp_path
try: try:
if done == 0: if done == 0:
pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal) pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal,format0)
infos.append("%s->Success" % (os.path.basename(inp_path))) infos.append("%s->Success" % (os.path.basename(inp_path)))
yield "\n".join(infos) yield "\n".join(infos)
except: except:
@ -1341,6 +1342,12 @@ with gr.Blocks() as app:
file_count="multiple", label=i18n("也可批量输入音频文件, 二选一, 优先读文件夹") file_count="multiple", label=i18n("也可批量输入音频文件, 二选一, 优先读文件夹")
) )
with gr.Row(): with gr.Row():
format1= gr.Radio(
label=i18n("导出文件格式"),
choices=["wav", "flac","mp3","m4a"],
value="flac",
interactive=True,
)
but1 = gr.Button(i18n("转换"), variant="primary") but1 = gr.Button(i18n("转换"), variant="primary")
vc_output3 = gr.Textbox(label=i18n("输出信息")) vc_output3 = gr.Textbox(label=i18n("输出信息"))
but1.click( but1.click(
@ -1359,7 +1366,8 @@ with gr.Blocks() as app:
filter_radius1, filter_radius1,
resample_sr1, resample_sr1,
rms_mix_rate1, rms_mix_rate1,
protect1 protect1,
format1
], ],
[vc_output3], [vc_output3],
) )
@ -1402,9 +1410,15 @@ with gr.Blocks() as app:
visible=False, # 先不开放调整 visible=False, # 先不开放调整
) )
opt_vocal_root = gr.Textbox( opt_vocal_root = gr.Textbox(
label=i18n("指定输出人声文件夹"), value="opt" label=i18n("指定输出主人声文件夹"), value="opt"
)
opt_ins_root = gr.Textbox(label=i18n("指定输出非主人声文件夹"), value="opt")
format0= gr.Radio(
label=i18n("导出文件格式"),
choices=["wav", "flac","mp3","m4a"],
value="flac",
interactive=True,
) )
opt_ins_root = gr.Textbox(label=i18n("指定输出乐器文件夹"), value="opt")
but2 = gr.Button(i18n("转换"), variant="primary") but2 = gr.Button(i18n("转换"), variant="primary")
vc_output4 = gr.Textbox(label=i18n("输出信息")) vc_output4 = gr.Textbox(label=i18n("输出信息"))
but2.click( but2.click(
@ -1416,6 +1430,7 @@ with gr.Blocks() as app:
wav_inputs, wav_inputs,
opt_ins_root, opt_ins_root,
agg, agg,
format0
], ],
[vc_output4], [vc_output4],
) )

34
infer_uvr5.py
View File

@ -11,7 +11,7 @@ from tqdm import tqdm
from uvr5_pack.lib_v5 import spec_utils from uvr5_pack.lib_v5 import spec_utils
from uvr5_pack.utils import _get_name_params, inference from uvr5_pack.utils import _get_name_params, inference
from uvr5_pack.lib_v5.model_param_init import ModelParameters from uvr5_pack.lib_v5.model_param_init import ModelParameters
from scipy.io import wavfile import soundfile as sf
from uvr5_pack.lib_v5.nets_new import CascadedNet from uvr5_pack.lib_v5.nets_new import CascadedNet
from uvr5_pack.lib_v5 import nets_61968KB as nets from uvr5_pack.lib_v5 import nets_61968KB as nets
@ -41,7 +41,7 @@ class _audio_pre_:
self.mp = mp self.mp = mp
self.model = model self.model = model
def _path_audio_(self, music_file, ins_root=None, vocal_root=None): def _path_audio_(self, music_file, ins_root=None, vocal_root=None,format="flac"):
if ins_root is None and vocal_root is None: if ins_root is None and vocal_root is None:
return "No save root." return "No save root."
name = os.path.basename(music_file) name = os.path.basename(music_file)
@ -120,12 +120,11 @@ class _audio_pre_:
else: else:
wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp) wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp)
print("%s instruments done" % name) print("%s instruments done" % name)
wavfile.write( sf.write(
os.path.join( os.path.join(
ins_root, "instrument_{}_{}.wav".format(name, self.data["agg"]) ins_root, "instrument_{}_{}.{}".format(name, self.data["agg"],format)
), ),
self.mp.param["sr"], (np.array(wav_instrument) * 32768).astype("int16"), self.mp.param["sr"],
(np.array(wav_instrument) * 32768).astype("int16"),
) # ) #
if vocal_root is not None: if vocal_root is not None:
if self.data["high_end_process"].startswith("mirroring"): if self.data["high_end_process"].startswith("mirroring"):
@ -138,12 +137,11 @@ class _audio_pre_:
else: else:
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp) wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
print("%s vocals done" % name) print("%s vocals done" % name)
wavfile.write( sf.write(
os.path.join( os.path.join(
vocal_root, "vocal_{}_{}.wav".format(name, self.data["agg"]) vocal_root, "vocal_{}_{}.{}".format(name, self.data["agg"],format)
), ),
self.mp.param["sr"], (np.array(wav_vocals) * 32768).astype("int16"), self.mp.param["sr"],
(np.array(wav_vocals) * 32768).astype("int16"),
) )
class _audio_pre_new: class _audio_pre_new:
@ -173,7 +171,7 @@ class _audio_pre_new:
self.mp = mp self.mp = mp
self.model = model self.model = model
def _path_audio_(self, music_file, vocal_root=None, ins_root=None):#3个VR模型vocal和ins是反的 def _path_audio_(self, music_file, vocal_root=None, ins_root=None,format="flac"):#3个VR模型vocal和ins是反的
if ins_root is None and vocal_root is None: if ins_root is None and vocal_root is None:
return "No save root." return "No save root."
name = os.path.basename(music_file) name = os.path.basename(music_file)
@ -252,12 +250,11 @@ class _audio_pre_new:
else: else:
wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp) wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp)
print("%s instruments done" % name) print("%s instruments done" % name)
wavfile.write( sf.write(
os.path.join( os.path.join(
ins_root, "main_vocal_{}_{}.wav".format(name, self.data["agg"]) ins_root, "main_vocal_{}_{}.{}".format(name, self.data["agg"],format)
), ),
self.mp.param["sr"], (np.array(wav_instrument) * 32768).astype("int16"),self.mp.param["sr"],
(np.array(wav_instrument) * 32768).astype("int16"),
) # ) #
if vocal_root is not None: if vocal_root is not None:
if self.data["high_end_process"].startswith("mirroring"): if self.data["high_end_process"].startswith("mirroring"):
@ -270,12 +267,11 @@ class _audio_pre_new:
else: else:
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp) wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
print("%s vocals done" % name) print("%s vocals done" % name)
wavfile.write( sf.write(
os.path.join( os.path.join(
vocal_root, "others_{}_{}.wav".format(name, self.data["agg"]) vocal_root, "others_{}_{}.{}".format(name, self.data["agg"],format)
), ),
self.mp.param["sr"], (np.array(wav_vocals) * 32768).astype("int16"),self.mp.param["sr"],
(np.array(wav_vocals) * 32768).astype("int16"),
) )

2
requirements.txt
View File

@ -40,4 +40,4 @@ colorama>=0.4.5
pyworld>=0.3.2 pyworld>=0.3.2
httpx==0.23.0 httpx==0.23.0
onnxruntime-gpu onnxruntime-gpu
torchcrepe torchcrepe