forked from R3gm/SoniTranslate
-
Notifications
You must be signed in to change notification settings - Fork 0
/
app_rvc.py
2323 lines (2130 loc) · 92.5 KB
/
app_rvc.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
import gradio as gr
from soni_translate.logging_setup import (
logger,
set_logging_level,
configure_logging_libs,
); configure_logging_libs() # noqa
import whisperx
import torch
import os
from soni_translate.audio_segments import create_translated_audio
from soni_translate.text_to_speech import (
audio_segmentation_to_voice,
edge_tts_voices_list,
coqui_xtts_voices_list,
piper_tts_voices_list,
create_wav_file_vc,
accelerate_segments,
)
from soni_translate.translate_segments import translate_text
from soni_translate.preprocessor import (
audio_video_preprocessor,
audio_preprocessor,
)
from soni_translate.language_configuration import (
LANGUAGES,
LANGUAGES_LIST,
bark_voices_list,
vits_voices_list,
)
from soni_translate.utils import (
remove_files,
download_list,
upload_model_list,
download_manager,
run_command,
is_audio_file,
copy_files,
)
from soni_translate.mdx_net import (
UVR_MODELS,
MDX_DOWNLOAD_LINK,
mdxnet_models_dir,
)
from soni_translate.speech_segmentation import (
transcribe_speech,
align_speech,
diarize_speech,
diarization_models,
)
from soni_translate.text_multiformat_processor import (
srt_file_to_segments,
document_preprocessor,
determine_chunk_size,
plain_text_to_segments,
segments_to_plain_text,
process_subtitles,
break_aling_segments,
)
from soni_translate.languages_gui import language_data, news
import copy
import logging
import json
from pydub import AudioSegment
from voice_main import ClassVoices
import argparse
import time
import hashlib
# Custom voice
directories = [
"downloads",
"logs",
"weights",
"clean_song_output",
"_XTTS_",
f"audio2{os.sep}audio",
"audio",
]
[
os.makedirs(directory)
for directory in directories
if not os.path.exists(directory)
]
class TTS_Info:
def __init__(self, piper_enabled, xtts_enabled):
self.list_edge = edge_tts_voices_list()
self.list_bark = list(bark_voices_list.keys())
self.list_vits = list(vits_voices_list.keys())
self.piper_enabled = piper_enabled
self.list_vits_onnx = (
piper_tts_voices_list() if self.piper_enabled else []
)
self.xtts_enabled = xtts_enabled
def tts_list(self):
self.list_coqui_xtts = (
coqui_xtts_voices_list() if self.xtts_enabled else []
)
list_tts = sorted(
self.list_edge
+ self.list_bark
+ self.list_vits
+ self.list_vits_onnx
+ self.list_coqui_xtts
)
return list_tts
def custom_model_voice_enable(enable_custom_voice):
os.environ["VOICES_MODELS"] = (
"ENABLE" if enable_custom_voice else "DISABLE"
)
def prog_disp(msg, percent, is_gui, progress=None):
logger.info(msg)
if is_gui:
progress(percent, desc=msg)
def warn_disp(wrn_lang, is_gui):
logger.warning(wrn_lang)
if is_gui:
gr.Warning(wrn_lang)
class SoniTrCache:
def __init__(self):
self.cache = {
'media': [[]],
'transcript_align': [],
'break_align': [],
'diarize': [],
'translate': [],
'subs_and_edit': [],
'tts': [],
'acc_and_vc': [],
'mix_aud': [],
'output': []
}
self.cache_data = {
'media': [],
'transcript_align': [],
'break_align': [],
'diarize': [],
'translate': [],
'subs_and_edit': [],
'tts': [],
'acc_and_vc': [],
'mix_aud': [],
'output': []
}
self.cache_keys = list(self.cache.keys())
self.first_task = self.cache_keys[0]
self.last_task = self.cache_keys[-1]
self.pre_step = None
self.pre_params = []
def set_variable(self, variable_name, value):
setattr(self, variable_name, value)
def task_in_cache(self, step: str, params: list, previous_step_data: dict):
self.pre_step_cache = None
if step == self.first_task:
self.pre_step = None
if self.pre_step:
self.cache[self.pre_step] = self.pre_params
# Fill data in cache
self.cache_data[self.pre_step] = copy.deepcopy(previous_step_data)
self.pre_params = params
# logger.debug(f"Step: {str(step)}, Cache params: {str(self.cache)}")
if params == self.cache[step]:
logger.debug(f"In cache: {str(step)}")
# Set the var needed for next step
# Recovery from cache_data the current step
for key, value in self.cache_data[step].items():
self.set_variable(key, copy.deepcopy(value))
logger.debug(
f"Chache load: {str(key)}"
)
self.pre_step = step
return True
else:
logger.debug(f"Flush next and caching {str(step)}")
selected_index = self.cache_keys.index(step)
for idx, key in enumerate(self.cache.keys()):
if idx >= selected_index:
self.cache[key] = []
self.cache_data[key] = {}
# The last is now previous
self.pre_step = step
return False
def clear_cache(self, media, force=False):
if media != self.cache["media"][0] or force:
# Clear cache
self.cache = {key: [] for key in self.cache}
self.cache["media"] = [[]]
logger.info("Cache flushed")
def get_hash(filepath):
with open(filepath, 'rb') as f:
file_hash = hashlib.blake2b()
while chunk := f.read(8192):
file_hash.update(chunk)
return file_hash.hexdigest()[:18]
class SoniTranslate(SoniTrCache):
def __init__(self, dev=False):
super().__init__()
self.device = "cuda" if torch.cuda.is_available() else "cpu"
self.result_diarize = None
self.align_language = None
self.result_source_lang = None
self.edit_subs_complete = False
self.voiceless_id = None
self.burn_subs_id = None
logger.info(f"Working in: {self.device}")
def multilingual_media_conversion(
self,
media_file,
link_media,
directory_input,
YOUR_HF_TOKEN,
preview=False,
WHISPER_MODEL_SIZE="large-v3",
batch_size=16,
compute_type="float16",
SOURCE_LANGUAGE="Automatic detection",
TRANSLATE_AUDIO_TO="English (en)",
min_speakers=1,
max_speakers=2,
tts_voice00="en-AU-WilliamNeural-Male",
tts_voice01="en-CA-ClaraNeural-Female",
tts_voice02="en-GB-ThomasNeural-Male",
tts_voice03="en-GB-SoniaNeural-Female",
tts_voice04="en-NZ-MitchellNeural-Male",
tts_voice05="en-GB-MaisieNeural-Female",
video_output_name="video_dub",
AUDIO_MIX_METHOD="Adjusting volumes and mixing audio",
max_accelerate_audio=2.1,
acceleration_rate_regulation=False,
volume_original_audio=0.25,
volume_translated_audio=1.80,
output_format_subtitle="srt",
get_translated_text=False,
get_video_from_text_json=False,
text_json="{}",
diarization_model="pyannote_2.1",
translate_process="google_translator_batch",
subtitle_file=None,
output_type="video",
voiceless_track=False,
voice_imitation=False,
voice_imitation_max_segments=3,
voice_imitation_vocals_dereverb=False,
voice_imitation_remove_previous=True,
voice_imitation_method="freevc",
dereverb_automatic_xtts=True,
divide_text_segments_by="",
burn_subtitles_to_video=False,
enable_cache=False,
is_gui=False,
progress=gr.Progress(),
):
if not YOUR_HF_TOKEN:
YOUR_HF_TOKEN = os.getenv("YOUR_HF_TOKEN")
if not YOUR_HF_TOKEN:
raise ValueError("No valid Hugging Face token")
else:
os.environ["YOUR_HF_TOKEN"] = YOUR_HF_TOKEN
if get_translated_text:
self.edit_subs_complete = False
if get_video_from_text_json:
if not self.edit_subs_complete:
raise ValueError("Generate the transcription first.")
TRANSLATE_AUDIO_TO = LANGUAGES[TRANSLATE_AUDIO_TO]
SOURCE_LANGUAGE = LANGUAGES[SOURCE_LANGUAGE]
if tts_voice00[:2].lower() != TRANSLATE_AUDIO_TO[:2].lower():
wrn_lang = (
"Make sure to select a 'TTS Speaker' suitable for"
" the translation language to avoid errors with the TTS."
)
warn_disp(wrn_lang, is_gui)
if "_XTTS_" in tts_voice00 and voice_imitation:
wrn_lang = (
"When you select XTTS, it is advisable "
"to disable Voice Imitation."
)
warn_disp(wrn_lang, is_gui)
if os.getenv("VOICES_MODELS") == "ENABLE" and voice_imitation:
wrn_lang = (
"When you use R.V.C. models, it is advisable"
" to disable Voice Imitation."
)
warn_disp(wrn_lang, is_gui)
if media_file is None:
media_file = (
directory_input
if os.path.exists(directory_input)
else link_media
)
media_file = (
media_file if isinstance(media_file, str) else media_file.name
)
if not media_file and not subtitle_file:
raise ValueError(
"Specifify a media or SRT file in advanced settings"
)
if subtitle_file:
subtitle_file = (
subtitle_file
if isinstance(subtitle_file, str)
else subtitle_file.name
)
if not media_file and subtitle_file:
media_file = "audio_support.wav"
if not get_video_from_text_json:
remove_files(media_file)
srt_data = srt_file_to_segments(subtitle_file)
total_duration = srt_data["segments"][-1]["end"] + 30.
support_audio = AudioSegment.silent(
duration=int(total_duration * 1000)
)
support_audio.export(
media_file, format="wav"
)
logger.info("Supporting audio for the SRT file, created.")
if "SET_LIMIT" == os.getenv("DEMO"):
preview = True
AUDIO_MIX_METHOD = "Adjusting volumes and mixing audio"
WHISPER_MODEL_SIZE = "medium"
logger.info(
"DEMO; set preview=True; Generation is limited to "
"10 seconds to prevent CPU errors. No limitations with GPU.\n"
"DEMO; set Adjusting volumes and mixing audio\n"
"DEMO; set whisper model to medium"
)
# Check GPU
compute_type = "float32" if self.device == "cpu" else compute_type
base_video_file = "Video.mp4"
base_audio_wav = "audio.wav"
dub_audio_file = "audio_dub_solo.ogg"
voiceless_audio_file = "audio_Voiceless.wav"
mix_audio_file = "audio_mix.mp3"
vid_subs = "video_subs_file.mp4"
video_output_file = f"{video_output_name}.mp4"
if os.path.exists(media_file):
media_base_hash = get_hash(media_file)
else:
media_base_hash = media_file
self.clear_cache(media_base_hash, force=(not enable_cache))
if not get_video_from_text_json:
self.result_diarize = (
self.align_language
) = self.result_source_lang = None
if not self.task_in_cache("media", [media_base_hash, preview], {}):
if is_audio_file(media_file):
prog_disp(
"Processing audio...", 0.15, is_gui, progress=progress
)
audio_preprocessor(preview, media_file, base_audio_wav)
else:
prog_disp(
"Processing video...", 0.15, is_gui, progress=progress
)
audio_video_preprocessor(
preview, media_file, base_video_file, base_audio_wav
)
logger.debug("Set file complete.")
if not self.task_in_cache("transcript_align", [
subtitle_file,
SOURCE_LANGUAGE,
WHISPER_MODEL_SIZE,
compute_type,
batch_size
], {}):
if subtitle_file:
prog_disp(
"From SRT file...", 0.30, is_gui, progress=progress
)
if SOURCE_LANGUAGE == "Automatic detection":
raise Exception(
"To use an SRT file, you need to specify its "
"original language (Source language)"
)
audio = whisperx.load_audio(base_audio_wav)
self.result = srt_file_to_segments(subtitle_file)
self.result["language"] = SOURCE_LANGUAGE
else:
prog_disp(
"Transcribing...", 0.30, is_gui, progress=progress
)
SOURCE_LANGUAGE = (
None
if SOURCE_LANGUAGE == "Automatic detection"
else SOURCE_LANGUAGE
)
audio, self.result = transcribe_speech(
base_audio_wav,
WHISPER_MODEL_SIZE,
compute_type,
batch_size,
SOURCE_LANGUAGE,
)
logger.debug(
"Transcript complete, "
f"segments count {len(self.result['segments'])}"
)
self.align_language = self.result["language"]
if not subtitle_file:
prog_disp("Aligning...", 0.45, is_gui, progress=progress)
self.result = align_speech(audio, self.result)
logger.debug(
"Align complete, "
f"segments count {len(self.result['segments'])}"
)
if self.result["segments"] == []:
raise ValueError("No active speech found in audio")
if not self.task_in_cache("break_align", [
divide_text_segments_by,
self.align_language
], {
"result": self.result,
"align_language": self.align_language
}):
if self.align_language in ["ja", "zh"]:
divide_text_segments_by += "|!|?|...|。"
if divide_text_segments_by:
try:
self.result = break_aling_segments(
self.result,
break_characters=divide_text_segments_by,
)
except Exception as error:
logger.error(str(error))
if not self.task_in_cache("diarize", [
min_speakers,
max_speakers,
YOUR_HF_TOKEN[:len(YOUR_HF_TOKEN)//2],
diarization_model
], {
"result": self.result
}):
prog_disp("Diarizing...", 0.60, is_gui, progress=progress)
diarize_model_select = diarization_models[diarization_model]
self.result_diarize = diarize_speech(
base_audio_wav,
self.result,
min_speakers,
max_speakers,
YOUR_HF_TOKEN,
diarize_model_select,
)
logger.debug("Diarize complete")
self.result_source_lang = copy.deepcopy(self.result_diarize)
if not self.task_in_cache("translate", [
TRANSLATE_AUDIO_TO,
translate_process
], {
"result_diarize": self.result_diarize
}):
prog_disp("Translating...", 0.70, is_gui, progress=progress)
self.result_diarize["segments"] = translate_text(
self.result_diarize["segments"],
TRANSLATE_AUDIO_TO,
translate_process,
chunk_size=1800,
)
logger.debug("Translation complete")
logger.debug(self.result_diarize)
if get_translated_text:
json_data = []
for segment in self.result_diarize["segments"]:
start = segment["start"]
text = segment["text"]
speaker = int(segment.get("speaker", "SPEAKER_00")[-1]) + 1
json_data.append(
{"start": start, "text": text, "speaker": speaker}
)
# Convert list of dictionaries to a JSON string with indentation
json_string = json.dumps(json_data, indent=2)
logger.info("Done")
self.edit_subs_complete = True
return json_string.encode().decode("unicode_escape")
if get_video_from_text_json:
if self.result_diarize is None:
raise ValueError("Generate the transcription first.")
# with open('text_json.json', 'r') as file:
text_json_loaded = json.loads(text_json)
for i, segment in enumerate(self.result_diarize["segments"]):
segment["text"] = text_json_loaded[i]["text"]
segment["speaker"] = "SPEAKER_0" + str(
int(text_json_loaded[i]["speaker"]) - 1
)
# Write subtitle
if not self.task_in_cache("subs_and_edit", [
copy.deepcopy(self.result_diarize),
output_format_subtitle,
TRANSLATE_AUDIO_TO
], {
"result_diarize": self.result_diarize
}):
self.sub_file = process_subtitles(
self.result_source_lang,
self.align_language,
self.result_diarize,
output_format_subtitle,
TRANSLATE_AUDIO_TO,
)
if output_type == "subtitle":
logger.info(f"Done: {self.sub_file}")
return self.sub_file
if not self.task_in_cache("tts", [
TRANSLATE_AUDIO_TO,
tts_voice00,
tts_voice01,
tts_voice02,
tts_voice03,
tts_voice04,
tts_voice05,
dereverb_automatic_xtts
], {
"sub_file": self.sub_file
}):
prog_disp("Text to speech...", 0.80, is_gui, progress=progress)
self.valid_speakers = audio_segmentation_to_voice(
self.result_diarize,
TRANSLATE_AUDIO_TO,
True,
tts_voice00,
tts_voice01,
tts_voice02,
tts_voice03,
tts_voice04,
tts_voice05,
dereverb_automatic_xtts,
)
if not hasattr(vci, 'model_voice_path00'):
cc_transpose_values = cc_index_values = cc_model_paths = None
else:
cc_model_paths = [
vci.model_voice_path00,
vci.model_voice_path01,
vci.model_voice_path02,
vci.model_voice_path03,
vci.model_voice_path04,
vci.model_voice_path05,
vci.model_voice_path99
]
cc_index_values = [
vci.file_index200,
vci.file_index201,
vci.file_index202,
vci.file_index203,
vci.file_index204,
vci.file_index205,
vci.file_index299
]
cc_transpose_values = [
vci.f0method,
vci.transpose00,
vci.transpose01,
vci.transpose02,
vci.transpose03,
vci.transpose04,
vci.transpose05,
vci.transpose99
]
if not self.task_in_cache("acc_and_vc", [
max_accelerate_audio,
acceleration_rate_regulation,
voice_imitation,
voice_imitation_max_segments,
voice_imitation_remove_previous,
voice_imitation_vocals_dereverb,
voice_imitation_method,
os.getenv("VOICES_MODELS"),
cc_model_paths,
cc_index_values,
cc_transpose_values
], {
"valid_speakers": self.valid_speakers
}):
audio_files, speakers_list = accelerate_segments(
self.result_diarize,
max_accelerate_audio,
self.valid_speakers,
acceleration_rate_regulation,
)
# Voice Imitation (Tone color converter)
if voice_imitation:
prog_disp(
"Voice Imitation...", 0.85, is_gui, progress=progress
)
from soni_translate.text_to_speech import toneconverter
try:
toneconverter(
copy.deepcopy(self.result_diarize),
preprocessor_max_segments=voice_imitation_max_segments,
remove_previous_process=voice_imitation_remove_previous,
get_vocals_dereverb=voice_imitation_vocals_dereverb,
method_vc=voice_imitation_method,
)
except Exception as error:
logger.error(str(error))
# custom voice
if os.getenv("VOICES_MODELS") == "ENABLE":
prog_disp(
"Applying customized voices...",
0.90,
is_gui,
progress=progress,
)
if cc_model_paths is None:
logger.error("Apply the configuration!")
try:
vci(speakers_list, audio_files)
except Exception as error:
logger.error(str(error))
prog_disp(
"Creating final translated video...",
0.95,
is_gui,
progress=progress,
)
remove_files(dub_audio_file)
create_translated_audio(
self.result_diarize, audio_files, dub_audio_file
)
# Voiceless track, change with file
hash_base_audio_wav = get_hash(base_audio_wav)
if voiceless_track:
if self.voiceless_id != hash_base_audio_wav:
from soni_translate.mdx_net import process_uvr_task
try:
# voiceless_audio_file_dir = "clean_song_output/voiceless"
remove_files(voiceless_audio_file)
uvr_voiceless_audio_wav, _ = process_uvr_task(
orig_song_path=base_audio_wav,
song_id="voiceless",
only_voiceless=True,
remove_files_output_dir=False,
)
copy_files(uvr_voiceless_audio_wav, ".")
base_audio_wav = voiceless_audio_file
self.voiceless_id = hash_base_audio_wav
except Exception as error:
logger.error(str(error))
else:
base_audio_wav = voiceless_audio_file
if not self.task_in_cache("mix_aud", [
AUDIO_MIX_METHOD,
volume_original_audio,
volume_translated_audio,
voiceless_track
], {}):
# TYPE MIX AUDIO
remove_files(mix_audio_file)
command_volume_mix = f'ffmpeg -y -i {base_audio_wav} -i {dub_audio_file} -filter_complex "[0:0]volume={volume_original_audio}[a];[1:0]volume={volume_translated_audio}[b];[a][b]amix=inputs=2:duration=longest" -c:a libmp3lame {mix_audio_file}'
command_background_mix = f'ffmpeg -i {base_audio_wav} -i {dub_audio_file} -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress=threshold=0.003:ratio=20[bg]; [bg][mix]amerge[final]" -map [final] {mix_audio_file}'
if AUDIO_MIX_METHOD == "Adjusting volumes and mixing audio":
# volume mix
run_command(command_volume_mix)
else:
try:
# background mix
run_command(command_background_mix)
except Exception as error_mix:
# volume mix except
logger.error(str(error_mix))
run_command(command_volume_mix)
if output_type == "audio" or is_audio_file(media_file):
logger.info(f"Done: {mix_audio_file}")
return mix_audio_file
hash_base_video_file = get_hash(base_video_file)
if burn_subtitles_to_video:
hashvideo_text = [hash_base_video_file, [seg["text"] for seg in self.result_diarize["segments"]]]
if self.burn_subs_id != hashvideo_text:
try:
logger.info("Burn subtitles")
remove_files(vid_subs)
command = f"ffmpeg -i {base_video_file} -y -vf subtitles={self.sub_file} {vid_subs}"
run_command(command)
base_video_file = vid_subs
self.burn_subs_id = hashvideo_text
except Exception as error:
logger.error(str(error))
else:
base_video_file = vid_subs
if not self.task_in_cache("output", [
hash_base_video_file,
hash_base_audio_wav,
burn_subtitles_to_video
], {}):
# Merge new audio + video
remove_files(video_output_file)
run_command(
f"ffmpeg -i {base_video_file} -i {mix_audio_file} -c:v copy -c:a copy -map 0:v -map 1:a -shortest {video_output_file}"
)
logger.info(f"Done: {video_output_file}")
return video_output_file
def multilingual_docs_conversion(
self,
string_text, # string
document=None, # doc path gui
directory_input="", # doc path
SOURCE_LANGUAGE="English (en)",
TRANSLATE_AUDIO_TO="English (en)",
tts_voice00="en-AU-WilliamNeural-Male",
name_final_file="sample",
translate_process="google_translator_iterative",
output_type="audio",
chunk_size=None,
is_gui=False,
progress=gr.Progress(),
):
SOURCE_LANGUAGE = LANGUAGES[SOURCE_LANGUAGE]
if translate_process != "disable translation":
TRANSLATE_AUDIO_TO = LANGUAGES[TRANSLATE_AUDIO_TO]
else:
TRANSLATE_AUDIO_TO = SOURCE_LANGUAGE
logger.info("No translation")
if tts_voice00[:2].lower() != TRANSLATE_AUDIO_TO[:2].lower():
logger.debug(
"Make sure to select a 'TTS Speaker' suitable for the "
"translation language to avoid errors with the TTS."
)
self.clear_cache(string_text, force=True)
is_string = False
if document is None:
if os.path.exists(directory_input):
document = directory_input
else:
document = string_text
is_string = True
document = document if isinstance(document, str) else document.name
if not document:
raise Exception("No data found")
# audio_wav = "audio.wav"
final_wav_file = (
"audio_book.wav"
if not name_final_file
else f"{name_final_file}.wav"
)
prog_disp("Processing text...", 0.15, is_gui, progress=progress)
result_file_path, result_text = document_preprocessor(
document, is_string
)
if (
output_type == "text"
and translate_process == "disable translation"
):
return result_file_path
if "SET_LIMIT" == os.getenv("DEMO"):
result_text = result_text[:50]
logger.info(
"DEMO; Generation is limited to 50 characters to prevent "
"CPU errors. No limitations with GPU.\n"
)
if translate_process != "disable translation":
# chunks text for translation
result_diarize = plain_text_to_segments(result_text, 1700)
prog_disp("Translating...", 0.30, is_gui, progress=progress)
# not or iterative with 1700 chars
result_diarize["segments"] = translate_text(
result_diarize["segments"],
TRANSLATE_AUDIO_TO,
translate_process,
chunk_size=0,
)
txt_file_path, result_text = segments_to_plain_text(result_diarize)
if output_type == "text":
return txt_file_path
# (TTS limits) plain text to result_diarize
chunk_size = (
chunk_size if chunk_size else determine_chunk_size(tts_voice00)
)
result_diarize = plain_text_to_segments(result_text, chunk_size)
logger.debug(result_diarize)
prog_disp("Text to speech...", 0.45, is_gui, progress=progress)
valid_speakers = audio_segmentation_to_voice(
result_diarize,
TRANSLATE_AUDIO_TO,
is_gui,
tts_voice00,
"",
"",
"",
"",
"",
)
# fix format and set folder output
audio_files, speakers_list = accelerate_segments(
result_diarize,
1.0,
valid_speakers,
)
# custom voice
if os.getenv("VOICES_MODELS") == "ENABLE":
prog_disp(
"Applying customized voices...",
0.80,
is_gui,
progress=progress,
)
vci(speakers_list, audio_files)
prog_disp(
"Creating final audio file...", 0.90, is_gui, progress=progress
)
remove_files(final_wav_file)
create_translated_audio(
result_diarize, audio_files, final_wav_file, True
)
return final_wav_file
title = "<center><strong><font size='7'>📽️ SoniTranslate 🈷️</font></strong></center>"
def create_gui(theme, logs_in_gui=False):
with gr.Blocks(theme=theme) as app:
gr.Markdown(title)
gr.Markdown(lg_conf["description"])
with gr.Tab(lg_conf["tab_translate"]):
with gr.Row():
with gr.Column():
input_data_type = gr.Dropdown(
["SUBMIT VIDEO", "URL", "Find Video Path"],
value="SUBMIT VIDEO",
label=lg_conf["video_source"],
)
def swap_visibility(data_type):
if data_type == "URL":
return (
gr.update(visible=False, value=None),
gr.update(visible=True, value=""),
gr.update(visible=False, value=""),
)
elif data_type == "SUBMIT VIDEO":
return (
gr.update(visible=True, value=None),
gr.update(visible=False, value=""),
gr.update(visible=False, value=""),
)
elif data_type == "Find Video Path":
return (
gr.update(visible=False, value=None),
gr.update(visible=False, value=""),
gr.update(visible=True, value=""),
)
video_input = gr.File(label="VIDEO")
blink_input = gr.Textbox(
visible=False,
label=lg_conf["link_label"],
info=lg_conf["link_info"],
placeholder=lg_conf["link_ph"],
)
directory_input = gr.Textbox(
visible=False,
label=lg_conf["dir_label"],
info=lg_conf["dir_info"],
placeholder=lg_conf["dir_ph"],
)
input_data_type.change(
fn=swap_visibility,
inputs=input_data_type,
outputs=[video_input, blink_input, directory_input],
)
link = gr.HTML()
SOURCE_LANGUAGE = gr.Dropdown(
LANGUAGES_LIST,
value="Automatic detection",
label=lg_conf["sl_label"],
info=lg_conf["sl_info"],
)
TRANSLATE_AUDIO_TO = gr.Dropdown(
LANGUAGES_LIST[1:],
value="English (en)",
label=lg_conf["tat_label"],
info=lg_conf["tat_info"],
)
line_ = gr.HTML("<hr></h2>")
gr.Markdown(lg_conf["num_speakers"])
MAX_TTS = 6
min_speakers = gr.Slider(
1,
MAX_TTS,
value=1,
label=lg_conf["min_sk"],
step=1,
visible=False,
)
max_speakers = gr.Slider(
1,
MAX_TTS,
value=2,
step=1,
label=lg_conf["max_sk"],
)
gr.Markdown(lg_conf["tts_select"])
def submit(value):
visibility_dict = {
f"tts_voice{i:02d}": gr.update(visible=i < value)
for i in range(6)
}
return [value for value in visibility_dict.values()]