Spaces:
Sleeping
Sleeping
File size: 35,283 Bytes
9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 ac7919b 9bedde5 |
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 |
"""UltraSinger uses AI to automatically create UltraStar song files"""
import copy
import getopt
import os
import sys
import re
import Levenshtein
import librosa
from tqdm import tqdm
from packaging import version
import soundfile as sf
from modules import os_helper
from modules.Audio.denoise import ffmpeg_reduce_noise
from modules.Audio.separation import separate_audio
from modules.Audio.vocal_chunks import (
export_chunks_from_transcribed_data,
export_chunks_from_ultrastar_data,
)
from modules.Audio.silence_processing import remove_silence_from_transcription_data, get_silence_sections
from modules.csv_handler import export_transcribed_data_to_csv
from modules.Audio.convert_audio import convert_audio_to_mono_wav, convert_wav_to_mp3
from modules.Audio.youtube import (
download_youtube_audio,
download_youtube_thumbnail,
download_youtube_video,
get_youtube_title,
)
from modules.DeviceDetection.device_detection import check_gpu_support
from modules.console_colors import (
ULTRASINGER_HEAD,
blue_highlighted,
gold_highlighted,
light_blue_highlighted,
red_highlighted,
)
from modules.Midi import midi_creator
from modules.Midi.midi_creator import (
convert_frequencies_to_notes,
create_midi_notes_from_pitched_data,
most_frequent,
)
from modules.Pitcher.pitcher import (
get_frequencies_with_high_confidence,
get_pitch_with_crepe_file,
)
from modules.Pitcher.pitched_data import PitchedData
from modules.Speech_Recognition.hyphenation import hyphenation, language_check, create_hyphenator
from modules.Speech_Recognition.Whisper import transcribe_with_whisper
from modules.Ultrastar import ultrastar_score_calculator, ultrastar_writer, ultrastar_converter, ultrastar_parser
from modules.Ultrastar.ultrastar_txt import UltrastarTxtValue
from Settings import Settings
from modules.Speech_Recognition.TranscribedData import TranscribedData
from modules.plot import plot, plot_spectrogram
from modules.musicbrainz_client import get_music_infos
settings = Settings()
def convert_midi_notes_to_ultrastar_notes(midi_notes: list[str]) -> list[int]:
"""Convert midi notes to ultrastar notes"""
print(f"{ULTRASINGER_HEAD} Creating Ultrastar notes from midi data")
ultrastar_note_numbers = []
for i in enumerate(midi_notes):
pos = i[0]
note_number_librosa = librosa.note_to_midi(midi_notes[pos])
pitch = ultrastar_converter.midi_note_to_ultrastar_note(
note_number_librosa
)
ultrastar_note_numbers.append(pitch)
# todo: Progress?
# print(
# f"Note: {midi_notes[i]} midi_note: {str(note_number_librosa)} pitch: {str(pitch)}"
# )
return ultrastar_note_numbers
def pitch_each_chunk_with_crepe(directory: str) -> list[str]:
"""Pitch each chunk with crepe and return midi notes"""
print(
f"{ULTRASINGER_HEAD} Pitching each chunk with {blue_highlighted('crepe')}"
)
midi_notes = []
for filename in sorted(
[f for f in os.listdir(directory) if f.endswith(".wav")],
key=lambda x: int(x.split("_")[1]),
):
filepath = os.path.join(directory, filename)
# todo: stepsize = duration? then when shorter than "it" it should take the duration. Otherwise there a more notes
pitched_data = get_pitch_with_crepe_file(
filepath,
settings.crepe_model_capacity,
settings.crepe_step_size,
settings.tensorflow_device,
)
conf_f = get_frequencies_with_high_confidence(
pitched_data.frequencies, pitched_data.confidence
)
notes = convert_frequencies_to_notes(conf_f)
note = most_frequent(notes)[0][0]
midi_notes.append(note)
# todo: Progress?
# print(filename + " f: " + str(mean))
return midi_notes
def add_hyphen_to_data(transcribed_data: list[TranscribedData], hyphen_words: list[list[str]]):
"""Add hyphen to transcribed data return new data list"""
new_data = []
for i, data in enumerate(transcribed_data):
if not hyphen_words[i]:
new_data.append(data)
else:
chunk_duration = data.end - data.start
chunk_duration = chunk_duration / (len(hyphen_words[i]))
next_start = data.start
for j in enumerate(hyphen_words[i]):
hyphenated_word_index = j[0]
dup = copy.copy(data)
dup.start = next_start
next_start = data.end - chunk_duration * (
len(hyphen_words[i]) - 1 - hyphenated_word_index
)
dup.end = next_start
dup.word = hyphen_words[i][hyphenated_word_index]
dup.is_hyphen = True
if hyphenated_word_index == len(hyphen_words[i]) - 1:
dup.is_word_end = True
else:
dup.is_word_end = False
new_data.append(dup)
return new_data
def get_bpm_from_data(data, sampling_rate):
"""Get real bpm from audio data"""
onset_env = librosa.onset.onset_strength(y=data, sr=sampling_rate)
wav_tempo = librosa.beat.tempo(onset_envelope=onset_env, sr=sampling_rate)
print(
f"{ULTRASINGER_HEAD} BPM is {blue_highlighted(str(round(wav_tempo[0], 2)))}"
)
return wav_tempo[0]
def get_bpm_from_file(wav_file: str) -> float:
"""Get real bpm from audio file"""
data, sampling_rate = librosa.load(wav_file, sr=None)
return get_bpm_from_data(data, sampling_rate)
def correct_words(recognized_words, word_list_file):
"""Docstring"""
with open(word_list_file, "r", encoding="utf-8") as file:
text = file.read()
word_list = text.split()
for i, rec_word in enumerate(recognized_words):
if rec_word.word in word_list:
continue
closest_word = min(
word_list, key=lambda x: Levenshtein.distance(rec_word.word, x)
)
print(recognized_words[i].word + " - " + closest_word)
recognized_words[i].word = closest_word
return recognized_words
def print_help() -> None:
"""Print help text"""
help_string = """
UltraSinger.py [opt] [mode] [transcription] [pitcher] [extra]
[opt]
-h This help text.
-i Ultrastar.txt
audio like .mp3, .wav, youtube link
-o Output folder
[mode]
## INPUT is audio ##
default Creates all
# Single file creation selection is in progress, you currently getting all!
(-u Create ultrastar txt file) # In Progress
(-m Create midi file) # In Progress
(-s Create sheet file) # In Progress
## INPUT is ultrastar.txt ##
default Creates all
# Single selection is in progress, you currently getting all!
(-r repitch Ultrastar.txt (input has to be audio)) # In Progress
(-p Check pitch of Ultrastar.txt input) # In Progress
(-m Create midi file) # In Progress
[transcription]
# Default is whisper
--whisper Multilingual model > tiny|base|small|medium|large-v1|large-v2 >> ((default) is large-v2
English-only model > tiny.en|base.en|small.en|medium.en
--whisper_align_model Use other languages model for Whisper provided from huggingface.co
--language Override the language detected by whisper, does not affect transcription but steps after transcription
--whisper_batch_size Reduce if low on GPU mem >> ((default) is 16)
--whisper_compute_type Change to "int8" if low on GPU mem (may reduce accuracy) >> ((default) is "float16" for cuda devices, "int8" for cpu)
[pitcher]
# Default is crepe
--crepe tiny|full >> ((default) is full)
--crepe_step_size unit is miliseconds >> ((default) is 10)
[extra]
--hyphenation True|False >> ((default) is True)
--disable_separation True|False >> ((default) is False)
--disable_karaoke True|False >> ((default) is False)
--create_audio_chunks True|False >> ((default) is False)
--keep_cache True|False >> ((default) is False)
--plot True|False >> ((default) is False)
--format_version 0.3.0|1.0.0|1.1.0 >> ((default) is 1.0.0)
[device]
--force_cpu True|False >> ((default) is False) All steps will be forced to cpu
--force_whisper_cpu True|False >> ((default) is False) Only whisper will be forced to cpu
--force_crepe_cpu True|False >> ((default) is False) Only crepe will be forced to cpu
"""
print(help_string)
def remove_unecessary_punctuations(transcribed_data: list[TranscribedData]) -> None:
"""Remove unecessary punctuations from transcribed data"""
punctuation = ".,"
for i, data in enumerate(transcribed_data):
data.word = data.word.translate(
{ord(i): None for i in punctuation}
)
def hyphenate_each_word(language: str, transcribed_data: list[TranscribedData]) -> list[list[str]] | None:
"""Hyphenate each word in the transcribed data."""
lang_region = language_check(language)
if lang_region is None:
print(
f"{ULTRASINGER_HEAD} {red_highlighted('Error in hyphenation for language ')} {blue_highlighted(language)}{red_highlighted(', maybe you want to disable it?')}"
)
return None
hyphenated_word = []
try:
hyphenator = create_hyphenator(lang_region)
for i in tqdm(enumerate(transcribed_data)):
pos = i[0]
hyphenated_word.append(
hyphenation(transcribed_data[pos].word, hyphenator)
)
except:
print(f"{ULTRASINGER_HEAD} {red_highlighted('Error in hyphenation for language ')} {blue_highlighted(language)}{red_highlighted(', maybe you want to disable it?')}")
return None
return hyphenated_word
def print_support() -> None:
"""Print support text"""
print()
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('Do you like UltraSinger? Want it to be even better? Then help with your')} {light_blue_highlighted('support')}{gold_highlighted('!')}"
)
print(
f"{ULTRASINGER_HEAD} See project page -> https://github.com/rakuri255/UltraSinger"
)
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('This will help a lot to keep this project alive and improved.')}"
)
def print_version() -> None:
"""Print version text"""
print()
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('*****************************')}"
)
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('UltraSinger Version:')} {light_blue_highlighted(settings.APP_VERSION)}"
)
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('*****************************')}"
)
def run() -> None:
"""The processing function of this program"""
is_audio = ".txt" not in settings.input_file_path
ultrastar_class = None
real_bpm = None
(title, artist, year, genre) = (None, None, None, None)
if not is_audio: # Parse Ultrastar txt
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('re-pitch mode')}"
)
(
basename_without_ext,
real_bpm,
song_output,
ultrastar_audio_input_path,
ultrastar_class,
) = parse_ultrastar_txt()
elif settings.input_file_path.startswith("https:"): # Youtube
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('full automatic mode')}"
)
(
basename_without_ext,
song_output,
ultrastar_audio_input_path,
(title, artist, year, genre)
) = download_from_youtube()
else: # Audio File
print(
f"{ULTRASINGER_HEAD} {gold_highlighted('full automatic mode')}"
)
(
basename_without_ext,
song_output,
ultrastar_audio_input_path,
(title, artist, year, genre)
) = infos_from_audio_input_file()
cache_path = os.path.join(song_output, "cache")
settings.processing_audio_path = os.path.join(
cache_path, basename_without_ext + ".wav"
)
os_helper.create_folder(cache_path)
# Separate vocal from audio
audio_separation_path = separate_vocal_from_audio(
basename_without_ext, cache_path, ultrastar_audio_input_path
)
vocals_path = os.path.join(audio_separation_path, "vocals.wav")
instrumental_path = os.path.join(audio_separation_path, "no_vocals.wav")
# Move instrumental and vocals
if settings.create_karaoke and version.parse(settings.format_version) < version.parse("1.1.0"):
karaoke_output_path = os.path.join(song_output, basename_without_ext + " [Karaoke].mp3")
convert_wav_to_mp3(instrumental_path, karaoke_output_path)
if version.parse(settings.format_version) >= version.parse("1.1.0"):
instrumental_output_path = os.path.join(song_output, basename_without_ext + " [Instrumental].mp3")
convert_wav_to_mp3(instrumental_path, instrumental_output_path)
vocals_output_path = os.path.join(song_output, basename_without_ext + " [Vocals].mp3")
convert_wav_to_mp3(vocals_path, vocals_output_path)
if settings.use_separated_vocal:
input_path = vocals_path
else:
input_path = ultrastar_audio_input_path
# Denoise vocal audio
denoised_output_path = os.path.join(
cache_path, basename_without_ext + "_denoised.wav"
)
denoise_vocal_audio(input_path, denoised_output_path)
# Convert to mono audio
mono_output_path = os.path.join(
cache_path, basename_without_ext + "_mono.wav"
)
convert_audio_to_mono_wav(denoised_output_path, mono_output_path)
# Mute silence sections
mute_output_path = os.path.join(
cache_path, basename_without_ext + "_mute.wav"
)
mute_no_singing_parts(mono_output_path, mute_output_path)
# Define the audio file to process
settings.processing_audio_path = mute_output_path
# Audio transcription
transcribed_data = None
language = settings.language
if is_audio:
detected_language, transcribed_data = transcribe_audio()
if language is None:
language = detected_language
remove_unecessary_punctuations(transcribed_data)
if settings.hyphenation:
hyphen_words = hyphenate_each_word(language, transcribed_data)
if hyphen_words is not None:
transcribed_data = add_hyphen_to_data(transcribed_data, hyphen_words)
transcribed_data = remove_silence_from_transcription_data(
settings.processing_audio_path, transcribed_data
)
# todo: do we need to correct words?
# lyric = 'input/faber_lyric.txt'
# --corrected_words = correct_words(vosk_speech, lyric)
# Create audio chunks
if settings.create_audio_chunks:
create_audio_chunks(
cache_path,
is_audio,
transcribed_data,
ultrastar_audio_input_path,
ultrastar_class,
)
# Pitch the audio
midi_notes, pitched_data, ultrastar_note_numbers = pitch_audio(
is_audio, transcribed_data, ultrastar_class
)
# Create plot
if settings.create_plot:
vocals_path = os.path.join(audio_separation_path, "vocals.wav")
plot_spectrogram(vocals_path, song_output, "vocals.wav")
plot_spectrogram(settings.processing_audio_path, song_output, "processing audio")
plot(pitched_data, song_output, transcribed_data, ultrastar_class, midi_notes)
# Write Ultrastar txt
if is_audio:
real_bpm, ultrastar_file_output = create_ultrastar_txt_from_automation(
basename_without_ext,
song_output,
transcribed_data,
ultrastar_audio_input_path,
ultrastar_note_numbers,
language,
title,
artist,
year,
genre
)
else:
ultrastar_file_output = create_ultrastar_txt_from_ultrastar_data(
song_output, ultrastar_class, ultrastar_note_numbers
)
# Calc Points
ultrastar_class, simple_score, accurate_score = calculate_score_points(
is_audio, pitched_data, ultrastar_class, ultrastar_file_output
)
# Add calculated score to Ultrastar txt #Todo: Missing Karaoke
ultrastar_writer.add_score_to_ultrastar_txt(
ultrastar_file_output, simple_score
)
# Midi
if settings.create_midi:
create_midi_file(real_bpm, song_output, ultrastar_class, basename_without_ext)
# Cleanup
if not settings.keep_cache:
remove_cache_folder(cache_path)
# Print Support
print_support()
def mute_no_singing_parts(mono_output_path, mute_output_path):
print(
f"{ULTRASINGER_HEAD} Mute audio parts with no singing"
)
silence_sections = get_silence_sections(mono_output_path)
y, sr = librosa.load(mono_output_path, sr=None)
# Mute the parts of the audio with no singing
for i in silence_sections:
# Define the time range to mute
start_time = i[0] # Start time in seconds
end_time = i[1] # End time in seconds
# Convert time to sample indices
start_sample = int(start_time * sr)
end_sample = int(end_time * sr)
y[start_sample:end_sample] = 0
sf.write(mute_output_path, y, sr)
def get_unused_song_output_dir(path: str) -> str:
"""Get an unused song output dir"""
# check if dir exists and add (i) if it does
i = 1
if os_helper.check_if_folder_exists(path):
path = f"{path} ({i})"
else:
return path
while os_helper.check_if_folder_exists(path):
path = path.replace(f"({i - 1})", f"({i})")
i += 1
if i > 999:
print(
f"{ULTRASINGER_HEAD} {red_highlighted('Error: Could not create output folder! (999) is the maximum number of tries.')}"
)
sys.exit(1)
return path
def transcribe_audio() -> (str, list[TranscribedData]):
"""Transcribe audio with AI"""
if settings.transcriber == "whisper":
device = "cpu" if settings.force_whisper_cpu else settings.pytorch_device
transcribed_data, detected_language = transcribe_with_whisper(
settings.processing_audio_path,
settings.whisper_model,
device,
settings.whisper_align_model,
settings.whisper_batch_size,
settings.whisper_compute_type,
settings.language,
)
else:
raise NotImplementedError
return detected_language, transcribed_data
def separate_vocal_from_audio(
basename_without_ext: str, cache_path: str, ultrastar_audio_input_path: str
) -> str:
"""Separate vocal from audio"""
audio_separation_path = os.path.join(
cache_path, "separated", "htdemucs", basename_without_ext
)
if settings.use_separated_vocal or settings.create_karaoke:
separate_audio(ultrastar_audio_input_path, cache_path, settings.pytorch_device)
return audio_separation_path
def calculate_score_points(
is_audio: bool, pitched_data: PitchedData, ultrastar_class: UltrastarTxtValue, ultrastar_file_output: str
):
"""Calculate score points"""
if is_audio:
ultrastar_class = ultrastar_parser.parse_ultrastar_txt(
ultrastar_file_output
)
(
simple_score,
accurate_score,
) = ultrastar_score_calculator.calculate_score(
pitched_data, ultrastar_class
)
ultrastar_score_calculator.print_score_calculation(
simple_score, accurate_score
)
else:
print(
f"{ULTRASINGER_HEAD} {blue_highlighted('Score of original Ultrastar txt')}"
)
(
simple_score,
accurate_score,
) = ultrastar_score_calculator.calculate_score(
pitched_data, ultrastar_class
)
ultrastar_score_calculator.print_score_calculation(
simple_score, accurate_score
)
print(
f"{ULTRASINGER_HEAD} {blue_highlighted('Score of re-pitched Ultrastar txt')}"
)
ultrastar_class = ultrastar_parser.parse_ultrastar_txt(
ultrastar_file_output
)
(
simple_score,
accurate_score,
) = ultrastar_score_calculator.calculate_score(
pitched_data, ultrastar_class
)
ultrastar_score_calculator.print_score_calculation(
simple_score, accurate_score
)
return ultrastar_class, simple_score, accurate_score
def create_ultrastar_txt_from_ultrastar_data(
song_output: str, ultrastar_class: UltrastarTxtValue, ultrastar_note_numbers: list[int]
) -> str:
"""Create Ultrastar txt from Ultrastar data"""
output_repitched_ultrastar = os.path.join(
song_output, ultrastar_class.title + ".txt"
)
ultrastar_writer.create_repitched_txt_from_ultrastar_data(
settings.input_file_path,
ultrastar_note_numbers,
output_repitched_ultrastar,
)
return output_repitched_ultrastar
def create_ultrastar_txt_from_automation(
basename_without_ext: str,
song_output: str,
transcribed_data: list[TranscribedData],
ultrastar_audio_input_path: str,
ultrastar_note_numbers: list[int],
language: str,
title: str,
artist: str,
year: str,
genre: str
):
"""Create Ultrastar txt from automation"""
ultrastar_header = UltrastarTxtValue()
ultrastar_header.version = settings.format_version
ultrastar_header.title = basename_without_ext
ultrastar_header.artist = basename_without_ext
ultrastar_header.mp3 = basename_without_ext + ".mp3"
ultrastar_header.audio = basename_without_ext + ".mp3"
ultrastar_header.vocals = basename_without_ext + " [Vocals].mp3"
ultrastar_header.instrumental = basename_without_ext + " [Instrumental].mp3"
ultrastar_header.video = basename_without_ext + ".mp4"
ultrastar_header.language = language
cover = basename_without_ext + " [CO].jpg"
ultrastar_header.cover = (
cover
if os_helper.check_file_exists(os.path.join(song_output, cover))
else None
)
ultrastar_header.creator = f"{ultrastar_header.creator} {Settings.APP_VERSION}"
ultrastar_header.comment = f"{ultrastar_header.comment} {Settings.APP_VERSION}"
# Additional data
if title is not None:
ultrastar_header.title = title
if artist is not None:
ultrastar_header.artist = artist
if year is not None:
ultrastar_header.year = extract_year(year)
if genre is not None:
ultrastar_header.genre = format_separated_string(genre)
real_bpm = get_bpm_from_file(ultrastar_audio_input_path)
ultrastar_file_output = os.path.join(
song_output, basename_without_ext + ".txt"
)
ultrastar_writer.create_ultrastar_txt_from_automation(
transcribed_data,
ultrastar_note_numbers,
ultrastar_file_output,
ultrastar_header,
real_bpm,
)
if settings.create_karaoke and version.parse(settings.format_version) < version.parse("1.1.0"):
title = basename_without_ext + " [Karaoke]"
ultrastar_header.title = title
ultrastar_header.mp3 = title + ".mp3"
karaoke_output_path = os.path.join(song_output, title)
karaoke_txt_output_path = karaoke_output_path + ".txt"
ultrastar_writer.create_ultrastar_txt_from_automation(
transcribed_data,
ultrastar_note_numbers,
karaoke_txt_output_path,
ultrastar_header,
real_bpm,
)
return real_bpm, ultrastar_file_output
def extract_year(date: str) -> str:
match = re.search(r'\b\d{4}\b', date)
if match:
return match.group(0)
else:
return date
def format_separated_string(data: str) -> str:
temp = re.sub(r'[;/]', ',', data)
words = temp.split(',')
words = [s for s in words if s.strip()]
for i, word in enumerate(words):
if "-" not in word:
words[i] = word.strip().capitalize() + ', '
else:
dash_words = word.split('-')
capitalized_dash_words = [dash_word.strip().capitalize() for dash_word in dash_words]
formatted_dash_word = '-'.join(capitalized_dash_words) + ', '
words[i] = formatted_dash_word
formatted_string = ''.join(words)
if formatted_string.endswith(', '):
formatted_string = formatted_string[:-2]
return formatted_string
def infos_from_audio_input_file() -> tuple[str, str, str, tuple[str, str, str, str]]:
"""Infos from audio input file"""
basename = os.path.basename(settings.input_file_path)
basename_without_ext = os.path.splitext(basename)[0]
artist, title = None, None
if " - " in basename_without_ext:
artist, title = basename_without_ext.split(" - ", 1)
search_string = f"{artist} - {title}"
else:
search_string = basename_without_ext
# Get additional data for song
(title_info, artist_info, year_info, genre_info) = get_music_infos(search_string)
if title_info is not None:
title = title_info
artist = artist_info
if artist is not None and title is not None:
basename_without_ext = f"{artist} - {title}"
extension = os.path.splitext(basename)[1]
basename = f"{basename_without_ext}{extension}"
song_output = os.path.join(settings.output_file_path, basename_without_ext)
song_output = get_unused_song_output_dir(song_output)
os_helper.create_folder(song_output)
os_helper.copy(settings.input_file_path, song_output)
os_helper.rename(os.path.join(song_output, os.path.basename(settings.input_file_path)), os.path.join(song_output, basename))
ultrastar_audio_input_path = os.path.join(song_output, basename)
return basename_without_ext, song_output, ultrastar_audio_input_path, (title, artist, year_info, genre_info)
FILENAME_REPLACEMENTS = (('?:"', ""), ("<", "("), (">", ")"), ("/\\|*", "-"))
def sanitize_filename(fname: str) -> str:
"""Sanitize filename"""
for old, new in FILENAME_REPLACEMENTS:
for char in old:
fname = fname.replace(char, new)
if fname.endswith("."):
fname = fname.rstrip(" .") # Windows does not like trailing periods
return fname
def download_from_youtube() -> tuple[str, str, str, tuple[str, str, str, str]]:
"""Download from YouTube"""
(artist, title) = get_youtube_title(settings.input_file_path)
# Get additional data for song
(title_info, artist_info, year_info, genre_info) = get_music_infos(f"{artist} - {title}")
if title_info is not None:
title = title_info
artist = artist_info
basename_without_ext = sanitize_filename(f"{artist} - {title}")
basename = basename_without_ext + ".mp3"
song_output = os.path.join(settings.output_file_path, basename_without_ext)
song_output = get_unused_song_output_dir(song_output)
os_helper.create_folder(song_output)
download_youtube_audio(
settings.input_file_path, basename_without_ext, song_output
)
download_youtube_video(
settings.input_file_path, basename_without_ext, song_output
)
download_youtube_thumbnail(
settings.input_file_path, basename_without_ext, song_output
)
ultrastar_audio_input_path = os.path.join(song_output, basename)
return basename_without_ext, song_output, ultrastar_audio_input_path, (title, artist, year_info, genre_info)
def parse_ultrastar_txt() -> tuple[str, float, str, str, UltrastarTxtValue]:
"""Parse Ultrastar txt"""
ultrastar_class = ultrastar_parser.parse_ultrastar_txt(
settings.input_file_path
)
real_bpm = ultrastar_converter.ultrastar_bpm_to_real_bpm(
float(ultrastar_class.bpm.replace(",", "."))
)
ultrastar_mp3_name = ultrastar_class.mp3
basename_without_ext = os.path.splitext(ultrastar_mp3_name)[0]
dirname = os.path.dirname(settings.input_file_path)
ultrastar_audio_input_path = os.path.join(dirname, ultrastar_mp3_name)
song_output = os.path.join(
settings.output_file_path,
ultrastar_class.artist.strip() + " - " + ultrastar_class.title.strip(),
)
song_output = get_unused_song_output_dir(str(song_output))
os_helper.create_folder(song_output)
return (
str(basename_without_ext),
real_bpm,
song_output,
str(ultrastar_audio_input_path),
ultrastar_class,
)
def create_midi_file(real_bpm: float,
song_output: str,
ultrastar_class: UltrastarTxtValue,
basename_without_ext: str) -> None:
"""Create midi file"""
print(
f"{ULTRASINGER_HEAD} Creating Midi with {blue_highlighted('pretty_midi')}"
)
voice_instrument = [
midi_creator.convert_ultrastar_to_midi_instrument(ultrastar_class)
]
midi_output = os.path.join(song_output, f"{basename_without_ext}.mid")
midi_creator.instruments_to_midi(
voice_instrument, real_bpm, midi_output
)
def pitch_audio(is_audio: bool, transcribed_data: list[TranscribedData], ultrastar_class: UltrastarTxtValue) -> tuple[
list[str], PitchedData, list[int]]:
"""Pitch audio"""
# todo: chunk pitching as option?
# midi_notes = pitch_each_chunk_with_crepe(chunk_folder_name)
device = "cpu" if settings.force_crepe_cpu else settings.tensorflow_device
pitched_data = get_pitch_with_crepe_file(
settings.processing_audio_path,
settings.crepe_model_capacity,
settings.crepe_step_size,
device,
)
if is_audio:
start_times = []
end_times = []
for i, data in enumerate(transcribed_data):
start_times.append(data.start)
end_times.append(data.end)
midi_notes = create_midi_notes_from_pitched_data(
start_times, end_times, pitched_data
)
else:
midi_notes = create_midi_notes_from_pitched_data(
ultrastar_class.startTimes, ultrastar_class.endTimes, pitched_data
)
ultrastar_note_numbers = convert_midi_notes_to_ultrastar_notes(midi_notes)
return midi_notes, pitched_data, ultrastar_note_numbers
def create_audio_chunks(
cache_path: str,
is_audio: bool,
transcribed_data: list[TranscribedData],
ultrastar_audio_input_path: str,
ultrastar_class: UltrastarTxtValue
) -> None:
"""Create audio chunks"""
audio_chunks_path = os.path.join(
cache_path, settings.audio_chunk_folder_name
)
os_helper.create_folder(audio_chunks_path)
if is_audio: # and csv
csv_filename = os.path.join(audio_chunks_path, "_chunks.csv")
export_chunks_from_transcribed_data(
settings.processing_audio_path, transcribed_data, audio_chunks_path
)
export_transcribed_data_to_csv(transcribed_data, csv_filename)
else:
export_chunks_from_ultrastar_data(
ultrastar_audio_input_path, ultrastar_class, audio_chunks_path
)
def denoise_vocal_audio(input_path: str, output_path: str) -> None:
"""Denoise vocal audio"""
ffmpeg_reduce_noise(input_path, output_path)
def main(argv: list[str]) -> None:
"""Main function"""
print_version()
init_settings(argv)
run()
sys.exit()
def remove_cache_folder(cache_path: str) -> None:
"""Remove cache folder"""
os_helper.remove_folder(cache_path)
def init_settings(argv: list[str]) -> None:
"""Init settings"""
long, short = arg_options()
opts, args = getopt.getopt(argv, short, long)
if len(opts) == 0:
print_help()
sys.exit()
for opt, arg in opts:
if opt == "-h":
print_help()
sys.exit()
elif opt in ("-i", "--ifile"):
settings.input_file_path = arg
elif opt in ("-o", "--ofile"):
settings.output_file_path = arg
elif opt in ("--whisper"):
settings.transcriber = "whisper"
settings.whisper_model = arg
elif opt in ("--whisper_align_model"):
settings.whisper_align_model = arg
elif opt in ("--whisper_batch_size"):
settings.whisper_batch_size = int(arg)
elif opt in ("--whisper_compute_type"):
settings.whisper_compute_type = arg
elif opt in ("--language"):
settings.language = arg
elif opt in ("--crepe"):
settings.crepe_model_capacity = arg
elif opt in ("--crepe_step_size"):
settings.crepe_step_size = int(arg)
elif opt in ("--plot"):
settings.create_plot = arg in ["True", "true"]
elif opt in ("--midi"):
settings.create_midi = arg in ["True", "true"]
elif opt in ("--hyphenation"):
settings.hyphenation = eval(arg.title())
elif opt in ("--disable_separation"):
settings.use_separated_vocal = not arg
elif opt in ("--disable_karaoke"):
settings.create_karaoke = not arg
elif opt in ("--create_audio_chunks"):
settings.create_audio_chunks = arg
elif opt in ("--force_cpu"):
settings.force_cpu = arg
if settings.force_cpu:
os.environ["CUDA_VISIBLE_DEVICES"] = "-1"
elif opt in ("--force_whisper_cpu"):
settings.force_whisper_cpu = eval(arg.title())
elif opt in ("--force_crepe_cpu"):
settings.force_crepe_cpu = eval(arg.title())
elif opt in ("--format_version"):
if arg != '0.3.0' and arg != '1.0.0' and arg != '1.1.0':
print(
f"{ULTRASINGER_HEAD} {red_highlighted('Error: Format version')} {blue_highlighted(arg)} {red_highlighted('is not supported.')}"
)
sys.exit(1)
settings.format_version = arg
elif opt in ("--keep_cache"):
settings.keep_cache = arg
if settings.output_file_path == "":
if settings.input_file_path.startswith("https:"):
dirname = os.getcwd()
else:
dirname = os.path.dirname(settings.input_file_path)
settings.output_file_path = os.path.join(dirname, "output")
if not settings.force_cpu:
settings.tensorflow_device, settings.pytorch_device = check_gpu_support()
def arg_options():
short = "hi:o:amv:"
long = [
"ifile=",
"ofile=",
"crepe=",
"crepe_step_size=",
"whisper=",
"whisper_align_model=",
"whisper_batch_size=",
"whisper_compute_type=",
"language=",
"plot=",
"midi=",
"hyphenation=",
"disable_separation=",
"disable_karaoke=",
"create_audio_chunks=",
"force_cpu=",
"force_whisper_cpu=",
"force_crepe_cpu=",
"format_version=",
"keep_cache"
]
return long, short
if __name__ == "__main__":
main(sys.argv[1:]) |