Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -22,7 +22,6 @@ from TTS.api import TTS
|
|
| 22 |
import base64
|
| 23 |
import pickle
|
| 24 |
import json
|
| 25 |
-
from moviepy.editor import TextClip, CompositeVideoClip, ColorClip, AudioFileClip
|
| 26 |
|
| 27 |
# Suppress warnings
|
| 28 |
warnings.filterwarnings("ignore")
|
|
@@ -264,7 +263,6 @@ def process_audio(audio_file, selected_effects, isolate_vocals, preset_name, exp
|
|
| 264 |
"Normalize": apply_normalize,
|
| 265 |
"Noise Gate": lambda x: apply_noise_gate(x, threshold=-50.0),
|
| 266 |
"Limiter": lambda x: apply_limiter(x, limit_dB=-1),
|
| 267 |
-
"Flanger": lambda x: apply_phaser(x, rate=1.2, depth=0.9, mix=0.7),
|
| 268 |
"Bitcrusher": lambda x: apply_bitcrush(x, bit_depth=8),
|
| 269 |
"Auto Gain": lambda x: apply_auto_gain(x, target_dB=-20),
|
| 270 |
"Vocal Distortion": lambda x: apply_vocal_distortion(x),
|
|
@@ -369,7 +367,7 @@ def batch_process_audio(files, selected_effects, isolate_vocals, preset_name, ex
|
|
| 369 |
results.append(processed_path)
|
| 370 |
session_logs.append(log)
|
| 371 |
|
| 372 |
-
zip_path = os.path.join(
|
| 373 |
with zipfile.ZipFile(zip_path, 'w') as zipf:
|
| 374 |
for i, res in enumerate(results):
|
| 375 |
filename = f"processed_{i}.{export_format.lower()}"
|
|
@@ -384,36 +382,15 @@ def batch_process_audio(files, selected_effects, isolate_vocals, preset_name, ex
|
|
| 384 |
# === Vocal Pitch Correction β Auto-Tune Style ===
|
| 385 |
def auto_tune_vocal(audio_path, target_key="C"):
|
| 386 |
try:
|
| 387 |
-
# Placeholder for real-time pitch detection
|
| 388 |
return apply_pitch_shift(AudioSegment.from_file(audio_path), 0.2)
|
| 389 |
except Exception as e:
|
| 390 |
return None
|
| 391 |
|
| 392 |
-
# === Create Karaoke Video from Audio + Lyrics ===
|
| 393 |
-
def create_karaoke_video(audio_path, lyrics, bg_image=None):
|
| 394 |
-
try:
|
| 395 |
-
from moviepy.editor import TextClip, CompositeVideoClip, ColorClip, AudioFileClip
|
| 396 |
-
|
| 397 |
-
audio = AudioFileClip(audio_path)
|
| 398 |
-
video = ColorClip(size=(1280, 720), color=(0, 0, 0), duration=audio.duration_seconds)
|
| 399 |
-
words = [(word.strip(), i * 3, (i+1)*3) for i, word in enumerate(lyrics.split())]
|
| 400 |
-
|
| 401 |
-
text_clips = [
|
| 402 |
-
TextClip(word, fontsize=60, color='white').set_position('center').set_duration(end - start).set_start(start)
|
| 403 |
-
for word, start, end in words
|
| 404 |
-
]
|
| 405 |
-
|
| 406 |
-
final_video = CompositeVideoClip([video] + text_clips).set_audio(audio)
|
| 407 |
-
out_path = os.path.join(tempfile.gettempdir(), "karaoke.mp4")
|
| 408 |
-
final_video.write_videofile(out_path, codec="libx264", audio_codec="aac")
|
| 409 |
-
return out_path
|
| 410 |
-
except Exception as e:
|
| 411 |
-
return f"β οΈ Failed: {str(e)}"
|
| 412 |
-
|
| 413 |
# === Real-Time Spectrum Analyzer + Live EQ Preview ===
|
| 414 |
def visualize_spectrum(audio_path):
|
| 415 |
y, sr = torchaudio.load(audio_path)
|
| 416 |
y_np = y.numpy().flatten()
|
|
|
|
| 417 |
stft = librosa.stft(y_np)
|
| 418 |
db = librosa.amplitude_to_db(abs(stft))
|
| 419 |
|
|
@@ -501,30 +478,6 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 501 |
clear_btn=None
|
| 502 |
)
|
| 503 |
|
| 504 |
-
# --- Preset Cards Gallery β Visual Selection ===
|
| 505 |
-
with gr.Tab("π Preset Gallery"):
|
| 506 |
-
gr.Markdown("### Select a preset visually")
|
| 507 |
-
preset_gallery = gr.Gallery(value=[
|
| 508 |
-
("images/pop_card.png", "Pop"),
|
| 509 |
-
("images/edm_card.png", "EDM"),
|
| 510 |
-
("images/rock_card.png", "Rock"),
|
| 511 |
-
("images/hiphop_card.png", "Hip-Hop"),
|
| 512 |
-
("images/acoustic_card.png", "Acoustic"),
|
| 513 |
-
("images/stage_mode_card.png", "Stage Mode"),
|
| 514 |
-
("images/vocal_distortion_card.png", "Vocal Distortion"),
|
| 515 |
-
("images/tube_saturation_card.png", "Tube Saturation")
|
| 516 |
-
], label="Preset Cards", columns=4, height="auto")
|
| 517 |
-
|
| 518 |
-
preset_name_out = gr.Dropdown(choices=preset_names, label="Selected Preset")
|
| 519 |
-
preset_effects_out = gr.CheckboxGroup(choices=list(preset_choices["Default"]), label="Effects")
|
| 520 |
-
|
| 521 |
-
def load_preset_by_card(evt: gr.SelectData):
|
| 522 |
-
index = evt.index % len(preset_names)
|
| 523 |
-
name = preset_names[index]
|
| 524 |
-
return name, preset_choices[name]
|
| 525 |
-
|
| 526 |
-
preset_gallery.select(fn=load_preset_by_card, inputs=[], outputs=[preset_name_out, preset_effects_out])
|
| 527 |
-
|
| 528 |
# --- Vocal Doubler / Harmonizer β Added ===
|
| 529 |
with gr.Tab("π§ Vocal Doubler / Harmonizer"):
|
| 530 |
gr.Interface(
|
|
@@ -570,20 +523,6 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 570 |
description="Correct vocal pitch automatically"
|
| 571 |
)
|
| 572 |
|
| 573 |
-
# --- Karaoke Video Creator β Added Back ===
|
| 574 |
-
with gr.Tab("πΉ Create Karaoke Video"):
|
| 575 |
-
gr.Interface(
|
| 576 |
-
fn=create_karaoke_video,
|
| 577 |
-
inputs=[
|
| 578 |
-
gr.Audio(label="Upload Track", type="filepath"),
|
| 579 |
-
gr.Textbox(label="Lyrics", lines=10),
|
| 580 |
-
gr.File(label="Background (Optional)")
|
| 581 |
-
],
|
| 582 |
-
outputs=gr.Video(label="Karaoke Video"),
|
| 583 |
-
title="Make Karaoke Videos from Audio + Lyrics",
|
| 584 |
-
description="Generate karaoke-style videos with real-time sync."
|
| 585 |
-
)
|
| 586 |
-
|
| 587 |
# --- Real-Time Spectrum Analyzer + Live EQ Preview β Added Back ===
|
| 588 |
with gr.Tab("π Frequency Spectrum"):
|
| 589 |
gr.Interface(
|
|
@@ -648,6 +587,30 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 648 |
description="Load your saved session"
|
| 649 |
)
|
| 650 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 651 |
# --- Prompt-Based Editing Tab β Added Back ===
|
| 652 |
def process_prompt(audio, prompt):
|
| 653 |
return apply_noise_reduction(audio)
|
|
|
|
| 22 |
import base64
|
| 23 |
import pickle
|
| 24 |
import json
|
|
|
|
| 25 |
|
| 26 |
# Suppress warnings
|
| 27 |
warnings.filterwarnings("ignore")
|
|
|
|
| 263 |
"Normalize": apply_normalize,
|
| 264 |
"Noise Gate": lambda x: apply_noise_gate(x, threshold=-50.0),
|
| 265 |
"Limiter": lambda x: apply_limiter(x, limit_dB=-1),
|
|
|
|
| 266 |
"Bitcrusher": lambda x: apply_bitcrush(x, bit_depth=8),
|
| 267 |
"Auto Gain": lambda x: apply_auto_gain(x, target_dB=-20),
|
| 268 |
"Vocal Distortion": lambda x: apply_vocal_distortion(x),
|
|
|
|
| 367 |
results.append(processed_path)
|
| 368 |
session_logs.append(log)
|
| 369 |
|
| 370 |
+
zip_path = os.path.join(tempfile.gettempdir(), "batch_output.zip")
|
| 371 |
with zipfile.ZipFile(zip_path, 'w') as zipf:
|
| 372 |
for i, res in enumerate(results):
|
| 373 |
filename = f"processed_{i}.{export_format.lower()}"
|
|
|
|
| 382 |
# === Vocal Pitch Correction β Auto-Tune Style ===
|
| 383 |
def auto_tune_vocal(audio_path, target_key="C"):
|
| 384 |
try:
|
|
|
|
| 385 |
return apply_pitch_shift(AudioSegment.from_file(audio_path), 0.2)
|
| 386 |
except Exception as e:
|
| 387 |
return None
|
| 388 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 389 |
# === Real-Time Spectrum Analyzer + Live EQ Preview ===
|
| 390 |
def visualize_spectrum(audio_path):
|
| 391 |
y, sr = torchaudio.load(audio_path)
|
| 392 |
y_np = y.numpy().flatten()
|
| 393 |
+
|
| 394 |
stft = librosa.stft(y_np)
|
| 395 |
db = librosa.amplitude_to_db(abs(stft))
|
| 396 |
|
|
|
|
| 478 |
clear_btn=None
|
| 479 |
)
|
| 480 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 481 |
# --- Vocal Doubler / Harmonizer β Added ===
|
| 482 |
with gr.Tab("π§ Vocal Doubler / Harmonizer"):
|
| 483 |
gr.Interface(
|
|
|
|
| 523 |
description="Correct vocal pitch automatically"
|
| 524 |
)
|
| 525 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 526 |
# --- Real-Time Spectrum Analyzer + Live EQ Preview β Added Back ===
|
| 527 |
with gr.Tab("π Frequency Spectrum"):
|
| 528 |
gr.Interface(
|
|
|
|
| 587 |
description="Load your saved session"
|
| 588 |
)
|
| 589 |
|
| 590 |
+
# --- Preset Cards Gallery β Visual Selection ===
|
| 591 |
+
with gr.Tab("π Preset Gallery"):
|
| 592 |
+
gr.Markdown("### Select a preset visually")
|
| 593 |
+
preset_gallery = gr.Gallery(value=[
|
| 594 |
+
("images/pop_card.png", "Pop"),
|
| 595 |
+
("images/edm_card.png", "EDM"),
|
| 596 |
+
("images/rock_card.png", "Rock"),
|
| 597 |
+
("images/hiphop_card.png", "Hip-Hop"),
|
| 598 |
+
("images/acoustic_card.png", "Acoustic"),
|
| 599 |
+
("images/stage_mode_card.png", "Stage Mode"),
|
| 600 |
+
("images/vocal_distortion_card.png", "Vocal Distortion"),
|
| 601 |
+
("images/tube_saturation_card.png", "Tube Saturation")
|
| 602 |
+
], label="Preset Cards", columns=4, height="auto")
|
| 603 |
+
|
| 604 |
+
preset_name_out = gr.Dropdown(choices=preset_names, label="Selected Preset")
|
| 605 |
+
preset_effects_out = gr.CheckboxGroup(choices=list(preset_choices["Default"]), label="Effects")
|
| 606 |
+
|
| 607 |
+
def load_preset_by_card(evt: gr.SelectData):
|
| 608 |
+
index = evt.index % len(preset_names)
|
| 609 |
+
name = preset_names[index]
|
| 610 |
+
return name, preset_choices[name]
|
| 611 |
+
|
| 612 |
+
preset_gallery.select(fn=load_preset_by_card, inputs=[], outputs=[preset_name_out, preset_effects_out])
|
| 613 |
+
|
| 614 |
# --- Prompt-Based Editing Tab β Added Back ===
|
| 615 |
def process_prompt(audio, prompt):
|
| 616 |
return apply_noise_reduction(audio)
|