Spaces:
Sleeping
Sleeping
File size: 9,110 Bytes
1f72b0c b808439 1088ad0 c7d476a 1f72b0c 0f64582 beea2cf 00d8b3c 1088ad0 1f72b0c 00d8b3c 1088ad0 0f64582 da777ed c7d476a 1f72b0c 00d8b3c 460ec26 1f72b0c 1088ad0 c7d476a 1f72b0c 02e97a3 c7d476a 02e97a3 f7e8b14 02e97a3 5accea7 02e97a3 1088ad0 00d8b3c f7e8b14 00d8b3c f7e8b14 6b58728 f7e8b14 6b58728 f7e8b14 7002418 f7e8b14 d446056 00d8b3c 5accea7 00d8b3c f7e8b14 5accea7 f7e8b14 4123c9b f8ec5f8 02e97a3 c7d476a f8ec5f8 1f72b0c 460ec26 8d8ac09 460ec26 8d8ac09 6e16e99 460ec26 f7e8b14 beea2cf 00d8b3c beea2cf f7e8b14 5accea7 0f64582 00d8b3c 6b58728 57af5e5 00d8b3c 5accea7 f8ec5f8 a666288 bc278d1 a666288 7fc09ab 460ec26 00d8b3c 460ec26 7002418 00d8b3c f7e8b14 a666288 00d8b3c f7e8b14 c7d476a 00d8b3c ced4e6e 0f64582 ced4e6e 00d8b3c 4123c9b f031dba 00d8b3c f031dba 4123c9b ced4e6e c7d476a beea2cf f8ec5f8 beea2cf 1f72b0c 00d8b3c 1f72b0c c7d476a 1088ad0 1f72b0c d446056 1088ad0 511d2ce |
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 |
import tempfile
import logging
import os
import asyncio
from moviepy.editor import *
import edge_tts
import gradio as gr
from pydub import AudioSegment
# Configuraci贸n de Logs
logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")
# CONSTANTES DE ARCHIVOS
INTRO_VIDEO = "introvideo.mp4"
OUTRO_VIDEO = "outrovideo.mp4"
MUSIC_BG = "musicafondo.mp3"
GLITCH_SOUND = "fxsound.mp3" # Efecto de sonido para glitches
EJEMPLO_VIDEO = "ejemplo.mp4"
# Validar existencia de archivos
for file in [INTRO_VIDEO, OUTRO_VIDEO, MUSIC_BG, GLITCH_SOUND, EJEMPLO_VIDEO]:
if not os.path.exists(file):
logging.error(f"Falta archivo necesario: {file}")
raise FileNotFoundError(f"Falta: {file}")
def eliminar_archivo_tiempo(ruta, delay=1800):
def eliminar():
try:
if os.path.exists(ruta):
os.remove(ruta)
logging.info(f"Archivo eliminado: {ruta}")
except Exception as e:
logging.error(f"Error al eliminar {ruta}: {e}")
from threading import Timer
Timer(delay, eliminar).start()
async def procesar_audio(texto, voz, duracion_video, audio_original):
temp_files = []
try:
# Validar texto
if not texto.strip():
raise ValueError("El texto para TTS no puede estar vac铆o.")
# Dividir el texto en fragmentos si es demasiado largo
def dividir_texto(texto, max_length=3000):
return [texto[i:i + max_length] for i in range(0, len(texto), max_length)]
fragmentos = dividir_texto(texto)
audios_tts = []
for fragmento in fragmentos:
# Generar TTS
communicate = edge_tts.Communicate(fragmento, voz)
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_tts:
try:
await communicate.save(tmp_tts.name)
except edge_tts.exceptions.NoAudioReceived as e:
logging.error(f"Error en TTS: {str(e)}")
raise ValueError("No se pudo generar el audio. Verifica tu conexi贸n o los par谩metros del TTS.")
tts_audio = AudioFileClip(tmp_tts.name)
temp_files.append(tmp_tts.name)
audios_tts.append(tts_audio)
# Combinar todos los fragmentos de TTS
tts_audio_final = concatenate_audioclips(audios_tts)
# Limitar TTS al video
if tts_audio_final.duration > duracion_video:
tts_audio_final = tts_audio_final.subclip(0, duracion_video)
# Preparar m煤sica de fondo en loop
bg_music = AudioSegment.from_mp3(MUSIC_BG)
needed_ms = int(duracion_video * 1000)
repeticiones = needed_ms // len(bg_music) + 1
bg_music = bg_music * repeticiones
bg_music = bg_music[:needed_ms].fade_out(1000)
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_bg:
bg_music.export(tmp_bg.name, format="mp3")
bg_audio = AudioFileClip(tmp_bg.name).volumex(0.15)
temp_files.append(tmp_bg.name)
# Combinar audios
audios = [bg_audio.set_duration(duracion_video)]
if audio_original:
audios.append(audio_original.volumex(0.7)) # Audio original al 70%
audios.append(tts_audio_final.volumex(0.85).set_start(0)) # TTS al 85%
audio_final = CompositeAudioClip(audios).set_duration(duracion_video)
return audio_final
except Exception as e:
logging.error(f" fallo en audio: {str(e)}")
raise
finally:
for file in temp_files:
try:
os.remove(file)
except Exception as e:
logging.warning(f"Error limpiando {file}: {e}")
def aplicar_glitch(video_clip):
"""Aplica un efecto de glitch al video."""
def glitch_effect(frame):
import numpy as np
height, width, _ = frame.shape
offset = np.random.randint(-10, 10)
if offset > 0:
offset = min(offset, height)
if offset < 0:
offset = max(offset, -height + 1)
if offset!= 0 and height > 0:
frame[offset:, :] = np.roll(frame[:-offset, :], -offset, axis=0)
return frame
return video_clip.fl_image(glitch_effect)
async def procesar_video(video_input, texto_tts, voz_seleccionada):
try:
# Cargar componentes
intro = VideoFileClip(INTRO_VIDEO)
outro = VideoFileClip(OUTRO_VIDEO)
video_original = VideoFileClip(video_input)
audio_original = video_original.audio
# Duraci贸n del video editado (sin intro/outro)
duracion_video = video_original.duration
# Procesar audio
audio_final = await procesar_audio(
texto_tts,
voz_seleccionada,
duracion_video,
audio_original
)
# Redimensionar todos los clips a 1920x1080
target_width = 1920
target_height = 1080
# Redimensionar intro
intro_resized = intro.resize((target_width, target_height))
# Redimensionar outro
outro_resized = outro.resize((target_width, target_height))
# Redimensionar video principal
video_resized = video_original.resize((target_width, target_height))
# Dividir el video en segmentos de 20 segundos y eliminar 2 segundos en cada corte
segment_duration = 20
overlap = 2 # Segundos a eliminar en cada corte
num_segments = int(duracion_video // (segment_duration - overlap)) + 1
segments = []
glitch_clips = []
glitch_sound = AudioFileClip(GLITCH_SOUND)
start_time = 0
for i in range(num_segments):
end_time = min(start_time + segment_duration, duracion_video)
if start_time >= duracion_video:
break
# Extraer el segmento
segment = video_resized.subclip(start_time, end_time)
# Aplicar glitch al inicio del segmento (excepto el primero)
if i > 0:
glitch_segment = aplicar_glitch(segment.subclip(0, 0.5)) # Glitch de 0.5 segundos
glitch_sound_clip = glitch_sound.set_start(start_time).volumex(0.5)
glitch_clips.append(glitch_sound_clip)
segment = concatenate_videoclips([glitch_segment, segment.subclip(0.5)], method="compose")
segments.append(segment)
# Avanzar al siguiente segmento, eliminando 2 segundos
start_time += segment_duration - overlap
# Combinar los segmentos procesados
video_final = concatenate_videoclips(segments)
# Combinar audio con efectos de glitch
audio_final = CompositeAudioClip([audio_final] + glitch_clips).set_duration(video_final.duration)
# Combinar video con audio
video_con_audio = video_final.set_audio(audio_final)
# Concatenar intro + video + outro SIN alteraciones
video_final = concatenate_videoclips(
[intro_resized, video_con_audio, outro_resized],
method="compose", # Evitar problemas de grid
padding=0 # Sin espacio entre clips
)
# Renderizar video final con metadatos correctos
with tempfile.NamedTemporaryFile(suffix=".mp4", delete=False) as tmp:
video_final.write_videofile(
tmp.name,
codec="libx264",
audio_codec="aac",
fps=video_original.fps, # Mantener FPS original
threads=4,
ffmpeg_params=[
"-aspect", "16:9", # Forzar relaci贸n de aspecto
"-vf", "scale=1920:1080" # Forzar escalado expl铆cito
],
verbose=False
)
eliminar_archivo_tiempo(tmp.name)
return tmp.name
except Exception as e:
logging.error(f" fallo general: {str(e)}")
raise
# Interfaz Gradio
with gr.Blocks() as demo:
gr.Markdown("# Editor de Video con IA")
with gr.Tab("Principal"):
video_input = gr.Video(label="Subir video")
texto_tts = gr.Textbox(
label="Texto para TTS",
lines=3,
placeholder="Escribe aqu铆 tu texto..."
)
voz_seleccionada = gr.Dropdown(
label="Voz",
choices=["es-ES-AlvaroNeural", "es-MX-BeatrizNeural"],
value="es-ES-AlvaroNeural"
)
procesar_btn = gr.Button("Generar Video")
video_output = gr.Video(label="Video Procesado")
with gr.Accordion("Ejemplos de Uso", open=False):
gr.Examples(
examples=[[EJEMPLO_VIDEO, "隆Hola! Esto es una prueba. Suscr铆bete al canal."]],
inputs=[video_input, texto_tts],
label="Ejemplos"
)
procesar_btn.click(
procesar_video,
inputs=[video_input, texto_tts, voz_seleccionada],
outputs=video_output
)
if __name__ == "__main__":
demo.queue().launch() |