Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -14,11 +14,10 @@ logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(
|
|
14 |
INTRO_VIDEO = "introvideo.mp4"
|
15 |
OUTRO_VIDEO = "outrovideo.mp4"
|
16 |
MUSIC_BG = "musicafondo.mp3"
|
17 |
-
GLITCH_SOUND = "fxsound.mp3"
|
18 |
EJEMPLO_VIDEO = "ejemplo.mp4"
|
19 |
|
20 |
# Validar existencia de archivos
|
21 |
-
for file in [INTRO_VIDEO, OUTRO_VIDEO, MUSIC_BG,
|
22 |
if not os.path.exists(file):
|
23 |
logging.error(f"Falta archivo necesario: {file}")
|
24 |
raise FileNotFoundError(f"Falta: {file}")
|
@@ -38,50 +37,26 @@ def eliminar_archivo_tiempo(ruta, delay=1800):
|
|
38 |
from threading import Timer
|
39 |
Timer(delay, eliminar).start()
|
40 |
|
41 |
-
async def
|
42 |
-
temp_files = []
|
43 |
try:
|
44 |
-
logging.info("
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
for fragmento in fragmentos:
|
56 |
-
communicate = edge_tts.Communicate(fragmento, voz)
|
57 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_tts:
|
58 |
-
await communicate.save(tmp_tts.name)
|
59 |
-
tts_audio = AudioFileClip(tmp_tts.name)
|
60 |
-
temp_files.append(tmp_tts.name)
|
61 |
-
audios_tts.append(tts_audio)
|
62 |
-
|
63 |
-
tts_audio_final = concatenate_audioclips(audios_tts)
|
64 |
-
|
65 |
-
if tts_audio_final.duration > duracion_video:
|
66 |
-
tts_audio_final = tts_audio_final.subclip(0, duracion_video)
|
67 |
-
|
68 |
-
logging.info("TTS procesado exitosamente")
|
69 |
-
return tts_audio_final, temp_files
|
70 |
-
|
71 |
except Exception as e:
|
72 |
-
logging.error(f"Fallo en
|
73 |
raise
|
74 |
-
finally:
|
75 |
-
for file in temp_files:
|
76 |
-
try:
|
77 |
-
os.remove(file)
|
78 |
-
except Exception as e:
|
79 |
-
logging.warning(f"Error limpiando {file}: {e}")
|
80 |
|
81 |
-
def crear_musica_fondo(
|
82 |
-
"""Crea un loop continuo de m煤sica de fondo."""
|
83 |
bg_music = AudioSegment.from_mp3(MUSIC_BG)
|
84 |
-
needed_ms = int(
|
85 |
repeticiones = needed_ms // len(bg_music) + 1
|
86 |
bg_music = bg_music * repeticiones
|
87 |
bg_music = bg_music[:needed_ms].fade_out(1000)
|
@@ -90,91 +65,51 @@ def crear_musica_fondo(duracion_video):
|
|
90 |
bg_music.export(tmp_bg.name, format="mp3")
|
91 |
return AudioFileClip(tmp_bg.name).volumex(0.15), tmp_bg.name
|
92 |
|
93 |
-
async def procesar_fragmento(chunk, texto_tts, voz_seleccionada, start_time):
|
94 |
-
try:
|
95 |
-
duracion_chunk = chunk.duration
|
96 |
-
|
97 |
-
# Procesar TTS para este chunk
|
98 |
-
tts_audio_final, tts_temp_files = await procesar_audio_tts(
|
99 |
-
texto_tts,
|
100 |
-
voz_seleccionada,
|
101 |
-
duracion_chunk
|
102 |
-
)
|
103 |
-
|
104 |
-
# Crear m煤sica de fondo continua
|
105 |
-
bg_audio, bg_temp_file = crear_musica_fondo(duracion_chunk)
|
106 |
-
|
107 |
-
# Combinar pistas de audio
|
108 |
-
audio_original = chunk.audio
|
109 |
-
audios = [bg_audio.set_duration(duracion_chunk)]
|
110 |
-
if audio_original:
|
111 |
-
audios.append(audio_original.volumex(0.7))
|
112 |
-
audios.append(tts_audio_final.volumex(0.85).set_start(0))
|
113 |
-
|
114 |
-
audio_final = CompositeAudioClip(audios).set_duration(duracion_chunk)
|
115 |
-
|
116 |
-
# Dividir el chunk en segmentos con cortes de 2 segundos
|
117 |
-
segment_duration = 18
|
118 |
-
overlap = 2
|
119 |
-
segments = []
|
120 |
-
|
121 |
-
current_time = 0
|
122 |
-
while current_time < duracion_chunk:
|
123 |
-
end_time = current_time + segment_duration
|
124 |
-
end_time = min(end_time, duracion_chunk)
|
125 |
-
|
126 |
-
full_segment = chunk.subclip(current_time, end_time)
|
127 |
-
segments.append(full_segment)
|
128 |
-
current_time += (segment_duration - overlap)
|
129 |
-
|
130 |
-
# Asegurar que haya al menos un segmento
|
131 |
-
if not segments:
|
132 |
-
logging.warning("Chunk demasiado corto, devolviendo el chunk original.")
|
133 |
-
video_chunk = chunk.set_audio(audio_final)
|
134 |
-
else:
|
135 |
-
video_chunk = concatenate_videoclips(segments, method="compose")
|
136 |
-
video_chunk = video_chunk.set_audio(audio_final)
|
137 |
-
|
138 |
-
return video_chunk, tts_temp_files + [bg_temp_file]
|
139 |
-
|
140 |
-
except Exception as e:
|
141 |
-
logging.error(f"Fallo procesando fragmento: {str(e)}")
|
142 |
-
raise
|
143 |
-
|
144 |
async def procesar_video(video_input, texto_tts, voz_seleccionada):
|
145 |
temp_files = []
|
146 |
-
intro, outro = None, None
|
147 |
try:
|
148 |
-
logging.info("Iniciando procesamiento
|
149 |
video_original = VideoFileClip(video_input, target_resolution=(720, 1280))
|
150 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
151 |
|
152 |
-
# Dividir en chunks
|
153 |
chunks = []
|
154 |
-
for start in range(0, int(
|
155 |
-
end = min(start + CHUNK_SIZE,
|
156 |
chunk = video_original.subclip(start, end)
|
157 |
-
chunks.append(
|
158 |
|
159 |
# Procesar cada chunk
|
160 |
processed_clips = []
|
161 |
-
for
|
162 |
-
|
163 |
-
processed_chunk, chunk_temp_files = await procesar_fragmento(chunk, texto_tts, voz_seleccionada, start_time)
|
164 |
-
processed_clips.append(processed_chunk)
|
165 |
-
temp_files.extend(chunk_temp_files)
|
166 |
|
167 |
-
# Combinar chunks
|
168 |
-
|
|
|
169 |
|
170 |
# Agregar intro y outro
|
171 |
intro = VideoFileClip(INTRO_VIDEO, target_resolution=(720, 1280))
|
172 |
outro = VideoFileClip(OUTRO_VIDEO, target_resolution=(720, 1280))
|
173 |
-
|
174 |
|
175 |
# Renderizado final
|
176 |
with tempfile.NamedTemporaryFile(suffix=".mp4", delete=False) as tmp:
|
177 |
-
|
178 |
tmp.name,
|
179 |
codec="libx264",
|
180 |
audio_codec="aac",
|
|
|
14 |
INTRO_VIDEO = "introvideo.mp4"
|
15 |
OUTRO_VIDEO = "outrovideo.mp4"
|
16 |
MUSIC_BG = "musicafondo.mp3"
|
|
|
17 |
EJEMPLO_VIDEO = "ejemplo.mp4"
|
18 |
|
19 |
# Validar existencia de archivos
|
20 |
+
for file in [INTRO_VIDEO, OUTRO_VIDEO, MUSIC_BG, EJEMPLO_VIDEO]:
|
21 |
if not os.path.exists(file):
|
22 |
logging.error(f"Falta archivo necesario: {file}")
|
23 |
raise FileNotFoundError(f"Falta: {file}")
|
|
|
37 |
from threading import Timer
|
38 |
Timer(delay, eliminar).start()
|
39 |
|
40 |
+
async def generar_tts(texto, voz, duracion_total):
|
|
|
41 |
try:
|
42 |
+
logging.info("Generando TTS")
|
43 |
+
communicate = edge_tts.Communicate(texto, voz)
|
44 |
+
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_tts:
|
45 |
+
await communicate.save(tmp_tts.name)
|
46 |
+
tts_audio = AudioFileClip(tmp_tts.name)
|
47 |
+
|
48 |
+
# Asegurar que el TTS no exceda la duraci贸n del video
|
49 |
+
if tts_audio.duration > duracion_total:
|
50 |
+
tts_audio = tts_audio.subclip(0, duracion_total)
|
51 |
+
|
52 |
+
return tts_audio, tmp_tts.name
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
except Exception as e:
|
54 |
+
logging.error(f"Fallo en TTS: {str(e)}")
|
55 |
raise
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
|
57 |
+
def crear_musica_fondo(duracion_total):
|
|
|
58 |
bg_music = AudioSegment.from_mp3(MUSIC_BG)
|
59 |
+
needed_ms = int(duracion_total * 1000)
|
60 |
repeticiones = needed_ms // len(bg_music) + 1
|
61 |
bg_music = bg_music * repeticiones
|
62 |
bg_music = bg_music[:needed_ms].fade_out(1000)
|
|
|
65 |
bg_music.export(tmp_bg.name, format="mp3")
|
66 |
return AudioFileClip(tmp_bg.name).volumex(0.15), tmp_bg.name
|
67 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
68 |
async def procesar_video(video_input, texto_tts, voz_seleccionada):
|
69 |
temp_files = []
|
70 |
+
intro, outro, video_original = None, None, None
|
71 |
try:
|
72 |
+
logging.info("Iniciando procesamiento")
|
73 |
video_original = VideoFileClip(video_input, target_resolution=(720, 1280))
|
74 |
+
duracion_video = video_original.duration
|
75 |
+
|
76 |
+
# Generar TTS y m煤sica de fondo para todo el video
|
77 |
+
tts_audio, tts_path = await generar_tts(texto_tts, voz_seleccionada, duracion_video)
|
78 |
+
bg_audio, bg_path = crear_musica_fondo(duracion_video)
|
79 |
+
temp_files.extend([tts_path, bg_path])
|
80 |
+
|
81 |
+
# Combinar audios
|
82 |
+
audio_original = video_original.audio.volumex(0.7) if video_original.audio else None
|
83 |
+
audios = [bg_audio.set_duration(duracion_video)]
|
84 |
+
if audio_original:
|
85 |
+
audios.append(audio_original)
|
86 |
+
audios.append(tts_audio.set_start(0).volumex(0.85))
|
87 |
+
audio_final = CompositeAudioClip(audios).set_duration(duracion_video)
|
88 |
|
89 |
+
# Dividir video en chunks para procesamiento
|
90 |
chunks = []
|
91 |
+
for start in range(0, int(duracion_video), CHUNK_SIZE):
|
92 |
+
end = min(start + CHUNK_SIZE, duracion_video)
|
93 |
chunk = video_original.subclip(start, end)
|
94 |
+
chunks.append(chunk)
|
95 |
|
96 |
# Procesar cada chunk
|
97 |
processed_clips = []
|
98 |
+
for chunk in chunks:
|
99 |
+
processed_clips.append(chunk)
|
|
|
|
|
|
|
100 |
|
101 |
+
# Combinar chunks (sin efectos)
|
102 |
+
video_final = concatenate_videoclips(processed_clips, method="compose")
|
103 |
+
video_final = video_final.set_audio(audio_final)
|
104 |
|
105 |
# Agregar intro y outro
|
106 |
intro = VideoFileClip(INTRO_VIDEO, target_resolution=(720, 1280))
|
107 |
outro = VideoFileClip(OUTRO_VIDEO, target_resolution=(720, 1280))
|
108 |
+
video_final = concatenate_videoclips([intro, video_final, outro], method="compose")
|
109 |
|
110 |
# Renderizado final
|
111 |
with tempfile.NamedTemporaryFile(suffix=".mp4", delete=False) as tmp:
|
112 |
+
video_final.write_videofile(
|
113 |
tmp.name,
|
114 |
codec="libx264",
|
115 |
audio_codec="aac",
|