Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -6,7 +6,7 @@ import subprocess
|
|
| 6 |
import shutil
|
| 7 |
import os
|
| 8 |
|
| 9 |
-
#
|
| 10 |
WORD_FADE_DURATION = 0.2
|
| 11 |
FFMPEG_ESCAPE_CHAR = "\\"
|
| 12 |
|
|
@@ -24,32 +24,22 @@ def get_font_path():
|
|
| 24 |
for font in possible_fonts:
|
| 25 |
if os.path.exists(font):
|
| 26 |
return font
|
| 27 |
-
return None
|
| 28 |
|
| 29 |
def save_temp_audio(audio_file_path):
|
| 30 |
-
"""
|
| 31 |
-
Speichert die hochgeladene Audio-Datei in einem temporären Verzeichnis.
|
| 32 |
-
Erwartet einen Dateipfad-String von Gradio.
|
| 33 |
-
"""
|
| 34 |
if not audio_file_path:
|
| 35 |
return None, None
|
| 36 |
-
|
| 37 |
-
# Gradio liefert einen String-Pfad zum temporären Speicherort
|
| 38 |
input_path = Path(audio_file_path)
|
| 39 |
-
|
| 40 |
-
# Bestimme die Erweiterung
|
| 41 |
ext = input_path.suffix
|
| 42 |
if ext.lower() not in allowed_audios:
|
| 43 |
ext = ".mp3"
|
| 44 |
|
| 45 |
-
# Erstelle das Zielverzeichnis und den Zielpfad
|
| 46 |
temp_audio_dir = Path(tempfile.mkdtemp())
|
| 47 |
temp_audio = temp_audio_dir / f"input{ext}"
|
| 48 |
|
| 49 |
-
# Kopiere die Datei vom Gradio-Temp-Pfad in unseren eigenen Temp-Pfad
|
| 50 |
try:
|
| 51 |
shutil.copyfile(input_path, temp_audio)
|
| 52 |
-
# Rückgabe des Verzeichnisses, das später gelöscht werden kann, und des Dateipfads
|
| 53 |
return temp_audio_dir, temp_audio
|
| 54 |
except Exception as e:
|
| 55 |
print(f"Fehler beim Kopieren der Audiodatei: {e}")
|
|
@@ -57,86 +47,118 @@ def save_temp_audio(audio_file_path):
|
|
| 57 |
shutil.rmtree(temp_audio_dir)
|
| 58 |
return None, None
|
| 59 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
|
| 61 |
-
|
| 62 |
-
"""Erstellt einen FFmpeg drawtext Filter, der ein Wort mit weichen Übergängen (Alpha-Kanal) einblendet,
|
| 63 |
-
basierend auf dem gewählten Stil."""
|
| 64 |
-
global FFMPEG_ESCAPE_CHAR
|
| 65 |
-
global WORD_FADE_DURATION
|
| 66 |
-
|
| 67 |
-
# 1. Escaping: Ersetze alle ":" durch "\:" für FFmpeg
|
| 68 |
-
escaped_word = word.replace(':', f"{FFMPEG_ESCAPE_CHAR}:")
|
| 69 |
|
| 70 |
-
#
|
| 71 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
|
| 73 |
-
#
|
| 74 |
-
|
| 75 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
|
| 77 |
# Alpha-Ausdruck für smooth Fade-In und Fade-Out
|
| 78 |
alpha_expression = (
|
| 79 |
f"if(lt(t,{start_time}), 0, "
|
| 80 |
-
f"if(lt(t,{
|
| 81 |
-
f"if(lt(t,{
|
| 82 |
-
f"if(lt(t,{
|
| 83 |
)
|
| 84 |
-
|
| 85 |
-
#
|
| 86 |
-
|
| 87 |
-
# Default-Werte (Modern-Stil)
|
| 88 |
params = {
|
| 89 |
-
"fontcolor": "
|
| 90 |
-
"borderw":
|
| 91 |
"bordercolor": "black",
|
| 92 |
-
"box": 0,
|
| 93 |
-
"boxcolor": "",
|
| 94 |
"fontsize_override": font_size
|
| 95 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
|
| 97 |
-
style_lower
|
| 98 |
-
|
| 99 |
-
if style_lower == "pop":
|
| 100 |
-
# Heller, auffälliger Text (Gelb mit Kontur, etwas größer)
|
| 101 |
params["fontcolor"] = "yellow"
|
| 102 |
-
params["borderw"] =
|
| 103 |
params["fontsize_override"] = font_size * 1.1
|
| 104 |
|
| 105 |
-
elif style_lower == "
|
| 106 |
-
#
|
| 107 |
-
params["fontcolor"] = "
|
| 108 |
-
params["borderw"] =
|
| 109 |
-
params["fontsize_override"] = font_size * 1.
|
| 110 |
|
| 111 |
elif style_lower == "badge":
|
| 112 |
-
#
|
| 113 |
params["fontcolor"] = "white"
|
| 114 |
params["borderw"] = 0
|
| 115 |
params["box"] = 1
|
| 116 |
-
|
| 117 |
-
params["
|
| 118 |
|
| 119 |
elif style_lower == "word":
|
| 120 |
-
#
|
| 121 |
-
params["fontcolor"] = "
|
| 122 |
params["borderw"] = 0
|
| 123 |
params["box"] = 1
|
| 124 |
-
|
| 125 |
-
params["
|
| 126 |
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
|
|
|
|
|
|
|
| 130 |
f"fontcolor={params['fontcolor']}:"
|
| 131 |
f"fontsize={params['fontsize_override']}:"
|
| 132 |
f"borderw={params['borderw']}:"
|
| 133 |
f"bordercolor={params['bordercolor']}:"
|
| 134 |
-
# Füge Box-Parameter nur hinzu, wenn box=1 (Badge- oder Word-Stil)
|
| 135 |
+ (f"box={params['box']}:boxcolor={params['boxcolor']}:boxborderw=10:" if params["box"] else "") +
|
| 136 |
-
f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}:"
|
| 137 |
f"alpha='{alpha_expression}'"
|
| 138 |
)
|
| 139 |
-
return drawtext_filter
|
| 140 |
|
| 141 |
|
| 142 |
def generate_slideshow_with_audio(images, input_text, duration_per_word, duration_per_image, fade_duration, font_size, y_pos, audio_file, subtitle_style):
|
|
@@ -156,7 +178,7 @@ def generate_slideshow_with_audio(images, input_text, duration_per_word, duratio
|
|
| 156 |
remainder = total_words % num_images
|
| 157 |
|
| 158 |
current_word_index = 0
|
| 159 |
-
clips_with_text = []
|
| 160 |
|
| 161 |
# Schriftart finden
|
| 162 |
font_path = get_font_path()
|
|
@@ -173,26 +195,31 @@ def generate_slideshow_with_audio(images, input_text, duration_per_word, duratio
|
|
| 173 |
|
| 174 |
# 1. Bestimme das Wortsegment für diesen Clip
|
| 175 |
words_on_this_clip = base_words_per_clip + (1 if i < remainder else 0)
|
| 176 |
-
|
| 177 |
-
# Extrahieren des Segments aus der Gesamtliste der Wörter
|
| 178 |
word_segment = words[current_word_index : current_word_index + words_on_this_clip]
|
| 179 |
current_word_index += len(word_segment)
|
| 180 |
|
|
|
|
|
|
|
| 181 |
# 2. Berechne die Clip-Dauer
|
| 182 |
text_duration = len(word_segment) * duration_per_word
|
| 183 |
-
# Die Dauer ist das Maximum aus der gewünschten Bilddauer und der benötigten Textdauer
|
| 184 |
duration_clip = max(duration_per_image, text_duration)
|
| 185 |
|
| 186 |
-
# 3. Generiere Drawtext Filter (Startzeit ist relativ zum Clip-Start, also 0)
|
| 187 |
drawtext_filters = []
|
| 188 |
-
|
| 189 |
-
|
| 190 |
-
#
|
| 191 |
-
|
| 192 |
-
drawtext_filters.append(
|
| 193 |
-
|
| 194 |
-
|
| 195 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 196 |
base_filters = (
|
| 197 |
"scale=w=1280:h=720:force_original_aspect_ratio=decrease,"
|
| 198 |
"pad=1280:720:(ow-iw)/2:(oh-ih)/2:color=black,"
|
|
@@ -203,7 +230,7 @@ def generate_slideshow_with_audio(images, input_text, duration_per_word, duratio
|
|
| 203 |
if fade_out_start < 0: fade_out_start = 0
|
| 204 |
fade_img_filter = f"fade=t=in:st=0:d={fade_duration},fade=t=out:st={fade_out_start}:d={fade_duration}"
|
| 205 |
|
| 206 |
-
#
|
| 207 |
if drawtext_filters:
|
| 208 |
all_drawtext_filters = ",".join(drawtext_filters)
|
| 209 |
vf_filters_clip = f"{base_filters},{all_drawtext_filters},{fade_img_filter}"
|
|
@@ -211,7 +238,7 @@ def generate_slideshow_with_audio(images, input_text, duration_per_word, duratio
|
|
| 211 |
# Kein Text mehr: Nur Bild mit Fade
|
| 212 |
vf_filters_clip = f"{base_filters},{fade_img_filter}"
|
| 213 |
|
| 214 |
-
#
|
| 215 |
cmd = [
|
| 216 |
"ffmpeg", "-y", "-loop", "1", "-i", str(img_path),
|
| 217 |
"-t", str(duration_clip),
|
|
@@ -280,18 +307,18 @@ with gr.Blocks() as demo:
|
|
| 280 |
|
| 281 |
with gr.Row():
|
| 282 |
img_input = gr.Files(label="Bilder", file_types=allowed_medias)
|
| 283 |
-
text_input = gr.Textbox(label="Text (Wörter werden gleichmäßig auf alle Bilder verteilt)", lines=5, placeholder="
|
| 284 |
|
| 285 |
with gr.Row():
|
| 286 |
duration_image_input = gr.Number(value=3, label="Mindest-Dauer pro BILD (s)")
|
| 287 |
-
duration_word_input = gr.Number(value=1.0, label="Dauer pro WORT (s) [bestimmt Geschwindigkeit der
|
| 288 |
fade_input = gr.Number(value=0.5, label="Bild-Fade Dauer (s)")
|
| 289 |
|
| 290 |
with gr.Row():
|
| 291 |
font_size_input = gr.Number(value=80, label="Schriftgröße (px)")
|
| 292 |
ypos_input = gr.Slider(0.0, 1.0, value=0.9, label="Y-Position (0=Oben, 1=Unten)")
|
| 293 |
|
| 294 |
-
#
|
| 295 |
subtitle_style_input = gr.Dropdown(
|
| 296 |
["Modern", "Pop", "Bold", "Badge", "Word"],
|
| 297 |
label="Untertitel-Stil",
|
|
@@ -305,8 +332,6 @@ with gr.Blocks() as demo:
|
|
| 305 |
out_video = gr.Video(label="Ergebnis")
|
| 306 |
status = gr.Textbox(label="Status")
|
| 307 |
|
| 308 |
-
# KORREKTE REIHENFOLGE DER INPUTS aktualisiert um 'subtitle_style_input':
|
| 309 |
-
# (images, input_text, duration_per_word, duration_per_image, fade_duration, font_size, y_pos, audio_file, subtitle_style)
|
| 310 |
btn.click(
|
| 311 |
fn=generate_slideshow_with_audio,
|
| 312 |
inputs=[
|
|
@@ -318,7 +343,7 @@ with gr.Blocks() as demo:
|
|
| 318 |
font_size_input,
|
| 319 |
ypos_input,
|
| 320 |
audio_input,
|
| 321 |
-
subtitle_style_input
|
| 322 |
],
|
| 323 |
outputs=[out_video, status]
|
| 324 |
)
|
|
|
|
| 6 |
import shutil
|
| 7 |
import os
|
| 8 |
|
| 9 |
+
# Konstanten
|
| 10 |
WORD_FADE_DURATION = 0.2
|
| 11 |
FFMPEG_ESCAPE_CHAR = "\\"
|
| 12 |
|
|
|
|
| 24 |
for font in possible_fonts:
|
| 25 |
if os.path.exists(font):
|
| 26 |
return font
|
| 27 |
+
return None
|
| 28 |
|
| 29 |
def save_temp_audio(audio_file_path):
|
| 30 |
+
"""Speichert die hochgeladene Audio-Datei in einem temporären Verzeichnis."""
|
|
|
|
|
|
|
|
|
|
| 31 |
if not audio_file_path:
|
| 32 |
return None, None
|
|
|
|
|
|
|
| 33 |
input_path = Path(audio_file_path)
|
|
|
|
|
|
|
| 34 |
ext = input_path.suffix
|
| 35 |
if ext.lower() not in allowed_audios:
|
| 36 |
ext = ".mp3"
|
| 37 |
|
|
|
|
| 38 |
temp_audio_dir = Path(tempfile.mkdtemp())
|
| 39 |
temp_audio = temp_audio_dir / f"input{ext}"
|
| 40 |
|
|
|
|
| 41 |
try:
|
| 42 |
shutil.copyfile(input_path, temp_audio)
|
|
|
|
| 43 |
return temp_audio_dir, temp_audio
|
| 44 |
except Exception as e:
|
| 45 |
print(f"Fehler beim Kopieren der Audiodatei: {e}")
|
|
|
|
| 47 |
shutil.rmtree(temp_audio_dir)
|
| 48 |
return None, None
|
| 49 |
|
| 50 |
+
def create_sentence_base_filter(full_text, duration_clip, font_option, font_size, y_pos, style):
|
| 51 |
+
"""
|
| 52 |
+
Erstellt den FFmpeg drawtext Filter für die Basisschicht (den gesamten Satzabschnitt),
|
| 53 |
+
der für die volle Clip-Dauer sichtbar ist.
|
| 54 |
+
"""
|
| 55 |
+
# Standard-Stil für inaktiven Text (Modern)
|
| 56 |
+
base_params = {
|
| 57 |
+
"fontcolor": "white",
|
| 58 |
+
"borderw": 2,
|
| 59 |
+
"bordercolor": "black",
|
| 60 |
+
"box": 0, "boxcolor": "",
|
| 61 |
+
"fontsize": font_size
|
| 62 |
+
}
|
| 63 |
|
| 64 |
+
style_lower = style.lower()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
+
# SPEZIALFALL: Pop Style (IMG_1456)
|
| 67 |
+
if style_lower == "pop":
|
| 68 |
+
# Ganzer Satzabschnitt in dunkler, semi-transparenter Box
|
| 69 |
+
base_params["box"] = 1
|
| 70 |
+
base_params["boxcolor"] = "[email protected]"
|
| 71 |
+
base_params["fontsize"] = font_size * 1.1
|
| 72 |
+
|
| 73 |
+
escaped_text = full_text.replace(':', FFMPEG_ESCAPE_CHAR + ':')
|
| 74 |
|
| 75 |
+
# Filter für den gesamten Satz, sichtbar für die gesamte Clip-Dauer
|
| 76 |
+
return (
|
| 77 |
+
f"drawtext=text='{escaped_text}':"
|
| 78 |
+
f"fontcolor={base_params['fontcolor']}:"
|
| 79 |
+
f"fontsize={base_params['fontsize']}:"
|
| 80 |
+
f"borderw={base_params['borderw']}:"
|
| 81 |
+
f"bordercolor={base_params['bordercolor']}:"
|
| 82 |
+
+ (f"box={base_params['box']}:boxcolor={base_params['boxcolor']}:boxborderw=10:" if base_params["box"] else "") +
|
| 83 |
+
f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}:{font_option}:"
|
| 84 |
+
f"enable='between(t, 0, {duration_clip})'"
|
| 85 |
+
)
|
| 86 |
+
|
| 87 |
+
|
| 88 |
+
def create_highlight_word_filter(word, full_text, start_time, duration, font_option, font_size, y_pos, style):
|
| 89 |
+
"""
|
| 90 |
+
Erstellt den FFmpeg drawtext Filter für die Highlight-Schicht (nur das aktive Wort).
|
| 91 |
+
Dieses Wort wird in der Mitte des Bildschirms angezeigt und überblendet die Basisschicht
|
| 92 |
+
während seiner Aktivierungszeit.
|
| 93 |
+
"""
|
| 94 |
+
word_end_time = start_time + duration
|
| 95 |
|
| 96 |
# Alpha-Ausdruck für smooth Fade-In und Fade-Out
|
| 97 |
alpha_expression = (
|
| 98 |
f"if(lt(t,{start_time}), 0, "
|
| 99 |
+
f"if(lt(t,{start_time + WORD_FADE_DURATION}), (t-{start_time})/{WORD_FADE_DURATION}, "
|
| 100 |
+
f"if(lt(t,{word_end_time - WORD_FADE_DURATION}), 1, "
|
| 101 |
+
f"if(lt(t,{word_end_time}), ({word_end_time}-t)/{WORD_FADE_DURATION}, 0))))"
|
| 102 |
)
|
| 103 |
+
|
| 104 |
+
# Styling Parameter (Standard: Gelb/Bold-Highlight)
|
|
|
|
|
|
|
| 105 |
params = {
|
| 106 |
+
"fontcolor": "yellow",
|
| 107 |
+
"borderw": 3,
|
| 108 |
"bordercolor": "black",
|
| 109 |
+
"box": 0, "boxcolor": "",
|
|
|
|
| 110 |
"fontsize_override": font_size
|
| 111 |
}
|
| 112 |
+
|
| 113 |
+
style_lower = style.lower()
|
| 114 |
+
|
| 115 |
+
if style_lower == "modern":
|
| 116 |
+
# Modern: Minimaler Highlight (etwas größer)
|
| 117 |
+
params["fontcolor"] = "white"
|
| 118 |
+
params["borderw"] = 2
|
| 119 |
+
params["fontsize_override"] = font_size * 1.05
|
| 120 |
|
| 121 |
+
elif style_lower == "bold":
|
| 122 |
+
# Bold: Gelb mit starkem Rand (wie in IMG_1455)
|
|
|
|
|
|
|
| 123 |
params["fontcolor"] = "yellow"
|
| 124 |
+
params["borderw"] = 4
|
| 125 |
params["fontsize_override"] = font_size * 1.1
|
| 126 |
|
| 127 |
+
elif style_lower == "pop":
|
| 128 |
+
# Pop: Gelbe Schrift, kein Rand (Box wird von der Basisschicht gezeichnet)
|
| 129 |
+
params["fontcolor"] = "yellow"
|
| 130 |
+
params["borderw"] = 0
|
| 131 |
+
params["fontsize_override"] = font_size * 1.1
|
| 132 |
|
| 133 |
elif style_lower == "badge":
|
| 134 |
+
# Badge: Mint Green Box um das Wort (wie in IMG_1453)
|
| 135 |
params["fontcolor"] = "white"
|
| 136 |
params["borderw"] = 0
|
| 137 |
params["box"] = 1
|
| 138 |
+
params["boxcolor"] = "[email protected]" # Mint Green
|
| 139 |
+
params["fontsize_override"] = font_size * 1.05
|
| 140 |
|
| 141 |
elif style_lower == "word":
|
| 142 |
+
# Word: Gelbe Box um das Wort (wie in IMG_1454)
|
| 143 |
+
params["fontcolor"] = "black" # Textfarbe auf schwarz, da Box gelb ist
|
| 144 |
params["borderw"] = 0
|
| 145 |
params["box"] = 1
|
| 146 |
+
params["boxcolor"] = "[email protected]" # Yellow
|
| 147 |
+
params["fontsize_override"] = font_size * 1.05
|
| 148 |
|
| 149 |
+
escaped_word = word.replace(':', FFMPEG_ESCAPE_CHAR + ':')
|
| 150 |
+
|
| 151 |
+
# Filter für das einzelne, hervorgehobene Wort
|
| 152 |
+
return (
|
| 153 |
+
f"drawtext=text='{escaped_word}':"
|
| 154 |
f"fontcolor={params['fontcolor']}:"
|
| 155 |
f"fontsize={params['fontsize_override']}:"
|
| 156 |
f"borderw={params['borderw']}:"
|
| 157 |
f"bordercolor={params['bordercolor']}:"
|
|
|
|
| 158 |
+ (f"box={params['box']}:boxcolor={params['boxcolor']}:boxborderw=10:" if params["box"] else "") +
|
| 159 |
+
f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}:{font_option}:"
|
| 160 |
f"alpha='{alpha_expression}'"
|
| 161 |
)
|
|
|
|
| 162 |
|
| 163 |
|
| 164 |
def generate_slideshow_with_audio(images, input_text, duration_per_word, duration_per_image, fade_duration, font_size, y_pos, audio_file, subtitle_style):
|
|
|
|
| 178 |
remainder = total_words % num_images
|
| 179 |
|
| 180 |
current_word_index = 0
|
| 181 |
+
clips_with_text = []
|
| 182 |
|
| 183 |
# Schriftart finden
|
| 184 |
font_path = get_font_path()
|
|
|
|
| 195 |
|
| 196 |
# 1. Bestimme das Wortsegment für diesen Clip
|
| 197 |
words_on_this_clip = base_words_per_clip + (1 if i < remainder else 0)
|
|
|
|
|
|
|
| 198 |
word_segment = words[current_word_index : current_word_index + words_on_this_clip]
|
| 199 |
current_word_index += len(word_segment)
|
| 200 |
|
| 201 |
+
full_text = " ".join(word_segment)
|
| 202 |
+
|
| 203 |
# 2. Berechne die Clip-Dauer
|
| 204 |
text_duration = len(word_segment) * duration_per_word
|
|
|
|
| 205 |
duration_clip = max(duration_per_image, text_duration)
|
| 206 |
|
|
|
|
| 207 |
drawtext_filters = []
|
| 208 |
+
|
| 209 |
+
if full_text:
|
| 210 |
+
# ERSTE SCHICHT: Der gesamte Satz (als Basis)
|
| 211 |
+
base_filter = create_sentence_base_filter(full_text, duration_clip, font_option, font_size, y_pos, subtitle_style)
|
| 212 |
+
drawtext_filters.append(base_filter)
|
| 213 |
+
|
| 214 |
+
# ZWEITE SCHICHT: Highlight-Layer für jedes Wort
|
| 215 |
+
word_start_time = 0.0
|
| 216 |
+
for word in word_segment:
|
| 217 |
+
highlight_filter = create_highlight_word_filter(word, full_text, word_start_time, duration_per_word, font_option, font_size, y_pos, subtitle_style)
|
| 218 |
+
drawtext_filters.append(highlight_filter)
|
| 219 |
+
word_start_time += duration_per_word
|
| 220 |
+
|
| 221 |
+
|
| 222 |
+
# 3. Basis- und Fade-Filter
|
| 223 |
base_filters = (
|
| 224 |
"scale=w=1280:h=720:force_original_aspect_ratio=decrease,"
|
| 225 |
"pad=1280:720:(ow-iw)/2:(oh-ih)/2:color=black,"
|
|
|
|
| 230 |
if fade_out_start < 0: fade_out_start = 0
|
| 231 |
fade_img_filter = f"fade=t=in:st=0:d={fade_duration},fade=t=out:st={fade_out_start}:d={fade_duration}"
|
| 232 |
|
| 233 |
+
# 4. Kombiniere alle Filter
|
| 234 |
if drawtext_filters:
|
| 235 |
all_drawtext_filters = ",".join(drawtext_filters)
|
| 236 |
vf_filters_clip = f"{base_filters},{all_drawtext_filters},{fade_img_filter}"
|
|
|
|
| 238 |
# Kein Text mehr: Nur Bild mit Fade
|
| 239 |
vf_filters_clip = f"{base_filters},{fade_img_filter}"
|
| 240 |
|
| 241 |
+
# 5. FFmpeg Command zum Erstellen des Clips
|
| 242 |
cmd = [
|
| 243 |
"ffmpeg", "-y", "-loop", "1", "-i", str(img_path),
|
| 244 |
"-t", str(duration_clip),
|
|
|
|
| 307 |
|
| 308 |
with gr.Row():
|
| 309 |
img_input = gr.Files(label="Bilder", file_types=allowed_medias)
|
| 310 |
+
text_input = gr.Textbox(label="Text (Wörter werden gleichmäßig auf alle Bilder verteilt)", lines=5, placeholder="Der Satzabschnitt ist pro Clip sichtbar. Das aktive Wort wird hervorgehoben.")
|
| 311 |
|
| 312 |
with gr.Row():
|
| 313 |
duration_image_input = gr.Number(value=3, label="Mindest-Dauer pro BILD (s)")
|
| 314 |
+
duration_word_input = gr.Number(value=1.0, label="Dauer pro WORT (s) [bestimmt Geschwindigkeit der Hervorhebung]")
|
| 315 |
fade_input = gr.Number(value=0.5, label="Bild-Fade Dauer (s)")
|
| 316 |
|
| 317 |
with gr.Row():
|
| 318 |
font_size_input = gr.Number(value=80, label="Schriftgröße (px)")
|
| 319 |
ypos_input = gr.Slider(0.0, 1.0, value=0.9, label="Y-Position (0=Oben, 1=Unten)")
|
| 320 |
|
| 321 |
+
# Untertitel-Stile
|
| 322 |
subtitle_style_input = gr.Dropdown(
|
| 323 |
["Modern", "Pop", "Bold", "Badge", "Word"],
|
| 324 |
label="Untertitel-Stil",
|
|
|
|
| 332 |
out_video = gr.Video(label="Ergebnis")
|
| 333 |
status = gr.Textbox(label="Status")
|
| 334 |
|
|
|
|
|
|
|
| 335 |
btn.click(
|
| 336 |
fn=generate_slideshow_with_audio,
|
| 337 |
inputs=[
|
|
|
|
| 343 |
font_size_input,
|
| 344 |
ypos_input,
|
| 345 |
audio_input,
|
| 346 |
+
subtitle_style_input
|
| 347 |
],
|
| 348 |
outputs=[out_video, status]
|
| 349 |
)
|