Update app.py

app.py

@@ -139,12 +139,12 @@ def add_narration_to_video(narrations, input_video, output_dir, output_file, tex
     cap = cv2.VideoCapture(input_video)
     fourcc = cv2.VideoWriter_fourcc(*'XVID')
     temp_video = os.path.join(output_dir, "with_transcript.avi")
-    out = cv2.VideoWriter(temp_video, fourcc, 30, (int(cap.get(3)), int(cap.get(4))))
+    out = cv2.VideoWriter(temp_video, fourcc, 60, (int(cap.get(3)), int(cap.get(4))))
     full_narration = AudioSegment.empty()
     for i, narration in enumerate(narrations):
         audio = os.path.join(output_dir, "narrations", f"narration_{i+1}.wav")
         duration = get_audio_duration(audio)
-        narration_frames = math.floor(duration / 1000 * 30)
+        narration_frames = math.floor(duration / 2000 * 60)
         full_narration += AudioSegment.from_file(audio)
         char_count = len(narration.replace(" ", ""))
         ms_per_char = duration / char_count
@@ -156,7 +156,7 @@ def add_narration_to_video(narrations, input_video, output_dir, output_file, tex
                 word_ms -= offset
                 if word_ms < 0:
                     word_ms = 0
-            for _ in range(math.floor(word_ms/1000*30)):
+            for _ in range(math.floor(word_ms/2000*60)):
                 ret, frame = cap.read()
                 if not ret:
                     break
@@ -193,8 +193,8 @@ def add_narration_to_video(narrations, input_video, output_dir, output_file, tex
 
 def create_video(narrations, output_dir, output_file, text_color, text_position):  # Add text_color and text_position parameters here
     width, height = 1080, 1920
-    frame_rate = 30
-    fade_time = 1000
+    frame_rate = 60
+    fade_time = 2000
     fourcc = cv2.VideoWriter_fourcc(*'XVID')
     temp_video = os.path.join(output_dir, "temp_video.avi")
     out = cv2.VideoWriter(temp_video, fourcc, frame_rate, (width, height))
@@ -214,7 +214,7 @@ def create_video(narrations, output_dir, output_file, text_color, text_position)
             duration -= fade_time
         if i == image_count-1:
             duration -= fade_time
-        for _ in range(math.floor(duration/1000*30)):
+        for _ in range(math.floor(duration/2000*60)):
             vertical_video_frame = np.zeros((height, width, 3), dtype=np.uint8)
             vertical_video_frame[:image1.shape[0], :] = image1
             out.write(vertical_video_frame)
@@ -251,7 +251,7 @@ def generate_video(topic, voice_choice):
             },
             {
                 "role": "user",
-                "content": f"""make a short video on: \n\n{topic} Generate 30 seconds to 1 minute of video. You will need to generate a very short description of images for each of the sentences. They will be used for background images. Note that the script will be fed into a text-to-speech engine, so dont use special characters. Respond with a pair of an image description in square brackets and a script below it. Both of them should be on their own lines, as follows: ###
+                "content": f"""make a short video on: \n\n{topic} Generate 60 seconds to 1 minute of video. You will need to generate a very short description of images for each of the sentences. They will be used for background images. Note that the script will be fed into a text-to-speech engine, so dont use special characters. Respond with a pair of an image description in square brackets and a script below it. Both of them should be on their own lines, as follows: ###
 [Description of a background image]
 Narrator: "One sentence of narration"
 ### The short should be 6 sentences maximum."""