heygen api video gen, timestamp clipping

audio_clips/clips_metadata.json

@@ -0,0 +1,194 @@
+{[
+    {
+        "clip_number": 0,
+        "start_time_ms": 0,
+        "end_time_ms": 316,
+        "file_name": "clip_0.wav"
+    },
+    {
+        "clip_number": 1,
+        "start_time_ms": 696,
+        "end_time_ms": 828,
+        "file_name": "clip_1.wav"
+    },
+    {
+        "clip_number": 2,
+        "start_time_ms": 1239,
+        "end_time_ms": 1391,
+        "file_name": "clip_2.wav"
+    },
+    {
+        "clip_number": 3,
+        "start_time_ms": 1822,
+        "end_time_ms": 1977,
+        "file_name": "clip_3.wav"
+    },
+    {
+        "clip_number": 4,
+        "start_time_ms": 2433,
+        "end_time_ms": 2661,
+        "file_name": "clip_4.wav"
+    },
+    {
+        "clip_number": 5,
+        "start_time_ms": 3067,
+        "end_time_ms": 3173,
+        "file_name": "clip_5.wav"
+    },
+    {
+        "clip_number": 7,
+        "start_time_ms": 3803,
+        "end_time_ms": 4103,
+        "file_name": "clip_7.wav"
+    },
+    {
+        "clip_number": 8,
+        "start_time_ms": 4450,
+        "end_time_ms": 4578,
+        "file_name": "clip_8.wav"
+    },
+    {
+        "clip_number": 9,
+        "start_time_ms": 5016,
+        "end_time_ms": 5394,
+        "file_name": "clip_9.wav"
+    },
+    {
+        "clip_number": 10,
+        "start_time_ms": 5772,
+        "end_time_ms": 5976,
+        "file_name": "clip_10.wav"
+    },
+    {
+        "clip_number": 11,
+        "start_time_ms": 6369,
+        "end_time_ms": 6665,
+        "file_name": "clip_11.wav"
+    },
+    {
+        "clip_number": 12,
+        "start_time_ms": 7080,
+        "end_time_ms": 7352,
+        "file_name": "clip_12.wav"
+    },
+    {
+        "clip_number": 13,
+        "start_time_ms": 7809,
+        "end_time_ms": 8056,
+        "file_name": "clip_13.wav"
+    },
+    {
+        "clip_number": 14,
+        "start_time_ms": 8503,
+        "end_time_ms": 8763,
+        "file_name": "clip_14.wav"
+    },
+    {
+        "clip_number": 15,
+        "start_time_ms": 9167,
+        "end_time_ms": 9532,
+        "file_name": "clip_15.wav"
+    },
+    {
+        "clip_number": 16,
+        "start_time_ms": 9951,
+        "end_time_ms": 10311,
+        "file_name": "clip_16.wav"
+    },
+    {
+        "clip_number": 17,
+        "start_time_ms": 10667,
+        "end_time_ms": 10900,
+        "file_name": "clip_17.wav"
+    },
+    {
+        "clip_number": 18,
+        "start_time_ms": 11242,
+        "end_time_ms": 11648,
+        "file_name": "clip_18.wav"
+    },
+    {
+        "clip_number": 19,
+        "start_time_ms": 11998,
+        "end_time_ms": 12374,
+        "file_name": "clip_19.wav"
+    },
+    {
+        "clip_number": 20,
+        "start_time_ms": 12782,
+        "end_time_ms": 13029,
+        "file_name": "clip_20.wav"
+    },
+    {
+        "clip_number": 21,
+        "start_time_ms": 13454,
+        "end_time_ms": 13805,
+        "file_name": "clip_21.wav"
+    },
+    {
+        "clip_number": 22,
+        "start_time_ms": 14243,
+        "end_time_ms": 14561,
+        "file_name": "clip_22.wav"
+    },
+    {
+        "clip_number": 23,
+        "start_time_ms": 14999,
+        "end_time_ms": 15351,
+        "file_name": "clip_23.wav"
+    },
+    {
+        "clip_number": 24,
+        "start_time_ms": 15744,
+        "end_time_ms": 16220,
+        "file_name": "clip_24.wav"
+    },
+    {
+        "clip_number": 25,
+        "start_time_ms": 16632,
+        "end_time_ms": 17093,
+        "file_name": "clip_25.wav"
+    },
+    {
+        "clip_number": 26,
+        "start_time_ms": 17443,
+        "end_time_ms": 17767,
+        "file_name": "clip_26.wav"
+    },
+    {
+        "clip_number": 27,
+        "start_time_ms": 18088,
+        "end_time_ms": 18564,
+        "file_name": "clip_27.wav"
+    },
+    {
+        "clip_number": 28,
+        "start_time_ms": 18970,
+        "end_time_ms": 19241,
+        "file_name": "clip_28.wav"
+    },
+    {
+        "clip_number": 29,
+        "start_time_ms": 19638,
+        "end_time_ms": 19986,
+        "file_name": "clip_29.wav"
+    },
+    {
+        "clip_number": 30,
+        "start_time_ms": 20328,
+        "end_time_ms": 20566,
+        "file_name": "clip_30.wav"
+    },
+    {
+        "clip_number": 31,
+        "start_time_ms": 21078,
+        "end_time_ms": 22167,
+        "file_name": "clip_31.wav"
+    },
+    {
+        "clip_number": 32,
+        "start_time_ms": 22752,
+        "end_time_ms": 23158,
+        "file_name": "clip_32.wav"
+    }
+]}

clips_metadata.json

@@ -0,0 +1,194 @@
+[
+    {
+        "clip_number": 0,
+        "start_time_ms": 0,
+        "end_time_ms": 356,
+        "file_name": "1_cut.mp4"
+    },
+    {
+        "clip_number": 1,
+        "start_time_ms": 656,
+        "end_time_ms": 868,
+        "file_name": "2_cut.mp4"
+    },
+    {
+        "clip_number": 2,
+        "start_time_ms": 1199,
+        "end_time_ms": 1431,
+        "file_name": "3_cut.mp4"
+    },
+    {
+        "clip_number": 3,
+        "start_time_ms": 1782,
+        "end_time_ms": 2017,
+        "file_name": "4_cut.mp4"
+    },
+    {
+        "clip_number": 4,
+        "start_time_ms": 2393,
+        "end_time_ms": 2701,
+        "file_name": "5_cut.mp4"
+    },
+    {
+        "clip_number": 5,
+        "start_time_ms": 3027,
+        "end_time_ms": 3213,
+        "file_name": "6_cut.mp4"
+    },
+    {
+        "clip_number": 7,
+        "start_time_ms": 3763,
+        "end_time_ms": 4143,
+        "file_name": "7_cut.mp4"
+    },
+    {
+        "clip_number": 8,
+        "start_time_ms": 4410,
+        "end_time_ms": 4618,
+        "file_name": "8_cut.mp4"
+    },
+    {
+        "clip_number": 9,
+        "start_time_ms": 4976,
+        "end_time_ms": 5434,
+        "file_name": "9_cut.mp4"
+    },
+    {
+        "clip_number": 10,
+        "start_time_ms": 5732,
+        "end_time_ms": 6016,
+        "file_name": "10_cut.mp4"
+    },
+    {
+        "clip_number": 11,
+        "start_time_ms": 6329,
+        "end_time_ms": 6705,
+        "file_name": "20_cut.mp4"
+    },
+    {
+        "clip_number": 12,
+        "start_time_ms": 7040,
+        "end_time_ms": 7392,
+        "file_name": "30_cut.mp4"
+    },
+    {
+        "clip_number": 13,
+        "start_time_ms": 7769,
+        "end_time_ms": 8096,
+        "file_name": "40_cut.mp4"
+    },
+    {
+        "clip_number": 14,
+        "start_time_ms": 8463,
+        "end_time_ms": 8803,
+        "file_name": "50_cut.mp4"
+    },
+    {
+        "clip_number": 15,
+        "start_time_ms": 9127,
+        "end_time_ms": 9572,
+        "file_name": "60_cut.mp4"
+    },
+    {
+        "clip_number": 16,
+        "start_time_ms": 9911,
+        "end_time_ms": 10351,
+        "file_name": "70_cut.mp4"
+    },
+    {
+        "clip_number": 17,
+        "start_time_ms": 10627,
+        "end_time_ms": 10940,
+        "file_name": "80_cut.mp4"
+    },
+    {
+        "clip_number": 18,
+        "start_time_ms": 11202,
+        "end_time_ms": 11688,
+        "file_name": "90_cut.mp4"
+    },
+    {
+        "clip_number": 19,
+        "start_time_ms": 11958,
+        "end_time_ms": 12414,
+        "file_name": "11_cut.mp4"
+    },
+    {
+        "clip_number": 20,
+        "start_time_ms": 12742,
+        "end_time_ms": 13069,
+        "file_name": "12_cut.mp4"
+    },
+    {
+        "clip_number": 21,
+        "start_time_ms": 13414,
+        "end_time_ms": 13845,
+        "file_name": "13_cut.mp4"
+    },
+    {
+        "clip_number": 22,
+        "start_time_ms": 14203,
+        "end_time_ms": 14601,
+        "file_name": "14_cut.mp4"
+    },
+    {
+        "clip_number": 23,
+        "start_time_ms": 14959,
+        "end_time_ms": 15391,
+        "file_name": "15_cut.mp4"
+    },
+    {
+        "clip_number": 24,
+        "start_time_ms": 15704,
+        "end_time_ms": 16260,
+        "file_name": "16_cut.mp4"
+    },
+    {
+        "clip_number": 25,
+        "start_time_ms": 16592,
+        "end_time_ms": 17133,
+        "file_name": "17_cut.mp4"
+    },
+    {
+        "clip_number": 26,
+        "start_time_ms": 17403,
+        "end_time_ms": 17807,
+        "file_name": "18_cut.mp4"
+    },
+    {
+        "clip_number": 27,
+        "start_time_ms": 18048,
+        "end_time_ms": 18604,
+        "file_name": "19_cut.mp4"
+    },
+    {
+        "clip_number": 28,
+        "start_time_ms": 18930,
+        "end_time_ms": 19281,
+        "file_name": "100_cut.mp4"
+    },
+    {
+        "clip_number": 29,
+        "start_time_ms": 19598,
+        "end_time_ms": 20026,
+        "file_name": "1000_cut.mp4"
+    },
+    {
+        "clip_number": 30,
+        "start_time_ms": 20288,
+        "end_time_ms": 20606,
+        "file_name": "100000_cut.mp4"
+    },
+    {
+        "clip_number": 31,
+        "start_time_ms": 21038,
+        "end_time_ms": 22207,
+        "file_name": "Intro_cut.mp4"
+    },
+    {
+        "clip_number": 32,
+        "start_time_ms": 22712,
+        "end_time_ms": 23198,
+        "file_name": "Conclusion_cut.mp4"
+    }
+]

process_audio.py

@@ -0,0 +1,92 @@
+from pydub import AudioSegment
+from pydub.silence import detect_silence
+import os
+import json
+
+def split_audio_on_silence(input_audio_path, output_dir, silence_thresh=-40, min_silence_len=120, keep_silence=30):
+    """
+    Splits an audio file into multiple clips when silence is detected and generates a JSON with timestamps.
+    
+    :param input_audio_path: Path to the input audio file
+    :param output_dir: Directory to save the output clips
+    :param silence_thresh: Silence threshold in dBFS (default: -40 dBFS)
+    :param min_silence_len: Minimum silence length to detect in milliseconds (default: 500 ms)
+    :param keep_silence: Milliseconds of silence to retain at the beginning and end of each clip (default: 200 ms)
+    """
+    # Load the audio file
+    audio = AudioSegment.from_file(input_audio_path)
+    
+    # Detect silences in the audio
+    silence_ranges = detect_silence(audio, min_silence_len=min_silence_len, silence_thresh=silence_thresh)
+    
+    # Adjust silence ranges to keep a bit of silence in each segment
+    adjusted_ranges = [(start - keep_silence, end + keep_silence) for start, end in silence_ranges]
+    adjusted_ranges = [(max(0, start), min(len(audio), end)) for start, end in adjusted_ranges]
+
+    # List to store clip metadata
+    clips_metadata = []
+
+    # If no silences are detected, save the whole audio as one clip
+    if not adjusted_ranges:
+        output_path = os.path.join(output_dir, "clip_0.wav")
+        audio.export(output_path, format="wav")
+        clips_metadata.append({
+            "clip_number": 0,
+            "start_time_ms": 0,
+            "end_time_ms": len(audio),
+            "file_name": "clip_0.wav"
+        })
+        print(f"No silences detected. Entire audio saved as {output_path}")
+        # Save metadata to JSON
+        with open(os.path.join(output_dir, "clips_metadata.json"), "w") as json_file:
+            json.dump(clips_metadata, json_file, indent=4)
+        return
+
+    # Split audio into clips based on silence
+    start_idx = 0
+    for i, (start, end) in enumerate(adjusted_ranges):
+        if start > start_idx:
+            clip = audio[start_idx:start]
+            file_name = f"clip_{i}.wav"
+            output_path = os.path.join(output_dir, file_name)
+            clip.export(output_path, format="wav")
+            print(f"Saved clip: {output_path}")
+            
+            # Add metadata for the clip
+            clips_metadata.append({
+                "clip_number": i,
+                "start_time_ms": start_idx,
+                "end_time_ms": start,
+                "file_name": file_name
+            })
+        start_idx = end
+
+    # Save the last segment after the final silence
+    if start_idx < len(audio):
+        file_name = f"clip_{len(adjusted_ranges)}.wav"
+        output_path = os.path.join(output_dir, file_name)
+        audio[start_idx:].export(output_path, format="wav")
+        print(f"Saved clip: {output_path}")
+        
+        # Add metadata for the last clip
+        clips_metadata.append({
+            "clip_number": len(adjusted_ranges),
+            "start_time_ms": start_idx,
+            "end_time_ms": len(audio),
+            "file_name": file_name
+        })
+
+    # Save metadata to JSON
+    json_path = os.path.join(output_dir, "clips_metadata.json")
+    with open(json_path, "w") as json_file:
+        json.dump(clips_metadata, json_file, indent=4)
+    print(f"Clip metadata saved to {json_path}")
+
+
+if __name__ == "__main__":
+    input_audio = "male_audio.m4a"
+    output_directory = "audio_clips"
+
+    os.makedirs(output_directory, exist_ok=True)
+
+    split_audio_on_silence(input_audio, output_directory)