feat: extract audio alongside WebP image sequence

Adds build_audio_extract_command and runs it in ExportWorker after the
frame sequence completes. Audio written to <sequence_dir>.wav (lossless
pcm_s16le). Extraction failure (no audio stream) is silently ignored.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

main.py

@@ -77,6 +77,20 @@ def build_ffmpeg_command(
     return cmd
 
 
+def build_audio_extract_command(input_path: str, start: float, sequence_dir: str) -> list[str]:
+    """Return an ffmpeg command that extracts audio to <sequence_dir>.wav."""
+    audio_path = sequence_dir + ".wav"
+    return [
+        "ffmpeg", "-y",
+        "-ss", str(start),
+        "-i", input_path,
+        "-t", "8",
+        "-vn",
+        "-c:a", "pcm_s16le",
+        audio_path,
+    ]
+
+
 def build_mask_output_dir(video_path: str) -> str:
     """Return path of mask output directory: <stem>_masks/ next to the video."""
     p = Path(video_path)
@@ -232,6 +246,13 @@ class ExportWorker(QThread):
             )
             result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
             if result.returncode == 0:
+                if self._image_sequence:
+                    audio_cmd = build_audio_extract_command(
+                        self._input, self._start, self._output
+                    )
+                    subprocess.run(audio_cmd, capture_output=True, text=True, timeout=60)
+                    # Audio extraction failure (e.g. no audio stream) is ignored —
+                    # the frame sequence is the primary output.
                 self.finished.emit(self._output)
             else:
                 self.error.emit(result.stderr[-500:])

tests/test_utils.py

@@ -1,5 +1,5 @@
 import tempfile, os
-from main import build_export_path, format_time, build_ffmpeg_command, build_mask_output_dir, build_sequence_dir
+from main import build_export_path, format_time, build_ffmpeg_command, build_mask_output_dir, build_sequence_dir, build_audio_extract_command
 from main import _normalize_filename, ProcessedDB
 
 
@@ -184,6 +184,29 @@ def test_mask_output_dir_nested():
     assert build_mask_output_dir("/a/b/c/shot_042.mp4") == "/a/b/c/shot_042_masks"
 
 
+# --- build_audio_extract_command ---
+
+def test_audio_extract_output_path():
+    cmd = build_audio_extract_command("/in/v.mp4", 0.0, "/out/clip_001")
+    assert cmd[-1] == "/out/clip_001.wav"
+
+def test_audio_extract_no_video():
+    cmd = build_audio_extract_command("/in/v.mp4", 0.0, "/out/clip_001")
+    assert "-vn" in cmd
+
+def test_audio_extract_lossless_codec():
+    cmd = build_audio_extract_command("/in/v.mp4", 0.0, "/out/clip_001")
+    assert "-c:a" in cmd
+    assert cmd[cmd.index("-c:a") + 1] == "pcm_s16le"
+
+def test_audio_extract_timing():
+    cmd = build_audio_extract_command("/in/v.mp4", 12.5, "/out/clip_001")
+    assert "-ss" in cmd
+    assert cmd[cmd.index("-ss") + 1] == "12.5"
+    assert "-t" in cmd
+    assert cmd[cmd.index("-t") + 1] == "8"
+
+
 def test_build_sequence_dir_basic():
     assert build_sequence_dir("/out", "clip", 1) == "/out/clip_001"