Implement sync_logic.py and ffmpeg_export.py (build order steps 2-3)

sync_logic.py: pure delay_ms computation with the documented sign
convention. Caught and fixed a sign error in my own test-fixture
ground-truth doc while writing the unit tests: the audio+350ms fixture
actually requires delay_ms=-350 (delay video), not +350, since the
audio is already the late one.

ffmpeg_export.py: ports ffdelay's two-branch command structure
(offset video input vs audio input depending on delay_ms sign) plus
its documented quirks (leading-zero float formatting, -ss/-t applied
to both inputs for clip mode, non-fatal stream-copy warnings surfaced
via stderr rather than swallowed).

Added an end-to-end integration test (scripts/test_export_roundtrip.py)
that independently re-detects event timing in the exported output via
ffmpeg scene-change/silence detection (not the same code path used to
build the export) and confirms both the positive- and negative-delay
branches bring drift down from ~350ms to under 1ms.

tests/test_sync_logic.py

@@ -0,0 +1,38 @@
+import sys
+import unittest
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+from sync_logic import compute_offset
+
+
+class TestComputeOffset(unittest.TestCase):
+    def test_audio_ahead_needs_delay(self):
+        # Flash seen at 5.000s, beep heard at 4.500s -> audio is running
+        # ahead of video, so audio needs to be delayed to catch up.
+        offset = compute_offset(video_mark_sec=5.000, audio_mark_sec=4.500)
+        self.assertAlmostEqual(offset.delay_ms, 500.0)
+        self.assertEqual(offset.target, "audio")
+
+    def test_audio_behind_needs_video_delay(self):
+        # This is our synthetic test fixture: audio was shifted +350ms
+        # later than video, so the fix is to delay video to catch up to
+        # the already-late audio -- NOT to delay audio further.
+        offset = compute_offset(video_mark_sec=5.000, audio_mark_sec=5.350)
+        self.assertAlmostEqual(offset.delay_ms, -350.0)
+        self.assertEqual(offset.target, "video")
+
+    def test_already_in_sync(self):
+        offset = compute_offset(video_mark_sec=5.000, audio_mark_sec=5.000)
+        self.assertEqual(offset.delay_ms, 0.0)
+        self.assertEqual(offset.target, "none")
+
+    def test_describe_formatting(self):
+        self.assertIn("audio will be delayed", compute_offset(5.5, 5.0).describe())
+        self.assertIn("video will be delayed", compute_offset(5.0, 5.5).describe())
+        self.assertIn("+500", compute_offset(5.5, 5.0).describe())
+        self.assertIn("-500", compute_offset(5.0, 5.5).describe())
+
+
+if __name__ == "__main__":
+    unittest.main()