Highlight Window Resolution Algorithm

Convert per-slide audio durations into deterministic frame windows, then map highlight slide ranges onto that timeline.

Step 1: Build Slide Timeline

Use fps=30 and add one second padding per slide:

audioFrames = ceil((audioDurationMs / 1000) * fps)
slideDurationFrames = audioFrames + fps

Each timeline entry stores:

• startFrame, endFrame
• startMs, endMs, durationMs
• source audioDurationMs

Step 2: Normalize Highlight Ranges

Allowed shapes:

• [n]
• [n, n+1]

Reject:

• empty ranges
• negative indexes
• non-consecutive two-slide ranges

Step 3: Resolve or Drop

For each highlight:

• Find first and last slide in timeline.
• If out of bounds, record in dropped with reason.
• Else produce window with start/end frame and millisecond bounds.

Why This Pattern

• Deterministic clip extraction boundaries.
• Simple validation surface.
• Clear audit trail for dropped highlight definitions.