Zero-Overlap Linear Resolution Engine
Sorts cues chronologically and trims each colliding cue's end time cleanly to next.startMs − minGapMs, guaranteeing 0 remaining timestamp overlaps across the entire file.
cue[i].endMs ≤ cue[i+1].startMs − 42ms
Detect every cue whose end timestamp bleeds into the next cue and auto-resolve collisions with a clean configurable frame gap (default 42ms / 1 frame at 24fps).
Drop .srt, .vtt, .ass or .ssa file here
Multi-file drop supported · Processed 100% locally in your browser
Sorts cues chronologically and trims each colliding cue's end time cleanly to next.startMs − minGapMs, guaranteeing 0 remaining timestamp overlaps across the entire file.
cue[i].endMs ≤ cue[i+1].startMs − 42ms
Choose 0ms for back-to-back cuts, 42ms for 1 frame at 23.976/24fps, or 80–84ms for 2-frame Netflix shot-change breathing room.
1 frame @ 24fps ≈ 42ms | 2 frames ≈ 83ms
When two cues start at nearly the same millisecond (common in raw YouTube VTT captures), SubSyncer prevents zero-duration or negative-duration cues by cleanly spacing them.
Guaranteed endMs > startMs for every cue
When Cue 1 ends after Cue 2 begins, players stack both captions awkwardly, flash lines off screen, or reject the file completely during Premiere Pro, YouTube Studio, or OTT QC ingestion.
At 23.976 or 24 frames per second, one video frame lasts ~41.7 milliseconds. Leaving a 42ms gap creates a crisp 1-frame visual blink so viewers immediately notice the new line.
Never. SubSyncer enforces a strict minimum duration floor and resolves near-simultaneous cue collisions safely without inverting start and end timestamps.