Quality
Controls how much visual detail is preserved versus how small the output file becomes. Under the hood this maps to x264's Constant Rate Factor (CRF), a mathematical target for the encoder.
- High quality (CRF 18) — near-lossless output, larger file. Choose this when you'll be projecting the output for teaching, or when fine visual details matter (fluorescein staining patterns, subtle vessel appearance, granular corneal changes).
- Balanced (CRF 23) — visually near-transparent for most clinical footage. Sensible file sizes. Right choice for day-to-day conversion.
- Compact (CRF 27) — noticeably more compression, smaller files. Useful for long-form recordings or rough previews where size matters more than pixel-peep quality.
Each step is roughly a doubling of file size in either direction. Lower CRF numbers = higher quality; the scale runs 0 (lossless) to 51 (unwatchable).
Speed
Controls how much CPU time the encoder spends looking for compression opportunities. It does not change target quality — that's what Quality does. It changes how efficiently the encoder reaches that target.
- Fast — quickest encode, ~30–40% larger file at the same Quality setting. Good for quick turnarounds.
- Balanced — sensible middle ground. Reasonable encode time, close-to-optimal file sizes.
- Best compression — slower encode, smaller files. Only practical for short jobs — the browser's WebAssembly FFmpeg is already 2–5× slower than native.
For long batch jobs (multi-hour theatre recordings, whole surgical case archives), stay on Fast or Balanced. Best compression can turn a 5-minute encode into 30 minutes.
Audio
Decides what happens to the source audio track.
- Keep original (passthrough) — carries the original audio unchanged into the output. Fastest, no quality loss. Right choice for dictated case commentary or spoken narration you want to preserve. If the source codec isn't MP4-compatible the app automatically falls back to Re-encode.
- Re-encode (AAC 128k) — transcodes audio to AAC at a fixed bitrate. Ensures broad MP4 compatibility across all players and editing software.
- Strip audio — drops the audio track entirely. Smallest file. Useful for batch-processing silent surgical footage, or when operating-theatre background noise isn't wanted in the final output.
Keep processing when minimised
Modern browsers slow down or pause background tabs to save battery — normally sensible, but a problem for long batch jobs. Without this option, the moment you switch to another tab or minimise the window, the FFmpeg conversion would slow to roughly 1% of normal speed.
When checked, the app plays a silent audio loop and requests a screen wake-lock while a job is processing. Browsers exempt tabs playing audio from background throttling, so the CPU stays fully engaged whether you're looking at the tab or not. You may see a small speaker icon appear in the tab title bar while it's active — that's expected and the audio really is silent.
Leave it checked for anything longer than a few seconds' worth of conversion. Zero effect on the output — purely about keeping the encoder running when your attention is elsewhere.