Filming tips for perfect lip-sync

While Dubly is designed to be highly versatile and can successfully handle a wide variety of "edge cases," it is important to remember that the quality and complexity of your input significantly impact the final output. Great Lip-Sync starts on set. Follow this checklist when filming content destined for Dubly.AI.

Framing

Face fills 10–40 % of the frame. Too small → less detail to animate; too close → any imperfection is magnified.
Front-facing or 3/4 profile. Strict side profiles are much harder to re-animate.
Keep head movement gentle — no quick turns mid-sentence.

Lighting

Even light on the face, no harsh shadows under the nose or around the mouth.
Avoid silhouettes — backlight without fill lighting makes the mouth hard to track.
Match color temperature across shots in the same scene so skin tones stay consistent.

Camera settings

Resolution: 1080p minimum, 4K ideal.
Frame rate: constant (CFR) 24, 25, or 30 fps. Avoid variable frame rate (VFR) recordings from some phones.
Shutter speed: double the frame rate (e.g., 1/50 for 25 fps) to avoid blur.
Focus: lock on the speaker's eyes so the mouth stays sharp.

Mouth visibility

No hands, props, or graphics covering the lips during speech.

Speaker direction

Coach your speaker to:

Speak clearly at a natural pace — don't over-enunciate.
Pause between sentences — gives Dubly.AI cleaner segment boundaries.
Limit side conversations and laughter during takes.
Avoid chewing gum or snacking on camera.

Audio

Lip-Sync still needs clean source audio to produce clean translations.

Post-production

Deliver un-graded or lightly graded footage if possible; heavy skin-smoothing can confuse face tracking.
Skip built-in subtitles and lower-thirds over the mouth; export a clean version.
Export as MP4 H.264 1080p, 48 kHz stereo. A five-minute setup following this checklist saves hours of cleanup later.