AI Lip Sync Singer

Use the provided avatar image as STRICT identity reference. Do not alter identity, face, proportions, or lighting.

PRIMARY GOAL: Demonstrate perfect, phoneme-accurate lip sync. Mouth precision has absolute priority over all motion.

AUDIO: Use the provided track "Can You Feel Me" (46s). Maintain exact timing. No delay, no drift, no reinterpretation.

CAMERA: Static tripod shot. No movement, no zoom, no shake. Tight bust-up framing, 3/4 angle. Microphone visible slightly in foreground.

PERFORMANCE (CORE): - Mouth movement must match phonemes exactly at all times. - Clear articulation for consonants (b, p, m = full lip closure; f, v = lip-to-teeth contact). - Distinct vowel shaping (A / E / I / O / U must be visually different and readable). - Natural jaw and lip coordination (no sliding or floating lips).

LONG TONE CONTROL (CRITICAL): - Sustain identical mouth shape during long vowels ("loooove", "gooooo", "stayyyyy"). - No jitter, no drift, no premature closing. - Only minimal natural jaw relaxation allowed while preserving shape consistency.

FAST PHRASE CONTROL (CRITICAL): - Preserve full articulation during fast phrases ("every little syllable"). - Do NOT compress or skip mouth transitions. - Each syllable must produce a visible mouth movement.

BREATH & TIMING: - Subtle inhale before first line and key phrases. - Micro pauses must be respected exactly. - Gentle exhale after long notes and ending phrase.

FACIAL EXPRESSION: - Emotionally immersive but controlled. - Slightly vulnerable gaze. - Eyes subtly moist during chorus. - Minimal natural blinking only (no random blinking).

HEAD / BODY: - Only micro head sway (very subtle, slow). - No large movement, no gestures, no shoulders exaggeration.

MIC INTERACTION: - Maintain consistent distance to microphone. - Slight forward lean during chorus for intensity. - No exaggerated motion.

LIGHTING / STYLE: - Preserve original cinematic studio lighting. - No stylization, no filters, no CG effects.

NEGATIVE (STRICT): - No lip sync lag - No vowel blending collapse - No mouth shape drift during sustained notes - No skipped articulation in fast phrases - No head bobbing - No exaggerated emotion - No timing changes - No camera movement

OUTPUT: Ultra-realistic human singing performance with stable, physically accurate lip sync, clearly readable articulation even in fast phrases and sustained vowels. Indistinguishable from real recorded footage. ---