T2VA integrated_multimodal_description: [Shot 1] A realistic, high-fashion film aesthetic with heavy 35mm grain and a cold neon grade—soft magenta and cyan against deep blacks—frames a vertical 9:16 close-up of a Japanese woman with fair skin, large dark-brown eyes, straight bangs, braided half-up dark-brown hair, teardrop pearl drop earrings, and a sleeveless white ruffled lace top. She stands against a dark wall lit by a single soft-magenta light strip and sings straight to camera with clear lip and jaw motion and a gentle head sway on the snare. She sings in a soft, breathy female voice with a melodic trap cadence, <d>[Japanese] 夜に溶けてく 私の声</d>. The Japanese text 「私の声」 punches into the mid-ground on the 808 hit at about 00:01.000 with a subtle scan-offset and chromatic split, placed beside her shoulder, never covering her eyes or mouth. The camera pushes in slightly with light handheld micro-shake synced to hi-hats. [Shot 2] At 00:05.000, a hard cut on the 808 bass hit to a second close space: a reflective wet-look floor with cyan light strips and the camera positioned slightly higher. The same woman, identical in appearance, lifts her gaze and sings the hook with more intensity, one hand rising softly on the accent, <d>[Japanese] 離さないで この光だけは</d>. The Japanese text 「この光だけは」 smashes into frame on the drop with low-frequency compression stretch, then sweeps out on the final beat. An upward tilt continues the same motion direction as Shot 1 for a kinetic match cut; handheld movement includes snare-synced micro frame skips. The shot ends on an end-freeze as the last 808 lands. overall_soundscape: Minimal dark room tone under the music; a faint airy reverb around the vocal. No spoken ambience or crowd sound. non_diegetic_music: A melodic trap instrumental at a hazy mid tempo: deep 808 sub-bass, rolling hi-hats, sparse backbeat snare, and a soft atmospheric synth pad. Steady groove with a small lift on the drop at 00:05.000, no big swell.
Parameters used to generate this content
The node graph used to generate this content — pan and zoom to explore, or download it to run in ComfyUI.
AI models used to generate this content