ככה תיצרו סרטון שלכם לעסק
בעזרת בינה מלאכותית

A high-resolution cinematic portrait of the person from [my photo], keeping their exact face, features and pose, but wearing a solid black hoodie. Replace the flat lighting with a professional three-point setup in a dark studio. On the left, a warm gold edge light acts as a rim light outlining the shoulder and hair, with a glowing gold light tube in the background. On the right, a deep royal purple light washes the wall and a potted plant. A soft front key light keeps the face sharp and detailed, while the gold and purple add subtle highlights on the skin. Shallow depth of field, the subject tack sharp against the atmospheric colored background. Photorealistic, ultra detailed.

Keep everything exactly the same, only change the two background lights to [color 1] and [color 2].

Animate this video as a visual restyle only. The audio, voice, words, rhythm and timing are locked and must stay identical to the source. This is the top priority.

Keep my original voice exactly as it is. Do not clone, re-synthesize, re-dub or replace my voice, tone or pitch. Copy the original spoken words exactly, in the same order, each word spoken once. Do not guess, invent, add, remove or shorten any word. No repeated words, no doubled words, no stutter, no echo.

Keep the exact original duration, frame count and speed 1:1. Do not shorten, trim, extend, stretch or loop the video. Never drop the first words. Do not add any intro or silence before the first word.

Preserve my mouth and lips frame by frame, perfectly synced to the original audio. Preserve my facial expressions, eye contact and head movements. Do not change my face, hair or identity.

Only change the visuals: dress me in a solid black hoodie, and place me in a dark professional studio. Add cinematic split lighting, a warm light on one side and a cool light on the other, a soft key light on the face with rich shadows, a glowing vertical light tube on one side and a softly blurred potted plant on the other. Match the exact colors and lighting of the reference image so every clip looks the same.

No subtitles, no captions, no on-screen text, no watermarks, no logos.

Negative prompt: dropped first words, added intro, invented words, added words, removed words, changed words, wrong words, re-dubbed audio, re-synthesized voice, cloned voice, different voice, altered tone, repeated words, doubled words, stutter, echo, shortened video, extended video, time-stretched motion, looping frames, out-of-sync lips, altered motion, warping, flickering, changing identity, distorted hands, color shift, inconsistent lighting, subtitles, captions, on-screen text, watermark, logo, bright background, flat lighting, daylight.