@sabrijefrii: Sebab sedap la aku repeat ni. Kopi Zus Dolce Latte

sabrijefrii
sabrijefrii
Open In TikTok:
Region: MY
Monday 02 November 2026 03:44:10 GMT
44
1
0
0

Music

Download

Comments

There are no more comments for this video.
To see more videos from user @sabrijefrii, please go to the Tikwm homepage.

Other Videos

How to Create Realistic AI Character Swaps like Stranger Things (Tutorial) 👇 I call this the
How to Create Realistic AI Character Swaps like Stranger Things (Tutorial) 👇 I call this the "Vecna Body Swap." Since Stranger Things is massive right now and I am a fan myself, I wanted to test if we could recreate those Hollywood-level practical effects using just AI. I wanted to see if the model could handle the extreme details-the muscles, the veins, the grotesque textures. The result? We are not just seeing better quality; we are seeing HIGH controllability. You can now become the actor, drive the performance, and execute high-end VFX shots literally from your phone with a decent internet connection. There's no way Hollywood won't be using this by the end of the year… so why shouldn’t you? The Tools: Nano Banana Pro + Kling 2.6 Motion Control. Here is the technical breakdown: 1. The Setup Matters I took the first frame of my video into Nano Banana Pro to create the Vecna image. The KEY here is that the generated character image must match your source video frame perfectly. The CLOSER they match in lighting and composition, the less the AI has to "guess," which KILLS the hallucinations and makes it look real. 2. Kling 2.6 Motion Control + "Match Video" Mode For the motion, I specifically used the "Match Video" setting in Motion Control (not Match Image). - Match Image is limited to 10 seconds. - Match Video allows for up to 30 seconds and handles more movement MUCH BETTER. 3. The Blend To finish it, I used Kling's Image-to-Video to extend the scene. This allows you to elegantly combine the character swap with environmental effects to sell the shot. Honest talk on prompting: You don't need complex or fancy prompting. I literally used speech-to-text and talked to my computer to describe the scene I imagined. Don't expect perfection on turn one-it usually takes me 2-3 tries to get the physics right. The real insight here is that the best creators aren't using one tool; they are chaining them. You use one tool for the texture, one for the motion, and another for the environment. That is how you get full control. Drop the word VECNA in the comments and I’ll DM you the prompts PS: I share AI workflows like this every day. Follow for more inspo 🔥

About