Attap Kia

Sometimes, the mood says everything. In this guide, I’ll show you how I created a slow, atmospheric 16-second cinematic video from scratch using just a few tools: ChatGPT for ideation and reference visuals, and Google’s Veo 3 for final video generation.

Whether you’re a creator, storyteller, or someone who enjoys visual poetry, here’s a simple guide you can follow.

Step 1: Choose Your Emotion

Start with the emotion you want to convey. For this video, I chose melancholy mixed with quiet companionship — the feeling of two people sharing a final moment as the world grows darker around them.

You can pick themes like longing, hope, betrayal, serenity, or even confusion. The goal is to select a core emotional tone that will drive your visuals.

Step 2: Visualize the Scene with AI Assistance

I used ChatGPT to brainstorm scene ideas and then created either a reference image or a mockup image to visualize the mood and setting more clearly. This step helped solidify the tone and direction before moving on to video generation.

Try asking ChatGPT to help you create:

Scene descriptions

Mood boards

Character positioning

Visual metaphors


You can also generate an image using tools like DALL·E or Midjourney to preview your concept.

Step 3: Craft the Final Prompt

Here’s the exact prompt I used in Veo 3:

Wide cinematic 8-second video. Two people sit close together in foldable camping chairs at the edge of a high coastal cliff, backs facing the camera. A small white cooler box is placed between them. They slowly clink beer bottles in a quiet, intimate toast, then sit in still silence, facing the sea.

The entire scene is captured in a wide-angle shot, showing the full expanse of the cliff edge, the vast ocean, and a long row of offshore wind turbines stretching into the misty distance.

The environment is dark, stormy, and atmospheric — thick overcast clouds, a deep blue-grey color palette, and heavy shadows. One distant lightning bolt strikes on the horizon during the shot.

Shot with a full-frame digital cinema camera (Sony FX6 or ARRI ALEXA Mini style), using a 35mm wide-angle lens. Cinematic framing with shallow depth of field in the foreground, gentle handheld movement or slow push-in.

Audio: Natural ambient sound only — soft wind blowing, low thunder rumbling in the distance, faint ocean waves, and a subtle beer bottle clink. No dialogue or music.

Ultra-realistic, high resolution. Gloomy and emotionally heavy. No text, no subtitles.

Feel free to modify this prompt to suit your own concept or emotion.

Step 4: Generate the 8-Second Cinematic Clip with Veo 3

Once I had the emotion and scene locked in, I used Google Veo 3 to create the original 8-second video.

The prompt I crafted was designed to deliver a cinematic, moody shot with natural motion and realistic lighting. Veo 3 handled everything from composition to camera movement, making it feel like a scene from a film.

This short 8-second clip captured the core moment I had envisioned:
two people clinking bottles at the edge of a stormy cliff, then sitting in still silence facing the sea.

Step 5: Slow Down the Video to 16 Seconds Using CapCut or TikTok

To deepen the mood and give the scene more emotional weight, I imported the 8-second video into CapCut (or TikTok’s editor) and slowed it down to 50% speed, stretching it to 16 seconds.

This simple edit added a sense of calm and gravity, allowing each moment to breathe without needing extra footage or transitions.

The final breakdown looked like this:

0–3s: Wide still shot of the two characters sitting close together

3–6s: A slow clink of beer bottles between them

6–10s: A brief moment of silence and stillness

10–14s: Distant lightning flashes in the background

14–16s: Final still moment, emphasizing the mood


This slowdown technique is especially useful for creators who want to extend the atmosphere of a short AI-generated clip without complicating the edit.

Optional: Add a Minimalist, Emotionally Charged Caption

Once I slow-motioned the video from 8 seconds to 16 seconds, I wanted to give it a bit more emotional punch — without overpowering the visuals.

I brainstormed with ChatGPT to come up with a melodramatic, minimalist caption that could sit subtly on screen or be used as the video’s title.

Here’s what I chose:

“Even if the world crumbles. As long as you’re beside me, I’ll never look back.”


This line added just enough narrative depth to suggest a backstory, while keeping the focus on the visual emotion. You can overlay it in small, tasteful font at the bottom of the frame, or use it as your upload caption on social platforms.

Final Thought

This process let me turn a raw feeling into a visual narrative — no actors, no crew, just creativity and the right tools. The result was a slow, intimate scene that resonated deeply without saying a word.

Try it yourself, and share your version. Sometimes, the quietest moments tell the loudest stories.