Transform a room while keeping the same character

Step by step

Run the workflow

1. 01

   The full workflow behind how I made this exact video

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. Follow the shown setup, keep the source order unchanged, and run one controlled test.

   

   

   

   

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
2. 02

   Prepare the source material

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. Capture as many images as you need, no pro camera gear required.

   

   

   

   

   

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
3. 03

   Open DirectorGPT

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. Your Personal Director
   2. We eliminate the guesswork from prompt engineering. This CustomGPT is calibrated to generate cinematic, ad-ready prompts optimized specifically for NanoBanana Pro.
   3. The CustomGPT functions as your Creative Director. It builds structured concepts and converts them into production-ready prompts.

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
4. 04

   Upload the source to DirectorGPT

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   • Click here to open ChatGPT (opens in a new tab)Direct link supplied in the source. Source page 6.

   1. Let's get started.
   2. Link Open the CustomGPT (DirectorGPT):
   3. Open the DirectorGPT using the link above.
   4. Upload your image into the DirectorGPT.

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
5. 05

   The DirectorGPT generates the prompt for you

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. Review the concept. If needed, ask for more variations in the same style.

   

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
6. 06

   Upload your image and send it

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. Ready to copy and paste prompt:
   2. Generate the prompt, don't change the person face, only the house. Lighting can change and match the human, don't change her facial features. Also make sure the lighting of background and subject matches

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
7. 07

   Open the image generator

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   • Click here to open Google Flow (opens in a new tab)Direct link supplied in the source. Source page 9.

   1. With your assets ready and your DirectorGPT prompts in hand, it's time to turn your phone photos into commercial-looking ad frames.
   2. Go to Google Flow and select "Create with Flow."
   3. Select + New project

   

   

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
8. 08

   Upload the source image

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. You may receive free credits (varies by account).
   2. Upload your source image.

   

   

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
9. 09

   Copy the selected prompt from DirectorGPT and paste it into Google Flow

   Use the source inputs to complete this stage and produce a result that can be checked before continuing.

   1. Follow the shown setup, keep the source order unchanged, and run one controlled test.

   

   

   Common failure modes

   • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
   • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
10. 10

    Select the "Omni" icon and select "VIDEO 3.0 Omni"

    Use the source inputs to complete this stage and produce a result that can be checked before continuing.

    • Click here to open Kling AI (opens in a new tab)Direct link supplied in the source. The Workflow Lab may earn a commission. Source page 12.

    1. The Generation Engine
    2. Now bring your ad frame to life in Kling AI. Precision here helps the model match your intended commercial look.
    3. Click here to access KlingAI
    4. Upload your source video as well as the generated image into the context window, and press generate.
    5. @video change background to @image

    

    

    

    Common failure modes

    • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
    • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
11. 11

    Compare the generated versions

    Use the source inputs to complete this stage and produce a result that can be checked before continuing.

    • Click here to open Kling AI (opens in a new tab)Direct link supplied in the source. The Workflow Lab may earn a commission. Source page 12.

    1. This workflow performs best in Professional Mode. Standard or free modes may produce lower resolution or unstable motion.
    2. Generation behavior
    3. Expect to run multiple generations before getting the result you want.
    4. Get the Tools Here

    

    Common failure modes

    • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
    • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
12. 12

    Set the edit rhythm

    Use the source inputs to complete this stage and produce a result that can be checked before continuing.

    1. Raw AI footage is your base. Post-production is where it becomes cinematic.
    2. Temporal Remapping (Speed Ramps):
    3. Use speed curves to create rhythm and emphasize key moments.

    

    

    

    Common failure modes

    • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
    • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
13. 13

    Grade the colour

    Use the source inputs to complete this stage and produce a result that can be checked before continuing.

    1. Visual Enhancement: Glow
    2. Apply a "Soft Glow" or "Edge Glow" effect to simulate high-end lens diffusion and luxury aesthetics.
    3. Color Grading: The Final Look
    4. Final color adjustments allow you to refine the mood beyond the AI's base

    

    

    

    

    

    Common failure modes

    • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
    • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.
14. 14

    Polish and export the final edit

    Use the source inputs to complete this stage and produce a result that can be checked before continuing.

    • Click here to open CapCut (opens in a new tab)Direct link supplied in the source. The Workflow Lab may earn a commission. Source page 17.

    1. Polish is what makes it feel expensive. Fine-tuning is what maximizes impact and gives your commercial a premium finish.

    

    Common failure modes

    • The model changes the subject, framing, text, product, or background instead of editing only the requested element.
    • The result looks polished at first glance but contains broken hands, unreadable text, warped objects, or inconsistent lighting.

Before you use it

Final checks

• Check the full result for changed people, products, text, timing, or layout.
• Confirm the tool, model, controls, and cost before generating.
• Keep private or unreleased material out of third-party tools.