Generative AI (GAI), like other AI models, learns from training data to generate new data. Google’s Veo is a GAI that creates videos. A content creator may use three forms of input—text, frames, and ingredients—in combination to help Veo generate the clips for a movie.
Ingredients to Video
The “ingredients-to-video” mode in Veo allows creators to maintain consistent visual elements—like avatars and props—across scenes. This familiarity strengthens the bond of trust between creator and viewer allowing continued storytelling. Each generation can include up to three ingredients plus accompanying text.
In our example, we establish a single ingredient, Zoey, and reuse it for 3 different scenes.
Step 1 – Create a Reference Clip
Using a text-to-video text prompt like the one below, generate a movie (even though we’ll only be using one frame). Note that SceneBuilder requires landscape. Press the arrow button.
- Prompt: “A static close-up of an attractive slightly smiling gen-z er on a green screen under natural lighting. She has unusual hair colorings.”
- Video: Zoey on Green Screen
Step 2 – Add to Scene
After video generation completes, select “Add to scene” button to add the clip to SceneBuilder.
Step 3 – Switch to SceneBuilder
Step 4 – Create a Reference Frame
Move the playhead to a desired frame in the timeline. Save this frame as a reusable asset. The first frame is frequently best.
Step 5 – Set to Ingredients to Video
Step 6 – Get Saved Asset
Select the “+” button to add an ingredient.
Step 7 – Tell Your Story
Write a new prompt. Let’s interview Zoey.
Summarizing:
- Set mode to ‘Ingredients to Video’
- Set aspect to landscape
- Get ingredients
- Include new text prompt
- Press arrow button
Step 8 – Repeat as Necessary
Repeat steps 5, 6, and 7 for each scene where you require ingredient consistency.
Let’s add a second clip, with Zoey jogging.
Now, let’s add a third and final clip, with Zoey coming home.
- Prompt1: “The camera shoots over the shoulder shot of interviewer towards Zoey, interviewee. Interviewer says ‘So tell me about yourself.’ Remove green screen. Place Zoey in a business suit in interviewee’s chair. She says “My name is Zoey. I’m curious, quick on my feet, and always ready to learn.””
- Prompt2: “Remove green screen. The camera shoots Zoey (attached) and her running buddy jogging through a park. Her buddy asks “How did your interview go?” Zoey responds “Pretty good; I’ll let you know for sure tomorrow.. They continue running.”
- Prompt3: “Remove green screen. The camera shoots Zoey (attached) being greeted after arriving home by her dog, Max, a Golden Retriever, runs at Zoey and begins licking her face incontrollably. Zoey says excitedly “Hi, Buddy. I missed you so much. You wanna go for a walk?”
- Video: Introducing Zoey
Step 9 – Reorient Video (Optional, but important)
- Download the connected clips as a single landscape video.
- While landscape suits computers, portrait is best for phones; use a post-production editor like iMovie (for Mac) or CapCut (free) to crop or adjust the aspect ratio of the screen.
- Upload a final clip to your favorite social media platform.
Step 10 – Upload Video
Upload a final clip to your favorite social media platform.
Final Notes:
Watch for character consistency in Introducing Zoey‘s 3 scenes (interview, jogging, and home sweet home).
Related:
- How to use text-to-video.
- How to use frames-to-video.
- How to use ingredients-to-video (this page).
P.S. I pwn the typos and other mistakes. We all love AI, but we value raw human input (warts and all) more.