LTX2.3_YOSETA_StartEnd_v02_FLOYO

Overview

This workflow generates a short video by creating controlled motion between a user-supplied start frame and end frame using LTX-Video 2.3.

It was adapted for the YOSETA commercial project to maintain character, composition, object placement, and shot continuity across a multi-shot production. Using approved images at both ends made it easier to constrain unwanted changes and reduce unnecessary retakes.

How to use

1. Upload a start image and an end image with matching dimensions and aspect ratios.

2. Describe the intended movement, performance, and camera behavior in the prompt.

3. Set the desired frame count and duration.

4. Run the workflow and review the generated motion between the two reference frames.

For subtle facial expressions or restrained movement, the same approved image can be used as both the start and end frame.

Production notes

• Matching image dimensions are important for stable execution.

• Prompts should focus on motion rather than repeating every visual detail already present in the images.

• Start and end frames help constrain composition, but generated intermediate frames should still be reviewed.

• Text, logos, interface graphics, and subtitles are best added during post-production.

• The workflow was also prepared for repeatable Floyo API operation during the YOSETA production.

Attribution

This workflow was adapted from an existing LTX-Video 2.3 Start/End Frame workflow and customized for Floyo-based production, continuity control, and API-assisted operation in the YOSETA project.