Turn static images into cinematic video sequences with this powerful LTXV Image-to-Video MP4 workflow in ComfyUI. By combining rich text prompts, visual conditioning, and the LTXV model, this setup animates a single frame into a flowing MP4 video, ideal for storytelling, mood pieces, or social media content.

---

🎯 Features

• LTXV Video Model – Uses ltx-video-2b-v0.9.safetensors for high-quality latent video generation.
• Image Animation – Animates a loaded image into a moving sequence with dynamic camera motion.
• MP4 Output – Saves final video using VHS_VideoCombine as H.264 MP4 with adjustable settings.
• Prompt-Driven Movement – Enhances realism and emotion using descriptive CLIP-based prompts.
• Scheduler & Sampler Control – Fine-tune motion consistency and temporal diffusion.
• Preview & Playback Support – Instant preview from VHS_VideoCombine makes iteration fast.

---

💡 Use Cases

• Cinematic Storyboarding – Animate still shots for visual storytelling or animatics.
• Social Media Clips – Turn art or photos into dynamic videos for platforms like TikTok or Reels.
• AI Art Motion – Add motion to static AI artworks for NFTs, presentations, or portfolios.
• Virtual Tourism & Scenes – Bring photos of landscapes or interiors to life.
• Creative Experiments – Explore how visuals transform under prompt-driven animation.

---

⚙️ How It Works

1. Load an Image – Use LoadImage to input your still frame (e.g., landscape or portrait).
2. Set Up LTXV – Load ltx-video-2b-v0.9.safetensors with CheckpointLoaderSimple.
3. Write Prompts – Use CLIPTextEncode for both positive and negative prompt descriptions.
4. Image-to-Video Node – Use LTXVImgToVideo to convert your image into latent video space.
5. Apply Conditioning – Send prompt encodings through LTXVConditioning.
6. Generate Animation – Use LTXVScheduler and SamplerCustom to render temporal sequences.
7. Decode to Frames – Convert latent outputs into video frames using VAEDecode.
8. Export MP4 – Combine frames into a final .mp4 using VHS_VideoCombine with preview enabled.

---

Credits: pixaroma