Join my Discord for help, workflows and news of new LoRAs!

🎈 Aether Splash – Water Balloon Impact LoRA for Wan 2.2 5B (i2v)

Aether Splash is a custom LoRA trained for slow-motion splashes — a single water balloon bursting on the subject’s face with full liquid impact.
It was designed specifically for image-to-video (i2v) workflows using Wan 2.2 5B, and works best with clear portraits or close-up images of people.
Best results are on human subjects, but it may work on others too!

Thanks to @bblink787 and @masslevel for video contributions!

💦 Trigger Phrase
Use this exact prompt to trigger the effect:

water balloon bursting on face

🧠 How to Use (i2v)
Start with an image around 768px resolution (512–768px works well)
– Portrait or landscape format is fine
– Should feature a clearly visible subject, preferably a person
– Front-facing or 3/4 angle works best

Prompt using the trigger phrase above
– Keep prompt minimal
– The LoRA tells the model to animate a balloon hitting and bursting on the face

⚙️ Inference Settings
• Mode: image-to-video (i2v)
• Model: Wan 2.2 5B
• Resolution: around 768px (landscape or portrait)
• FPS: 24
• Steps: 20
• CFG: 5
• Clip length: 5 seconds (121 frames)

✅ Best Practices
• Use a clean, well-lit portrait or close-up image
• Avoid clutter — splash effect is strongest with clear subjects
• Don’t include water balloons in the image — the animation adds them

Trained using Ostris's AI Toolkit.

https://tensor.art/u/635897593804625782