Playing Windows XP sounds on instruments is one thing, using custom MIDI dials to manipulate an AI image generator image in real-time is another. Johannes Stelzer accomplishes the latter by with Stable Diffusion‘s SDXL Turbo, which creates a baseline image of “photo of a red brick house, blue sky”.
New way to navigate latent space. It preservers the underlying image structure and feels a bit like a powerful style-transfer that can be applied to anything. The trick is to… pic.twitter.com/orFBysBpkT
— Johannes Stelzer (@j_stelzer) July 15, 2024
Put simply, the MIDI dials basically function as manual controls for applying the various embeddings to this baseline, including “coral”, “moss”, “fire”, “ice”, “sand”, “rusty steel” and “cookie”. Adjusting the dials applies the embeddings to the base image in varying strengths. All of the effects you see a generated in real-time, thanks to software toolkit lunar_tools being integrated directly into the MIDI controller.
Bose SoundLink Flex Bluetooth Speaker, Portable Speaker with Microphone, Wireless Waterproof Speaker for...
- STATE-OF-ART DESIGN: SoundLink Flex outdoor speaker is packed with exclusive technologies and a custom-engineered transducer for deep, clear, and...
- CLEAREST POSSIBLE SOUND: Proprietary Position Qtechnology automatically detects the position of your portable Bluetooth speaker for optimal sound...
- WATERPROOF WIRELESS SPEAKER FOR TRAVEL: SoundLink Flex is rigorously tested to meet IP67 waterproof speaker standards. Crafted and sealed with...
[Source]
