Generative Video Wall Art Reveals Paintings Brushstroke By Brushstroke
May 20, 2022 by Dave Haynes
Industry veteran Jim Nista sold his content and coding company Insteo to Almo a few years ago, and after spending some time working within the distribution company, he went back out on his own and is focused again on developing creative for digital signage projects.
This is a really nice take on generative art for large format displays – using AI and real-time data to create dynamic paintings that build brushstroke by brushstroke on a display.
“I’ve been learning traditional oil painting, but I kept dreaming about painting with code,” Nista writes on Linkedin. “This painting was generated entirely with code on a BrightSign player, creating itself slowly, stroke by stroke while you watch. Using AI & Machine Learning, I coded an entire process for generating unique works of art every few minutes. When it completes a painting, it waits a few minutes before erasing itself and starting something totally new.”
“Perfectly suited as ambient content for public spaces, the canvas creates infinite paintings. It evolves and changes very slowly. Color palettes and paint strokes are decided using the seasons, time of day, position of the sun and moon. The first version generates sunflowers as tribute to the people of Ukraine.”
I’m a big fan of generative art for things like video walls, in part because they are visually interesting but also because they have set-it-and-forget-it qualities. These are living pieces that are informed by external, real-time data, so they’re never quite the same and always shape-shifting. That keeps them fresh in a way not possible with rendered videos that are the same until the file is swapped for something new.
The challenge with a lot of the oozing, morphing and pulsing generative work out there is that it looks kind of amazing, but you don’t really know what you are looking at and why it is transforming. What Nista is doing, on the other hand, builds into something recognizable, like a landscape or flowers.
This page has a video showing the build, though it was buffering even with my fiber connection (BIG file, I assume): http://nista.co/generative/