Finally! I’ve completed the first stable version of Entrain. I’ve begun rendering designs across a high-performance GPU cluster—utilizing everything from my primary RTX 5090 to A6000s, A100s, and H100s.
I am producing stunning visual art at 1080p, 4k, and even 8k. These files are massive; rendering at 4k60 or 8k120 is computationally expensive, but the results are captivating. The underlying math is incredibly complex, often taking days to render a single sequence. By using unique random seeds, I ensure every piece is mathematically distinct.
The stack is built on PyTorch, Python, CuPy, and FFmpeg, supported by Node.js, Redis, and PipeWire. While the code is evolving too fast for a public GitHub repo just yet, I’m focused on creating “mathematic visual art” paired with lush tonal frequencies designed for brain entrainment, sleep, and meditation. I can’t wait to share these with you! Meanwhile, I am working on new tools for v7.0 and will likely set up some social media in the coming weeks. — Reev