HoloAttractor introduces a novel framework for synthesizing chaotic attractors through holomorphic flow dynamics combined with multi-format negative feedback mechanisms. The system explores the intersection of complex analysis, dynamical systems theory, and generative synthesis, demonstrating that holomorphic mappings in the complex plane can produce rich chaotic behavior when subjected to carefully designed negative feedback in multiple signal formats (amplitude, phase, and spectral domains). This work presents mathematical foundations, algorithmic implementations, and experimental results showing emergent attractor geometries with fractal dimensionality. The framework has applications in creative coding, scientific visualization, and the study of complex dynamical systems. Source code and implementation details are provided for reproducibility. Version 2. 0 adds MonaSeed: a DCT seed+residual image codec demonstrating that top-energy DCT coefficients capture the compositional essence of an image at extreme compression ratios. MonaSeed achieves up to 27x compression vs JPEG Q85 (1. 85 kB vs 49. 99 kB) at 22. 98 dB PSNR, with a tiled 8x8 block DCT residual codec adding only 2-4 kB to recover fine detail. New files: experiments/monaseedᵥ3. py, experiments/residualcodec. py. Updated paper/main. tex includes a full MonaSeed section with benchmark table and analysis.
Ege Berk Türk (2026) studied this question.