Original title: TARTARUS: A High-Performance Python Code for Ray Tracing in Curved Spacetimes
David F. Bambague, Alexis Larrañaga · September 2026
Telescopes recently captured images of the supermassive black holes M87* and Sagittarius A*. To understand these images, scientists must model how light bends around extreme objects. They need tools that trace paths of light particles, called photons, from glowing plasma to an observer. To solve this, researchers built TARTARUS. It is a new, modular framework written in Python to trace light paths in curved space. Python is easy to use, but it can run slowly. TARTARUS uses Just-In-Time compilation with Numba to match the fast execution speeds of languages like C++ or FORTRAN. The tool detects when light hits objects like accretion disks or event horizons. Tests on spinning Kerr black holes showed that the software is physically accurate and computationally efficient.
Planning a hike where the trail ahead changes shape depending on which way you step. The map you start with quickly becomes outdated, so you have to recalculate the path with each footfall. Here, light traveling near a black hole follows a similarly shifting path—the curvature of space alters the trajectory step by step—and the code recalculates the route at every point.
Observations of black holes like M87* and Sagittarius A* rely on modeling how light bends around them. This code makes such modeling faster and more accessible, allowing astrophysicists to test theories of accretion and gravity without needing to master low-level languages. It brings high-performance ray tracing into Python's flexible ecosystem.
The paper demonstrates TARTARUS only on Kerr black holes and thin accretion disks. It does not compare performance against existing ray-tracing codes or provide benchmarks. The accuracy tests are limited to Hamiltonian constraint conservation and visual similarity. Its performance on dynamical spacetimes or with photon scattering off plasma is not addressed.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
TARTARUS: A High-Performance Python Code for Ray Tracing in Curved Spacetimes — David F. Bambague, Alexis Larrañaga
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.