Revised with Joop Schaye’s comments: COLIBRE’s sonified galaxies made fly by triggered star formation audible, while Little Red Dots sit below the simulation’s current black hole resolution.
The James Webb Space Telescope has found 341 compact, reddish objects in the early Universe — small, active galaxies that shouldn't exist yet. They appear just 600 million to 1.6 billion years after the Big Bang, and their black holes are already massive. The standard model of cosmology doesn't explain how those black holes got there so fast. A new simulation has now resolved everything else.
COLIBRE, a simulation published this month in the Monthly Notices of the Royal Astronomical Society by Schaye et al., is the first large-volume cosmological simulation to model cold gas and cosmic dust inside galaxies directly. Previous simulations could not track gas cooling below about 10,000 degrees inside galaxies — hotter than the surface of the Sun — because the computation was intractable. COLIBRE, built over nearly a decade by an international team spanning Leiden University, Durham University, and the University of Portsmouth, includes the physics required to model that cold interstellar gas. It tracks three grain species and two grain sizes of cosmic dust, uses 20 times more resolution elements than prior simulations, and required 72 million CPU hours on the COSMA8 supercomputer at Durham's DiRAC facility. The largest runs tracked 136 billion particles.
When COLIBRE's synthetic universes are compared against what JWST actually observed, they line up. Cold gas, dust, star formation, black hole feedback — across 13 billion years of cosmic history, the standard model now matches the telescope's data in detail. "The standard model of the Universe can explain galaxy formation more accurately than previously thought," the Royal Astronomical Society said in its press release.
In comments to type0, Joop Schaye of Leiden University, who led COLIBRE, said the sonified simulations were useful rather than cosmetic. He was “very happy to experience that the sound provides very useful information”: in videos of a galaxy’s evolution, fly-bys and other interactions make the triggered star formation immediately clear. Sonified videos, he added, “enhance the overall experience” in public talks because they help audiences understand what is happening.
The standard model of cosmology — Lambda-CDM, the term physicists use — could not explain where the first supermassive black holes came from. JWST found the objects, now called the Little Red Dots. COLIBRE does not predict them because it assumes their black hole seeds already exist. It models how those seeds grow and how they affect their host galaxies. It does not model how the seeds formed. This is not a simulation flaw so much as a boundary of current physics. Schaye emphasized that COLIBRE was developed before the Little Red Dots were discovered. If the objects are black holes with the relatively low masses now suggested in several papers, he told type0, they “fall below the numerical resolution of the simulation.” The mismatch, in his view, is a reason to build higher-resolution extensions rather than a reason to abandon COLIBRE’s Lambda-CDM result.
The leading hypothesis is that the Little Red Dots are actively growing supermassive black holes embedded in dense gas, their light reddened and partially obscured by that same gas. JWST spectra show broad hydrogen emission lines consistent with accretion onto a black hole; electron scattering through a Compton-thick medium can explain the line profiles without requiring the black holes to be as massive as they initially appeared. Several competing theories remain in play — direct-collapse black hole nurseries, supermassive star precursors, and a more recent proposal that dark matter halo collapse in a specific model of self-interacting dark matter can produce seed black holes of sufficient mass by the time the Universe was 600 million years old. None is confirmed.
The COLIBRE team is transparent about this. Joop Schaye of Leiden University led the project. James Trayford at the University of Portsmouth built the dust model and the simulation's sonification framework — STRAUSS, an open-source package that maps physical quantities like density, temperature, and star formation rate to audio properties like frequency, amplitude, and panning. The result is a way to hear a galaxy evolve: changes in star formation hit higher frequencies, black hole growth adds lower registers, outflows shift the stereo field. The team has released videos and audio tracks publicly.
COLIBRE's highest-resolution simulations are still running. The full picture will take months more to complete. The black hole seed problem will take longer still — and it is the field's most active open question. Lambda-CDM just passed a serious test. The Little Red Dots are the loose thread, and the people who built the best simulation in the world know exactly where it is.