There are many apps and products on the market that showcase the Stars…but we want to be clear about what TheStarDB is, and what STEN can do. Currently available apps like Stellarium and Sky Safari show you where stars are in the sky tonight. These, unlike what STEN renders, which are full Stellar Class Objects (SCO), are just points of light on perhaps a star field background. Functional for a time, but we felt it was time to do better. Universe Sandbox is also a great simulation application, and it lets you play around with colliding planets and watching gravity do its work, but Universe Sandbox lacks the portability and simplicity of STEN. These tools and applications, while specialized, serve a specific focus.
STEN however, does something different. It takes a star’s real measured physical parameters, such as Effective Temperature, Luminosity, Surface Gravity and Mass, and renders what that star looks like at the photospheric level. Dark, slightly cooler granulation cells are rendered (where appropriate), driven by rigorous physics and other logarithmic mathematical calculations, sustaining the Star’s persistent convective turbulence models. Limb darkening was also computed from wavelength-dependent opacity.
Black Holes, however, presented a unique rendering challenge. Standard rasterization cannot simulate what happens to light in an extreme gravitational field. In general relativity, light near a black hole does not travel in straight lines. It follows geodesics, which are curved paths through warped spacetime. For this, STEN employs a custom Geodesic Raymarcher. STEN’s Raymarcher numerically integrates these curved light paths step by step around the gravitational field, computing exactly how photons bend, orbit, and escape. This produces the characteristic photon ring, the relativistic brightening of the approaching side of the accretion disk due to Doppler boosting, and the black hole shadow itself, which is that dark central void hiding the singularity from which no light can theoretically return. Accretion disk emission profiles are further shaped by the compact object’s mass and spin parameter, which determine the innermost stable circular orbit and therefore the disk’s inner edge temperature and luminosity. Coronal structure and flare probability for the Stars were tied to spectral class, with a slight “top up” for dramatic effect and uniformity. Essentially, every visual element has at its baseline, a physical computable source. Nothing is merely decorative for the sake of being decorative.
Then, for mass consumption and public viewing experience and pleasure, we added a few sounds for immersion and perceived realism. We are well aware that a Star may not give off a burning sound like wood burning in atmospheric oxygen on a pyle in the same distortion as perceived by human ears. But perhaps it can. This we will not argue, but rather the decision to add such sounds were the only decorative parts of the entire program. However, even in this, we tried to be as accurate and correct as possible, using sound frequencies native to true approximations of the object’s sound as recorded by real-life physical instruments. The Pulsar’s “fast hitting, rhythm-like beating sound” is real. That is what a specific type of Pulsar sounds like. The Magnetar’s high-pitch radio frequency sound is a sound simulated by our engine to give the user experience greater depth. Then to complete the immersion, we provide hand-selected and curated royalty-free relaxing “emotion-depth” tracks throughout the experience, that puts the user in the mood, perfect for Stellar contemplation and relaxation.
But even then, STEN is more than just a renderer. STEN prides itself on clean specifications, and accurate, usable data. That accuracy is only possible because the underlying data itself is extremely clean and mathematically processed. Our underlying data forms the backend of TheStarDB and for STEN, and it also drives our clean API. Starting with open-source information available everywhere, STEN imports Gaia’s DR3 dataset; and using its photometry specifications, this gives us an effective temperature for all 16.1 million stars in the catalog, not just the ones with observed spectra. This is a great starting foundation. But that data is also incomplete. To achieve our level of accuracy, we then take that seed data in various combinations, (for instance, using temperature and luminosity) and we derive the stellar radius via the Stefan-Boltzmann relation. STEN then maps those parameters directly to custom hand-programmed, AI-verified shader inputs. What you see is a full consequence of the physics, not an artist’s approximation.