For most people, simulating gravitational dynamics requires a deep knowledge of physics and other mathematical training. It also requires precision instruments and high-spec hardware, all in order to visualize the Cosmos…but not anymore. GLEN presents a new, open opportunity, for normal, everyday people, to experience galaxies up close and personal, without needing expensive equipment or a working knowledge of Astrophysics. We have done all the work for you.
The images that we most commonly see for galaxies, while nice looking, are nothing more than artist illustrations and computer generated graphics exaggerated for dramatic effect. These images have helped us understand and learn so much about the universe, and we are immensely grateful for them. But while being very attractive, they do not take into account the object’s actual physical properties, and this is why we created GLEN. GLEN stands for Galactic Layer & Environment Navigator, and it’s a new lightweight, portable renderer that takes a galaxy’s real measured morphological parameters, such as Hubble T-type, inclination, bar strength, integrated color, redshift, effective radius, and renders what that galaxy looks like based on our current rendering capabilities. The “Sa spiral” gets tight wound arms and a bright central bulge because the data says it is an Sa spiral, not because an artist chose that look. The E7 elliptical is stretched to its observed axis ratio. The barred SBb has its bar elongated in proportion to the measured T-type. What you see in GLEN is a direct consequence of the measurements in the catalog.
While some may argue that GLEN itself is nothing more than a renderer with pretty shader capabilities, GLEN does something different from the tools that came before it. The Hubble and James Webb Space Telescope images show you what specific galaxies look like through an actual lens. These images however, are mostly bright points of light on a black starfield background (which still look very cool, we might add) but does not show you what the galaxy might look like up close. The Sloan Digital Sky Survey plates give you photometric data at scale. These are irreplaceable scientific resources. But they are not live renders. They are records. GLEN takes those records, the measured T-types, axis ratios, color indices, redshifts, and bar strengths, all of it, and synthesizes what the galaxy looks like as a live, interactive render in any browser, on any device, with no installation and no expiration. The image you see is not a photograph. It is a consequence of the data, all done portably for your ease of use and enjoyment.
While it would appear that this task was easy to do (since we happily report that it runs in a browser, no install required), the scale of the catalog itself presented the first significant challenge. No single public source contains all near-observable galaxies, much less to have a dataset carrying consistent morphological coverage. This is a large gap that we had to first fill, if we were ever to produce a large enough catalog worthy of browsing and becoming GLEN’s backend. The solution then, was to implement a five-source union: SDSS DR17 provides 1.1 million spectroscopic redshifts and velocity dispersions. HyperLEDA contributes 52,000 morphological T-types and 4 million basic entries. HECATE supplies 51,000 stellar masses and star formation rates. PGC2003 provides the baseline identifiers that tie the catalog together. And Mangrove adds environmental context, that is, whether a galaxy lives in isolation, in a group, or in a dense cluster. Each source fills some of the gaps the others leave open, although still greatly incomplete. The result of this union however, is a catalog where mostly every entry has something physically grounded to render from, not just a blank placeholder. The data was always there. No one ever thought to aggregate it in this way.
Now that we had a unified set of data from which to manipulate, the morphological coverage (the actual shape and type of galaxy) was still the hardest gap to close. HyperLEDA’s T-types cover only roughly 52,000 galaxies. That is less than 0.25% of the entire catalog. For the remaining ~22 Million galaxies in the catalog, we needed to derive a way to assign class and type to each galaxy, a most daunting, and some might argue, an nigh impossible task for a one or two man team. But after studying some more excellent mathematical concepts, and thinking about how it relates to the type and class of objects in the deep field, we derived a series of methods to do just that. As such, GLEN applies a color-morphology calibration: redder integrated color implies earlier type, and bluer implies later type. Although rudimentary, this was the first step to allowing us to classify over 20M+ galaxies. Assigning this rule, we honestly labeled all galaxies using this concept in the catalog as PHOTOMETRIC. No entry is silently fabricated. Thus if GLEN renders an E3 elliptical for a galaxy that was never morphologically classified prior, that label is in the data so that any researcher or developer consuming the API knows exactly what they are working with. The commitment to labeling derived values is not merely a footnote. It’s the policy that the entire catalog was built on.
Now the rendering itself, varies by galaxy class in ways that required separate treatment for each type. Spiral galaxies appear simple, but they are the most complex. Their arm structure is generated from the Hubble T-type and bar strength: an Sc has loosely wound, prominent arms with minimal bulge; an Sa is tightly wound with a dominant central concentration. The bar, where present, is elongated in proportion to the measured morphological bar parameter. Ellipticals are rendered with a Sérsic profile at the observed axis ratio, their light concentrated toward the center and falling off at the rate the data specifies. Irregular galaxies, by definition, have no imposed structure, and thus GLEN can render them as asymmetric or even clumpy distributions driven by their measured color and size. Lenticulars sit between this characterization: their disk present, but spiral arms absent. Each type presented is not a style choice on our part, but rather the output of a different physical rendering path driven by its measured parameters.
Redshift also presented its own consideration. A galaxy at z = 0.05 looks different from one at z = 0.3, not just in distance but in observed color. The spectrum shifts, the apparent size shrinks, and the surface brightness dims. GLEN accounts for this cosmological dimming and color shifting in the render pipeline so that what you see for a distant galaxy reflects its observed appearance rather than its rest-frame appearance projected forward. This also is not decoration. It is the difference between showing you a galaxy as a point of white light as a dot zoomed in, and showing you what astronomers actually measured.
Then, just as we did for STEN, in lieu of immersion and public engagement, GLEN adds very slight embellishment and color enhancement. The choice was also made to add an ambient background sound layer for the galaxy renders. No one can currently know what a galaxy sounds like, as no man-made instrument has ever left the Milky Way, but we can superimpose instead, the sound of depthless breath and immensity; a sound suited quite well to representing what a galaxy would sound like, if ever mankind should so garner the privilege of ascertaining it. Space however, is not a medium for acoustic waves. This we know. However, for a galaxy kiosk in a science museum, or a classroom embed, or perhaps a personal exploration session, where children look up and forward, and gaze upon the Stars and the Galaxies in wonder and awe, we felt that sounds are an integral part of this experience. It is also most certainly correct to say, that such venues benefit from depth and atmosphere. Thus as with STEN, we also selected more royalty-free, ambient tracks for the engine, and the site handles the attribution rights, so you may play GLEN from any browser of your choice, on any device or hardware. The tracks are hand-selected and deeply curated, crafted and tuned to the scale of what is being rendered. The sounds are the only part of the render that is not physically derived by galactic measurements, but we argue are necessary for the overall experience.
But even here again, GLEN is more than just a renderer. The underlying data that drives it is the same data that powers the entire GalaxyDB’s catalog API. Every shader parameter maps directly to a column in the database. That means any developer with an API key can query the same physical parameters that GLEN reads at render time; the T-type, the axis ratio, the integrated color, the redshift, all of it, and build their own tools on top of them. GLEN is in itself, both the frontend and the proof of concept for the data. GLEN can render over 22 Million galaxies correctly from those columns, and the possibilities of using this are endless. While we want to make our hard work and contributions a sustainable opportunity for us, we also hope that you will use the API to create even more wonderful things, and we are excited especially, with what you do with GLEN.
The Galaxies Await.