Sources
Where every number on this site comes from, and the methods and tools that turn it into finds. All the space data is public.
How it fits togetherSpace dataCataloguesMethods & softwareHow it's builtSetup page dataAcknowledgementsLicence
How it fits together
NASA TESS→MAST (QLP light curves)→hunt.py→VSX / ASAS-SN / ZTF / Gaia check→Gaia, SIMBAD, ExoFOP→planet & deep checks→Star Log
A PC at home downloads every TESS light curve of a sector from NASA's archive, searches each one for dips and repeating changes, throws out everything already in the big variable-star catalogues, and adds what Gaia, SIMBAD and NASA's planet-candidate lists know about the star. Transit-like signals go through a planet check, strong variables through a deep check of all TESS sectors. The results are uploaded here every 10 minutes. Nothing on the site is typed in by hand except the finds marked By hand.
Space data
| Source | What is used | Used by |
|---|---|---|
| TESS (NASA) | The light curves: the brightness of every star TESS watched, every 10 or 30 minutes (200 s in newer sectors), for about 27 days per sector. Ricker et al. 2015, JATIS 1, 014003 | everything |
| QLP light curves at MAST | The Quick-Look Pipeline's light curves from TESS's full-frame images, downloaded per sector (FITS). Also all other sectors of a star, for the planet and deep checks. DOI 10.17909/t9-r086-e880. Huang et al. 2020, RNAAS 4, 204 & 206; Kunimoto et al. 2021, RNAAS 5, 234 | hunt.py, vet_transits.py, deep_check.py, export_star.py |
| TESS Input Catalog (TIC v8.2) | The star's size, mass, temperature, density, magnitudes, distance and contamination, for the planet check. Stassun et al. 2019, AJ 158, 138 | vet_transits.py |
| Gaia DR3 (ESA) | Position, parallax / distance, colour, brightness, temperature, size and mass (GSP-Phot, FLAME), RUWE (hidden-companion hint), Gaia's own variable classification and eclipsing-binary period, and Gaia's epoch photometry (its own light curve). Queried through the ARI Heidelberg Gaia mirror and the ESA Gaia Archive. Gaia Collaboration, Prusti et al. 2016, A&A 595, A1; Gaia Collaboration, Vallenari et al. 2023, A&A 674, A1; Eyer et al. 2023, A&A 674, A13; Mowlavi et al. 2023, A&A 674, A16; Creevey et al. 2023, A&A 674, A26 | gaia_enrich.py, sync_to_site.py, illustrate.py |
| ExoFOP-TESS | NASA's lists of TESS Objects of Interest (TOIs) and Community TOIs, with their status (PC, CP, FP…). Downloaded every 3 hours; every candidate on the site is checked against them. Guerrero et al. 2021, ApJS 254, 39; ExoFOP DOI 10.26134/ExoFOP5 | sync_to_site.py, vet_transits.py |
Catalogues (is it already known?)
Every flagged star is matched within 30″ against these, through VizieR and the CDS X-Match service. Only stars that are in none of the first three become New candidates.
| Catalogue | What it is | VizieR table |
|---|---|---|
| VSX (AAVSO) | The International Variable Star Index: the biggest list of known variable stars. Also used for the Nearest known variable within 5′. Watson, Henden & Price 2006, SASS 25, 47 | B/vsx |
| ASAS-SN | Variable stars found by the All-Sky Automated Survey for Supernovae. Jayasinghe et al. 2018, MNRAS 477, 3145 (and later papers in the series) | II/366 |
| ZTF | Periodic variable stars from the Zwicky Transient Facility. Chen et al. 2020, ApJS 249, 18 | J/ApJS/249/18 |
| Gaia DR3 variability | Gaia's own classification of variable stars (shown as a note, doesn't remove a find). Rimoldini et al. 2023, A&A 674, A14 | I/358 |
| SIMBAD (CDS) | The star's best-known name and type, and whether research papers already call it variable or a binary. Wenger et al. 2000, A&AS 143, 9 | — |
Methods & software
| Tool / method | What it does here |
|---|---|
| Box Least Squares | Finds repeating box-shaped dips: eclipses and transits. Kovács, Zucker & Mazeh 2002, A&A 391, 369 |
| Lomb–Scargle | Finds smooth repeating changes (pulsation, spots, ellipsoidal stars); with prewhitening in the deep check. Lomb 1976, Ap&SS 39, 447; Scargle 1982, ApJ 263, 835; VanderPlas 2018, ApJS 236, 16 |
| LEO-vetter | The automated pass/fail tests of the planet check (odd/even, secondary eclipse, shape, noise, …). Kunimoto et al. 2025, arXiv:2509.10619 |
| TRICERATOPS | The planet probability: compares a planet against every kind of binary star and nearby-star scenario. Giacalone et al. 2021, AJ 161, 24 |
| Astropy | The BLS and Lomb–Scargle code, FITS files, coordinates and units. Astropy Collaboration 2013, 2018, 2022 (ApJ 935, 167) |
| Lightkurve | Finds and downloads all TESS sectors of a star. Lightkurve Collaboration 2018, ASCL 1812.013 |
| astroquery & PyVO | Talk to MAST, VizieR, X-Match, SIMBAD and the Gaia archives. Ginsburg et al. 2019, AJ 157, 98 |
| NumPy, SciPy, Matplotlib | The maths, fitting and every plot on the site. Harris et al. 2020, Nature 585, 357; Virtanen et al. 2020, Nat. Methods 17, 261; Hunter 2007, CiSE 9, 90 |
| Blender (Cycles) | Renders the eclipsing binaries to scale from the fitted sizes, temperatures and orbit. |
| ComfyUI + Stable Diffusion XL | Paints the Blender render into an artist's impression, locally on the PC's GPU (DreamShaper XL model). Impressions are art, not data. Podell et al. 2023, arXiv:2307.01952 |
| Flask, gunicorn, nginx | This website. HTTPS certificates from Let's Encrypt. |
| ntfy | A push notification to the owner's phone when a 95%+ planet candidate turns up. |
How it's built
Star Log is run by these programs, written for this project. The code itself is not published (see Licence); every source of data and method they rely on is listed above.
| Part | What it does |
|---|---|
hunt.py | The search: downloads every QLP light curve of a TESS sector, runs BLS and Lomb-Scargle on each star, flags variables, and removes the ones already in VSX, ASAS-SN and ZTF. |
sync_to_site.py | Builds each find's page (plots, Gaia, SIMBAD, neighbours, NASA's TOI lists, planet and deep checks, statuses, notes) and uploads it here every 10 minutes; also the Setup page numbers. |
vet_transits.py | The planet check: LEO-vetter tests, the TRICERATOPS planet probability, the verification dossier for 95%+ candidates, and alerts. |
deep_check.py | Deep check of strong variable stars over all TESS sectors: true period, type (binary, spots, pulsation) and whether it is probably known. |
gaia_enrich.py | Looks every candidate up in Gaia DR3 and makes Gaia's own light curve where there is one. |
notify.py | Sends a push notification (ntfy) when a 95%+ planet candidate is found. |
export_star.py | Packages a star analysed by hand in a notebook into a find for this site. |
rescan_dips.py | Re-checks single-dip candidates with the latest filters. |
diagnose.py | Quick health check of the search and its files. |
auto_illustrate.py | Picks the best binaries and transit candidates and makes artist's impressions of them. |
illustrate.py | Fits an eclipse model to the light curve, adds Gaia's temperature and Kepler's third law, and renders the system. |
render_binary_blender.py | Runs inside Blender: builds both stars to scale with their real colours and renders them. |
ai_paint.py | Paints the Blender render into an artist's impression with Stable Diffusion XL, locally on the GPU. |
start_all.sh | Starts the background jobs (search, sync, impressions, planet check) in tmux. |
update_all.sh | Installs new script versions and restarts everything. |
app.py | This website (Flask, on the server). |
Python 3.12 under WSL2 Ubuntu on the home PC, kept running around the clock; the website runs on a separate server in Germany.
Setup page data
| Number | Where it comes from |
|---|---|
| Temperatures, load, power | LibreHardwareMonitor on the home PC, plus nvidia-smi for the RTX 4070. Read every minute. |
| Stock vs tuned watts | Measured full-load power against the limits set in the GPU drivers (AMD Adrenalin, NVIDIA). |
| Solar panels | Live sunshine on a flat panel from Open-Meteo (weather data by Open-Meteo.com, CC BY 4.0), for 4 × 400 W panels. The exact location stays on the PC. |
| Cost & CO₂ | The owner's electricity price (€0.26/kWh) and about 0.27 kg CO₂ per kWh, a rough Dutch grid average. |
| Wren | Typed over from the owner's Wren dashboard. |
| Uptime, disk, internet | Windows, WSL and the server itself; the fibre speed is the subscription's, not a live test. |
Acknowledgements
Using something from Star Log in your own work? Please credit the original data, as these projects ask:
This work includes data collected by the TESS mission, which are publicly available from the Mikulski Archive for Space Telescopes (MAST). Funding for the TESS mission is provided by NASA's Science Mission Directorate.
This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement.
This research has made use of the Exoplanet Follow-up Observation Program (ExoFOP; DOI: 10.26134/ExoFOP5) website, which is operated by the California Institute of Technology, under contract with the National Aeronautics and Space Administration under the Exoplanet Exploration Program.
This research has made use of the SIMBAD database and the VizieR catalogue access tool, CDS, Strasbourg, France, and of the International Variable Star Index (VSX) database, operated at AAVSO, Cambridge, Massachusetts, USA.
Licence
- The code
- © 2026 Toasty. All rights reserved. The programs behind Star Log are private: not published, and not to be copied or reused without permission.
- Site content
- The plots, tables, notes and verification dossiers on this site may be shared and reused under Creative Commons Attribution 4.0 (CC BY 4.0): credit Star Log (starlog.nl) and the original data sources above.
- The space data
- Belongs to the missions and catalogues it comes from (NASA, ESA, MAST, CDS, AAVSO, …) and is used under their own terms; please credit them as shown in Acknowledgements.
- Artist's impressions
- Generated pictures, not observations. Free to share under the same CC BY 4.0 terms.
- No warranty
- Everything here is automated, unreviewed research output, provided as-is. A candidate is a lead, not a discovery.
Questions or permission requests: Discord meliketoast_.