Graha-Swara Engine
Computes, minute by minute, the nostril (svara) and suṣumnā windows that the Śiva Svarodaya and Jyotiṣa lead one to expect from a birth chart and the live sky.
About Graha-Swara Engine
The engine behind Kaalchakra's Swara Code. Natal and moment terms — each tagged with its verse and labelled as the text's rule or our own construction — drive a breath phase integrated minute by minute through the day, locked so that every sunrise opens on its tithi's nāḍī (Śiva Svarodaya v.61), with the day's swara count fixed to its cosmic value. L1 is an audited fork of the earlier v3 engine, carrying nine fixes (F1–F9) from an audit across 904 days, 9 places and 8 charts. Pure, deterministic JavaScript that runs on the user's device; it reconstructs what the texts prescribe and makes no claim of predictive validity.
Highlights
- ▸Minute-by-minute iḍā / piṅgalā / suṣumnā from chart and sky, phase-locked to the tithi's nāḍī at sunrise
- ▸Birth-time robust — a 2-minute error changes under 2% of daytime minutes on average (test-asserted; up to 58% in v3)
- ▸Nine audit fixes (F1–F9), including a fixed cosmic swara count per day and merged suṣumnā slivers
- ▸Every term carries an id, a verse and a text/construction label; outputs are version-stamped
More from this lab
Tantra Vidyā Lab.
- software · beta · SEP 21, 2026Graha-Tattva Engine T1
Apportions the five tattvas — vāyu, agni, pṛthvī, jala, ākāśa — within each svara from ṣaḍbala, daśā lords and the Moon, inside the bounds the text gives.
Publications from this lab
- Research thesis · October 2026 · in preparation · OCT 31, 2026Kaalchakra: A Computational Reconstruction of Classical Indian Time-Keeping, Breath Science and Jyotiṣa
Kaalchakra is a browser-based, auditable reconstruction of classical Indian time-keeping, breath science (svarodaya) and Jyotiṣa. Planetary positions come from the Swiss Ephemeris (2.10.03, Moshier mode) compiled to WebAssembly, with an analytic Meeus–Standish fallback, and all computation runs on the user's device. Its central construction, the Swara Code, is produced by two engines. Graha-Swara (L1) combines natal and transit correspondences from the Śiva Svarodaya and the Bṛhat Parāśara Horā Śāstra into a minute-by-minute expectation of the flowing nostril and suṣumnā windows, phase-locked to the tithi at sunrise and robust to birth-time error. Graha-Tattva (T1) apportions the five tattvas within each svara from ṣaḍbala, daśā lords and the Moon's nakṣatra, inside bounds taken from the text. Every weight is declared as the authors' construction. Tables were rebuilt through a primary-source loop — fetch the scanned edition, diff, assert the text's own counts — and every table first written from secondary knowledge contained errors, including the elements of Jupiter and Mercury. Astronomical and calendrical outputs are calibrated against external references: NASA's eclipse canon (greatest eclipse within 10 s, 1801–2025), USNO sunrise times and Drik Panchang dates. The work claims fidelity to the texts, agreement with references and robustness — not predictive validity.
- Research dossier · in preparation · OCT 31, 2026Maantrik: A Graded, Source-Traceable Computational Reconstruction of Mantra, Breath and Letter Doctrines
Classical Indian sciences of sound and breath — the Śiva Svarodaya, Śikṣā phonetics, mātṛkā-nyāsa and the Tantrāloka's doctrine of letters rising in the breath — survive as verses and tables that are rarely made computable or checked against their editions. We present Maantrik, a local-first web instrument that encodes these doctrines as executable rules, each tagged with a provenance grade (A: verse located; B: attested; C: lineage-specific; D: modern synthesis) and traced to edition and verse in a graded research dossier. Its signal layer runs entirely on the device: an AudioWorklet syllable-onset detector for japa counting; a landmark-based phoneme estimator (YIN pitch, LPC formants) mapped to Śikṣā place and manner of articulation; in-ear capture of pulse, breath and candidate otoacoustic tones; and fingertip photoplethysmography. A Meeus/JPL ephemeris drives the pañcāṅga and eclipse computation, checked against published reference examples and ten eclipses of 2025–2027. A companion source study finds that the texts picture the body as a vīṇā, never a flute, and that none links the audible nostril tone to letters — so the breath module times letters and never reads them from sound. Body mappings are symbolic, not physiological; tests use synthetic signals, and no efficacy or predictive claim is made.
Choose the one that fits.
Get your app, venture or R&D done
An app, a venture or R&D — brief us on WhatsApp and get everything built, lightning fast.
Work with the commons For builders & researchersApply as a Fellow
Bring your app, your business — and apply for investment — or your own research. Not built yet? We'll build it with you first.
Apply For investorsInvest in the commons
Back Aitele itself, a lab's research programme, a venture, or a technology release on its way to market — one conversation covers all four.
Invest