Two traditions, same purpose
When you open a Tamil panchangam, you are looking at either Vakya or Thirukanitham numbers. Both try to answer the same question: where are the Sun and Moon right now? They just use radically different methods to find the answer.
Vakya panchangam (वाक्य, 'utterance/verse') is the oldest tradition. It uses pre-computed tables, verses that encode astronomical positions for specific epochs. A panchangam-maker looks up the table, applies corrections for the current date, and writes the result. The method originates from the Surya Siddhanta and has been refined over centuries, but the fundamental constants are the same ones used 500–1000 years ago.
Thirukanitham (திருக்கணிதம், 'sacred computation'), also called Drigganita (दृग्गणित, 'computation from observation'), is the modern tradition. It computes Sun and Moon positions directly from astronomical algorithms calibrated against telescope observations. The result matches what an observatory would measure.
What is the actual difference?
In practice, they agree most of the time. For an average day, tithi, nakshatra, and sunrise will be identical or differ by under a minute. But at tithi boundaries, the moment when one tithi ends and the next begins, Vakya and Thirukanitham can diverge by 1–3 hours.
| Vakya | Thirukanitham | |
|---|---|---|
| Method | Verse tables + corrections | Modern algorithms |
| Source | Surya Siddhanta and manuscripts | NOAA, Meeus, JPL data |
| Updates | Fixed constants | Calibrated against observations |
| Accuracy | Within minutes most days | Within seconds |
| Tithi boundary | Can differ by 1–3 hours | Matches observed position |
| Used by | Traditional temples, old almanacs | Modern panchangam publishers, apps |
| Advantage | Reverence for tradition | Precision for all locations |
Worked example
When does it make a difference?
Imagine Ekadasi begins according to both systems:
- Vakya calculation
- Ekadasi begins 05:12
- Thirukanitham calculation
- Ekadasi begins 03:48
- Difference
- 1 hour 24 minutes
- But rarely...
- The boundary falls at the sunrise moment
Based on verse table with corrections.
Based on actual Moon position from algorithm.
Normally irrelevant, both agree that Ekadasi prevails at sunrise.
Then Vakya says 'today' and Thirukanitham says 'yesterday', or vice versa.
In rare cases (a few days per year) Vakya and Thirukanitham can give different festival dates. The difference is always at most one day.
Why we chose Thirukanitham
For the diaspora, Thirukanitham is the right choice. Here is why:
- Per-city precision: Sunrise and Rahu Kalam differ by several minutes between cities just one degree of longitude apart. A calculation anchored to real astronomy, for your city's exact coordinates, gives more reliable times than a manuscript table built for a single location.
- Transparency: We can show exactly which algorithm we use (NOAA for sun, Meeus for moon, Lahiri ayanamsa). Anyone with astronomical knowledge can verify our numbers.
- No black box: Our engine is pure TypeScript with no external libraries. Everything is implemented from scratch, no dependency on a third-party ephemeris we do not control.
- Modern consensus: Most modern panchangam publishers (Drik Panchang, ProKerala, mypanchang.com) use Thirukanitham. It is the direction the field is moving.
Our computation engine in detail
Our panchangam engine is written in pure TypeScript, no external ephemeris, no third-party astronomy library. It computes everything from scratch:
| Element | Method | Source |
|---|---|---|
| Sunrise and sunset | NOAA solar position algorithm | gml.noaa.gov/grad/solcalc |
| Sun position | NOAA apparent sun longitude | Same |
| Moon position | Meeus, Astronomical Algorithms, chapter 47 | 60-term sigma-L series |
| Tithi | Sun-Moon angle in 12-degree steps | Ayanamsa-independent |
| Nakshatra | Moon's sidereal position | Lahiri ayanamsa |
| Ayanamsa | Lahiri (Chitrapaksha), linear model | Anchored at J2000 ≈ 23.85° |
| Moonrise and moonset | Meeus section 15 | Geocentric altitude |
| Lagnam (ascendant) | Tropical ascendant − Lahiri ayanamsa | Same Lahiri model |
The engine is pure and deterministic: the same call gives the same answer on the server and in the browser. It uses no randomness, no network calls, no external service. Everything is computed locally from the date, time, and coordinates.
What is Lahiri ayanamsa?
Ayanamsa is the one convention choice in a panchangam engine. It is the angle between the tropical zodiac (used by Western astronomy) and the sidereal zodiac (used by Indian astronomy). The Earth 'wobbles' slowly over 26,000 years, and the ayanamsa compensates for this precession.
There are several competing ayanamsa models. We use Lahiri (Chitrapaksha), which is the official standard of the Indian government and is used by the vast majority of Tamil panchangam. It is anchored at J2000 (year 2000.0) with a value of approximately 23.85° and grows by approximately 50.26 arc-seconds per year.
What to do if you see a difference
If your family panchangam shows a different festival day than our calendar, there are three likely causes:
- Vakya vs. Thirukanitham: Your family panchangam may use Vakya tables. The difference is always at most one day.
- Different ayanamsa: If they use Raman or KP instead of Lahiri, nakshatra boundaries may differ.
- Different reference city: Some panchangam use your local sunrise to determine tithi. We use Chennai as the reference to keep the diaspora synchronised.
None of these differences is 'wrong'. They are two valid traditions using slightly different input data. Follow the tradition your family uses, or use our calendar if you prefer the transparent, modern computation.