Nava Panchangam ← Back to app
Methodology

How Nava Calculates

நவ பஞ்சாங்கம் எவ்வாறு கணிக்கிறது — கணிப்பு முறைகளும் மரபு மரியாதைகளும்.

Two calculation systems

Nava offers two independent engines and shows which one is active. The two traditions genuinely disagree on some days — that is the point of offering both, not a defect in either.

SystemWhat it is
Vakya (default) A pure arithmetic engine in the Tamil Vakyakarana lineage. Sun and Moon longitudes are computed directly from the tradition's own printed tables — the Bhupadi table for the Sun, Vararuci's 248 candravākyas for the Moon (1962 critical edition) — with the classical manyadi lunar correction and the publisher's quarter-day Sun bija rule. This is the tradition the printed Tamil almanac in your home follows. Verified against printed almanac data from two publishers — sources that print ghati–vighati only — it faithfully reproduces those printed ghati–vighati values, which is why it is the default. Sunrise and sunset follow the almanac's own definition — the moment the Sun's centre crosses the true horizon, with no correction for atmospheric refraction — which matches the printed sunrise to within seconds. Measured against wall calendars that print actual clock times (Chennai 395 days, Coimbatore 377 and Delhi 363 — three complete years — plus Mumbai and 158 days of a US edition for New Jersey), the tithi or nakshatra named for the day matched the printed page on 2,664 of 2,666 comparisons — the two that differ are days where a limb ends within a minute of sunrise, so the page and the calculation name the limb on either side of the same boundary. The Tamil calendar date matched on every one of the 1,344 printed days of the 2026-27 editions we have been able to check. Source of the tables: T. S. Kuppanna Sastri & K. V. Sarma, Vākyakaraṇa (1962). You can switch between engines in your profile.
Thirukanitha (Drik Ganitha) True (drik) planetary positions from the Swiss Ephemeris with the Lahiri (Chitrapaksha) ayanamsa — the method Tamil tradition calls Thirukanitham, and the convention most computed panchangams use. Sunrise and sunset are the apparent (refraction-corrected) times for your exact coordinates, and the Tamil month begins by the sankranti-before-sunset rule. Checked against published drik panchangam data for Chennai (2026) and Lucknow (2025) — 725 days, every tithi, nakshatra, moon-rasi and paksha matching, and sunrise matching within seconds.

Two calendars, side by side

A Hindu date has two different months, and Nava computes both. The one it leads with is the Souramana (solar) Tamil month — Chithirai, Vaikasi, Aadi and so on — which changes when the Sun enters a new zodiac sign. Nava applies the Tamil Nadu convention specifically: the month begins on the day of the sankranti if that moment falls before sunset, and otherwise the next day. Kerala and coastal Karnataka are also solar, but they resolve that boundary differently (Kerala uses an aparahna rule), so a small number of days each year will not match an almanac printed for those regions.

Nava also computes the Chandramana (lunar) month — Chaitra, Vaisakha, Ashadha — which runs from one new moon to the next (amanta reckoning, as followed across South and Western India), from the same engine you have selected. It is deliberately not displayed next to the Tamil month. A Tamil almanac does not print one: its columns are the solar month and day, the tithi and the nakshatra, and the lunar cycle is carried by the tithi and paksha you already see. Showing two month names side by side would leave you with no way to tell which one is yours.

One lunar fact the solar month cannot express does get shown. Roughly every 32½ months a lunar month contains no solar sankranti at all; that month is an Adhika (intercalary) month, and it is why some festivals fall a month later in those years. On those days — and only those — Nava shows a small Adhika Masa label with the lunar month it applies to. The most recent began on 17 May 2026 (Adhika Jyeshtha). The same label appears for the far rarer Kshaya month, where one lunar month absorbs the next one's name (three times between 1900 and 2200; the one in range was January 1983).

The lunar festivals are now detected from the lunar month, not from the solar month as an approximation of it. Ugadi, Sri Rama Navami, Ratha Saptami, the Navaratri days, Deepavali and Naraka Chaturdashi, Gayatri Japam and Sama Veda Upakarma each name the lunar month they actually belong to, and an Adhika month is excluded — the correction that stops a festival landing a whole month early in the years where the lunar month it belongs to is doubled. Every one of those months was read off the printed almanacs rather than assumed, and each is checked against 952 transcribed almanac days on every change. That check covers the festivals a transcription actually names — about twenty of them — not every entry in the catalogue; the rest rest on the rule as stated rather than on a printed date we can point to.

A festival is also placed on the day the tithi prevails, not merely the day it happens to touch sunrise. Amavasya observances and their family are judged at aparahna (about 70% of the way through daylight); Pradosham is judged across the 90-minute pradosha kala that begins at sunset, because Trayodashi can begin after one sunset and end before the next and so prevail at neither. Where a tithi covers two such windows on consecutive days, the later day is taken. These are the almanac's own conventions.

One caution that remains, stated plainly: festivals keyed on a nakshatra rather than a tithi — the Veda Upakarmas, Karthigai and Bharani Deepam, Masi Magam, and the Thirunakshatram (star birthday) entries — are still matched against the nakshatra at the moment you ask, not at sunrise. On about nine days a year that means the app can name a different festival in the morning than in the afternoon. We found this while doing the lunar work, it is measured, and correcting it is the next piece.

Traditional conventions we deliberately follow

  • Vakya nazhigai origin. Ghati–vighati end times in the Vakya system are counted from sunrise as the printed Tamil almanac defines it — the Sun's geometric centre on the true horizon, not the refraction-corrected civil sunrise a weather app shows — so the ghati values you see match the book. The Thirukanitha system keeps the civil convention, which is the one its own published sources use, so the two systems print sunrise a few minutes apart on purpose.
  • One instant, expressed locally. A tithi or nakshatra ends at a single moment everywhere on Earth. The printed almanac solves that moment once, for its own base town in Tamil Nadu, and each city edition prints it in local time and local ghati — which is why its Chennai, Mumbai and New Jersey pages carry the same end times once converted to a common clock. We do the same: the end times you see are the same instant a reader of the printed book sees, converted to your clock, while sunrise, sunset, horai, the kalams and the day strips remain your own city's.
  • The Tamil month begins by your own sunset. A solar month starts on the day the Sun enters the new sign if that happens before sunset, and on the next day if it happens after. Sunset is local, so two cities can start the month a day apart — and the printed editions do exactly that: for August 2026 the Chennai page runs Adi to a 32nd day while the Coimbatore page has already begun Avani, because the Sun entered Simha at 18:31 between Chennai's sunset (18:26) and Coimbatore's (18:37). We follow your city, as the book does.
  • The kalams divide your actual daytime. Rahu kalam, Yamagandam and Gulikai are the classical eight equal parts of sunrise–to–sunset, one assigned to each by weekday. That is why a block runs about 92 minutes at Chennai in August rather than a tidy 90 — the day is 12 hours 20 minutes. The familiar “Monday 7:30–9:00” table is the same rule flattened onto a nominal 6am–6pm day. You may find our times later than a printed almanac by roughly a minute for every eighth of the day that has already passed — so a minute or two for a kalam in the early morning, and up to eight or nine for one in the late afternoon. That is a rounding difference, not a different method. The Nrisimhapriya almanac we compare against rounds twice, and both roundings go the same way: it takes sunrise to the whole minute below, and it gives the block in whole minutes too, so an eighth of a 740-minute day is written as 92 rather than 92½. The first costs up to a minute before any block has elapsed, and the second accumulates — a block's worth of the dropped fraction for every block already gone. Its own ghati column, meanwhile, carries the day length to the vinadi — the day is known precisely and the coarsening is in the printed clock. We divide the day exactly, which is why our eight parts close on sunset where that page's fall four to seven minutes short of its own printed sunset. Measured against that publisher's Chennai pages for 25 August, 1 September and 4 September 2026; we have not checked other publishers and this says nothing about how they round.
  • Vara turnover at sunrise. The weekday for scoring changes at sunrise, not midnight, exactly as in tradition.
  • Rahu as the mean node. Rahu/Ketu positions use the mean lunar node, the convention of classical Vimshottari dasha software; Ketu is opposite Rahu.
  • Whole-sign houses. The natal chart (Lagna, house placements, yogas) uses whole-sign bhavas.

Gowri Panchangam, Abhijit, Durmuhurtham, Varjyam

The Gowri day and night sequences are the fixed weekday tables of the Tamil tradition, including the Saturday-night table's repetition of Soram. Abhijit muhurtham is the 8th of the fifteen day-muhurtas (omitted on Wednesdays, per tradition). Durmuhurtham windows are the classical weekday muhurtas. Varjyam (nakshatra thyajyam) starts after the tabled ghatis of each nakshatra's own span and lasts four ghatis of that span, timed by whichever engine is active so it always agrees with the nakshatra end times shown.

கௌரி பஞ்சாங்கம் — மரபு வார அட்டவணை; அபிஜித் — பகலின் 8-வது முகூர்த்தம் (புதன்கிழமை தவிர); துர்முகூர்த்தம், வர்ஜ்யம் — சாஸ்திர விதிப்படி.

The auspiciousness score

Classical jyotisha has no 0–100 number; Nava's score is a documented, normalized mapping of how a muhurta is traditionally judged — the weekday–star yoga (Amrutha/Siddha/Marana) and the nakshatra and tithi classes dominate; nithya-yoga and karana refine; Rahu kalam, Yamagandam and Gulika kalam are strong time penalties. A severe dosha (Marana yoga, an active kalam) hard-caps the score — good factors are never allowed to dilute it away. The personal score adds Tarabalam, Chandrabalam, Chandrashtama, Sade Sati and your running Vimshottari dasha–bhukti, judged from your own Lagna where your birth details allow a full chart.


Nava is a religious and cultural reference.

© 2026 Nava Panchangam
Home About Privacy Delete account