A technical review of a widely used Tamil almanac — where it excels, and one boundary case it silently mishandles.
Among Tamil-language astrology sites, ePanchang occupies a clear niche: a modern, no-cost, astronomically current almanac. It publishes the daily panchangam (tithi, nakshatra, yoga, karana), muhurtham timings and a free marriage-matching (Jathagam Porutham) tool that accepts each partner's rasi and nakshatra and returns the familiar porutham verdicts. Its calculations follow the Thirukanitha (Drik-ganita) method — the modern, observationally corrected computation of planetary positions — rather than the older Vākya table tradition. For the very large majority of its users, who are born in India under a single, fixed civil-time zone, it does exactly what it should, quickly and for free.
This review is written from the standpoint of an engineer building a matching engine, not a critic. The goal is to characterise, with reproducible computation, both what ePanchang does well and the one place where its model of time breaks — a place that becomes important precisely for the diaspora audience that online matching increasingly serves.
ePanchang's use of the Drik/Thirukanitha method means its nakshatra and tithi boundaries reflect the actual observed positions of the Moon and Sun, corrected for precession through the Lahiri ayanamsa. This is the same family of computation used by the most rigorous reference almanacs, and it avoids the several-degree drift that older Vākya tables can carry. For a user who simply wants an accurate nakshatra for a domestic Indian birth, this is the right engine.
The service is free, requires no account, and presents results in clean Tamil. The matching tool is particularly convenient because it accepts rasi and nakshatra directly — a user who already knows their star and sign can obtain a porutham verdict in seconds without re-entering a full birth record. For everyday domestic use this lowers friction considerably.
ePanchang follows the common 14-yoni convention — merging the ox and cow into a single "cow" (māḍu) yoni and ignoring the male/female distinction — and uses varṇa as the twelfth porutham. These are the widely shared, pan-Indian conventions, which makes its results easy to compare against North-Indian and other regional tools. (Whether this convention is better than the finer, gendered Tamil scheme is a separate question, treated in our companion review of srirangaminfo.)
In short: for an Indian-clock birth, ePanchang is a reliable, modern, free Tamil almanac. The weakness below does not detract from any of this — it applies to a specific and growing class of births.
Daylight Saving Time is the seasonal one-hour shift of civil clocks observed by the United States, Canada, most of Europe, the United Kingdom, Australia and others (India, notably, does not observe it). A person born in Toledo, Ohio on 15 June is born under Eastern Daylight Time (EDT, UTC−4), not Eastern Standard Time (EST, UTC−5). A matching engine that stores only a fixed "standard" offset per location, and applies it year-round, will place every summer birth in that region one hour off. This is one of the most common defects in astrology software, because DST rules are irregular, historical and jurisdiction-specific, and are easy to omit if one models a timezone as a single number.
The Moon is the fastest-moving body in the chart, traversing the zodiac in about 27.3 days — roughly
13.2° per day, or 0.55° per hour. The sidereal circle of 360° is divided into 27 nakshatras of 13.33°
each, each nakshatra into 4 padas of 3.33°, and into 12 rasis of 30° each. A one-hour error therefore moves the
Moon about 0.5° — roughly one-sixth of a pada. That sounds small, and usually it is: if the true
Moon sits comfortably in the middle of a pada, a half-degree shift changes nothing. But whenever the true Moon
lies within about half a degree of a boundary — a pada edge, a nakshatra edge, or a rasi edge — the same
half-degree shift carries it across, and the classification changes. Because pada, nakshatra and rasi boundaries
recur every few degrees, a meaningful fraction of all births fall in this danger zone.
To test the effect concretely, we compute the Moon's sidereal ecliptic longitude (tropical longitude from the astronomy-engine ephemeris, minus the Lahiri ayanamsa of 23.83°) at two instants for the same local birth: (a) the correct UTC obtained with full DST awareness via the IANA timezone database, and (b) the DST-omitted UTC obtained by applying the location's standard offset year-round. We then classify each longitude into (nakshatra, pada, rasi) and record whether the two disagree.
It would be easy, and wrong, to claim the origin case demonstrates the bug. It does not. Sindhu (born
15 June 1997, 20:46, Toledo, Ohio) has her Moon at 182.50° under correct EDT and 183.02°
under the DST-omitted offset — a shift of +0.52°, exactly as predicted. Both fall in
Chithirai, pada 3, Thula. She sits mid-pada, so the error changes nothing. We state this plainly because
it is the crux of the problem: you cannot tell in advance whether a given birth is a boundary case without
already knowing the correct time. The defect is invisible in most reports and surfaces only in the minority
that happen to land on an edge.
Searching real births across the summer of 1997 in the US Eastern zone, we find many that flip. Four representative cases:
| Local birth (US Eastern, DST on) | Correct (DST-aware) | DST-omitted (+0.5°) | What changes |
|---|---|---|---|
| 1 Jun 1997, 10:00 | Moon 359.98° — Revathi p4, Meena | Moon 0.56° — Aswini p1, Mesha | nakshatra + rasi (full flip at 0°) |
| 2 Jun 1997, 09:00 | Moon 13.33° — Aswini p4, Mesha | Moon 13.90° — Bharani p1, Mesha | nakshatra |
| 3 Jun 1997, 14:00 | Moon 29.98° — Krittika p1, Mesha | Moon 30.54° — Krittika p2, Rishaba | pada + rasi (30° edge) |
| 5 Jun 1997, 19:00 | Moon 59.62° — Mrigashira p2, Rishaba | Moon 60.17° — Mrigashira p3, Mithuna | pada + rasi (60° edge) |
Figure 1. Real 1997 Eastern-US summer births in which omitting the DST hour changes the classification. Sidereal longitudes via astronomy-engine + Lahiri (23.83°).
Each row is a birth in which a family relying on a DST-unaware tool would be handed the wrong nakshatra or rasi. The first row is the worst kind: a full flip across the 0°/360° seam, turning Revathi (Meena) into Aswini (Mesha) — a different nakshatra and a different sign, which rewrites essentially the entire porutham table. The third and fourth rows sit on rasi edges (30°, 60°): the nakshatra is unchanged but the sign flips, so the Rasi, Rasi-adhipati and Vasya poruthams — and the displayed lagna — all change, and the total score moves.
Marriage matching is a function of the two nakshatras and two rasis. When DST omission changes even one partner's nakshatra, the effect cascades: Yoni, Gana, Rajju, Nadi, Dina and Stree-Deergha are all nakshatra functions, so a nakshatra flip can swing the score by several poruthams and can turn an "உண்டு" (present) into an "இல்லை" (absent) — or the reverse — on the very poruthams families weigh most. A rasi-only flip is narrower but still changes the Bhakoot/Rasi group and can shift the verdict on a marginal match. In a domain where a single porutham can decide whether an alliance proceeds, an unnecessary one-hour error is not a rounding detail.
The affected population is specific but rapidly growing: anyone born in a DST-observing country during its summer window — the United States, Canada, the United Kingdom, continental Europe, Australia and others — which is precisely the diaspora that online matrimony now serves at scale. A child born at 9 p.m. on a July evening in New Jersey or London is an EDT/BST birth; a tool that quietly treats it as standard time is wrong by an hour for that birth. The Gulf states (a large Indian-expatriate population) do not observe DST, so those births are unaffected — a useful reminder that the defect is regional, not universal. Historical Indian DST (a brief wartime measure in 1942–45) is essentially irrelevant to today's overwhelmingly younger applicant pool.
The deeper issue is silence. Because the tool returns a confident, well-formatted nakshatra either way, neither the family nor the astrologer has any signal that a boundary birth was misclassified. The error does not announce itself; it simply produces a plausible wrong answer some of the time.
Our engine resolves every birthplace to an IANA timezone (via the zoneinfo/tzdata database) and converts local birth time to UTC with full DST awareness, so a June Toledo birth is correctly treated as EDT (UTC−4), not EST. The historical DST rules are applied automatically from the timezone database rather than from a single stored offset. We also present both the traditional (Vākya-style) and modern (Thirukanitha) matching conventions side by side, so that a boundary birth — where the two traditions or the two possible padas diverge — is visible to the user rather than hidden.
ePanchang is a strong, modern, free Tamil almanac that is entirely fit for domestic Indian-clock births, with accurate Drik computation and a convenient rasi/nakshatra matching tool. Its one systematic weakness is the omission of Daylight Saving Time correction, which — for the specific and expanding class of diaspora summer births near a boundary — can change the nakshatra, pada or rasi and thereby the marriage-matching result. The effect is small in degrees (about half a degree per hour) but decisive at boundaries, and it is silent: most reports are unaffected, and the affected minority is indistinguishable from the rest without the correct time. For a matching engine, the remedy is not exotic — it is simply to model civil time properly, DST included, which is what a correct engine must do.