How these times are calculated

Every window on this site is computed from your city’s own coordinates and the real position of the sun. Nothing is copied from another site, nothing is interpolated from a nearby city, and nothing is fetched while you read the page. This is the whole method, including the places where we disagree with other panchangs and why.

Two layers, and only the first is astronomy

A panchang time is two things stacked. The bottom layer is astronomy: when the sun actually rises and sets where you are, which depends on your latitude, longitude and the day of the year. The top layer is division: Muhurta Shastra cuts that span into equal parts and names them. The astronomy can be right or wrong; the division is a convention recorded in the texts, and different traditions record it differently.

Keeping the two apart is the reason this page can be specific. When we differ from another panchang it is almost always the second layer, not the first, and we say which.

Sunrise is a definition, not an observation

There is no single moment called sunrise. It depends on whether you mean the sun’s centre or its upper edge, and whether you correct for the atmosphere bending light over the horizon. The choice moves every time on this page by one to three minutes, and it moves Rahu Kaal and the choghadiya with it.

We use the upper limb with atmospheric refraction, the civil convention, and the one Indian panchangs use. It was chosen by testing all four conventions against published reference times and keeping the one that matched.

What a panchang actually is

Pancha-anga means five limbs, and all five come from two numbers: the ecliptic longitudes of the sun and the moon. That is the whole system, and it is why a panchang can be computed for a date thirty years from now or three hundred years ago. It is orbital mechanics, not a record of observations.

Indian astronomy measures these against the fixed stars rather than the drifting equinox, so the longitudes have the ayanamsa subtracted. We use Lahiri (Chitrapaksha), the Indian government standard, and we compute it rather than looking it up: the Calendar Reform Committee fixed it at 23°15′ for 21 March 1956, and everything since is precession. Computing it from that definition means it stays right centuries out instead of running off the end of a table. As a check, our value passes through zero in 285 CE, the year the two zodiacs coincide, which is what “Chitrapaksha” means, and which we never tuned it to hit.

Each timing, exactly

Eclipses are computed, not listed

When a grahan falls in the day, the page says so and marks the muhurats void through sutak and the eclipse: twelve hours before a solar eclipse, nine before a lunar one.

Two details matter. First, we compute eclipses from orbital mechanics rather than shipping a list of dates, so there is no year in which the site quietly stops knowing about them. Second, visibility is worked out for your city, not for the country: an eclipse below your horizon is not your eclipse. On 26 June 2029 there is a total lunar eclipse that India cannot see at all, because it peaks at 8:53 in the morning with the moon down. A date-and-region list would mark that day void for every Indian city. Ours does not. A penumbral lunar eclipse is likewise not treated as a grahan, and carries no sutak.

How this is checked

Accuracy claims are cheap, so here is the evidence rather than the adjective.

• Against a published reference. Every window compared against an established panchang across sampled days from 2016 to 2026: sunrise, sunset, Rahu Kaal, Yamaganda, Gulika Kaal, Abhijit, Brahma Muhurta and Dur Muhurtam, plus1,360 individual choghadiya periods. Zero disagreements beyond rounding; the largest gap anywhere was 41 seconds.
• Against a second, independent implementation. An unrelated open-source panchang engine runs against ours on every build, across three cities and 2010–2035. Where they differ, it is on the two conventions listed below, and it is deliberate.
• Against itself, over 110 years. From 1950 to 2060 and from 53°S to 60°N, the eight choghadiya tile the daylight exactly, the twelve horas hand off to the next weekday’s ruling planet at every sunrise, and every window lands where its division says it should.

Where we differ from other panchangs

Panchang traditions genuinely disagree, and a site that hides that is not being accurate, only confident. Two places where you may see a different number elsewhere:

• Dur Muhurtam. Widely reproduced tables give Monday, Thursday, Friday and Saturday a second window. The reference we calibrate against lists one window on every weekday, and we follow it. If your family panchang shows a second, that is this disagreement, not an arithmetic error.
• Brahma Muhurta. Some sources place it a fixed interval before sunrise. We take the 14th of the fifteen muhurtas of the actual night, so its length and position shift with the season. The two can sit almost an hour apart in winter.

Times are shown to the minute. The underlying calculation carries seconds, so a boundary may round a minute differently from another source that also rounds.

What this site does not do yet

There is no festival calendar, and no Varjyam, Amrit Kala, Bhadra or Panchaka. Those are built on tithi and nakshatra boundaries, and they are not published here until they meet the same standard of checking as everything above.