Why two apps give you two different nakshatras

Put the same birth details into two apps and you can get two different nakshatras, or the same nakshatra with a different pada. Often neither has made an arithmetic mistake. They made different choices, or were given slightly different inputs, and your Moon sat close to a line.

The zodiac is cut into 27 nakshatras of 13 degrees 20 minutes and 108 padas of 3 degrees 20 minutes. Two apps disagree only when they put your Moon on opposite sides of one of those lines. So every disagreement has two numbers behind it: how far your Moon sat from the nearest line, and how far apart the two apps are.

If the other app's Moon is later by roughly the whole ayanamsa, 23 to 24 degrees for a birth since 1950, it is probably using the tropical zodiac, and that is a different question. What follows is about two sidereal apps.

The ayanamsa

Every sidereal position this app computes is a tropical longitude minus one number, the ayanamsa for that instant. It uses Lahiri, also called Chitrapaksha, computed as a straight line: 23.85306 degrees at noon on 1 January 2000, plus 0.0139601 degrees a year, about 50 arcseconds. No star is looked up and no short-period term is added.

Lahiri is not the only ayanamsa; Krishnamurti Paddhati (KP), for one, uses its own. A different one slides every sidereal longitude by the difference between the two figures. A difference of D arcminutes changes the nakshatra only for positions within D arcminutes of one of the 27 lines, on the side the shift carries them across: D/800 of the circle, and D/200 for the pada. A sample of 200,000 birth moments from 1950 to 2020, run through this app's own Moon, agrees.

This page cannot give you the gap between Lahiri and any other named ayanamsa: this app's code implements no other, and says outright that the project's library holds no value for KP. Finer divisions are more sensitive still: for the sub-lord, the same D arcminutes matters on D/88.9 of the circle.

One ayanamsa gap, ten chances to differ

The table above is for the Moon alone. But an ayanamsa difference does not act on the Moon alone: it shifts the Sun, every planet, both nodes and the ascendant by exactly the same amount. Each of those ten points has its own D/200 chance of sitting close enough to a pada line to cross it. So the chance that at least one of them changes pada is much higher than the chance for any one.

That is why two charts cast with different ayanamsas so often "look different" somewhere even when they agree about the Moon: some planet's pada, and with it its Navamsa sign, has moved. The Navamsa is the most visible casualty, because each pada is one Navamsa division. If a comparison of two apps shows the same rashi chart but a different Navamsa chart, a small ayanamsa difference is the first suspect.

The birth instant

The Moon is the fastest body in the chart. Measured from this app's own Moon between 1950 and 2030, it covers 11.8 to 15.4 degrees a day, roughly 29 to 38 arcminutes an hour, so a pada lasts about five to seven hours of clock time. In the 200,000-birth sample, a birth time 15 minutes later changed the pada 4% of the time; an hour later, about 16.5%, and the nakshatra 4%.

A missed summer-time hour is an error of exactly that size, and a clock time read in the wrong zone can be further out. This app's calculators use your birth place's time zone when they can resolve one and assume India Standard Time otherwise, and with no birth time they read 12:00 noon. An app that assumes differently is computing a different moment, and what a missing time costs depends on your specific date.

Old births and old clocks

For older births there is a subtler version of the zone problem: the clock the time was read from may not have been on the zone a modern app assumes. The case-study book discussed below shows this in its own chart headers. Most of its Indian charts are cast with a zone meridian of 82 degrees 30 minutes east - India Standard Time, five and a half hours ahead of Greenwich. But several older Indian charts use a zone of 88 degrees 20 minutes east instead, and one 1905 chart from Lahore uses the birthplace's own longitude, 74 degrees 56 minutes, as its zone - that is, local time by the Sun.

The arithmetic of those differences is simple. Each degree of longitude is four minutes of time, so 88 degrees 20 minutes east is 5 hours 53 minutes 20 seconds ahead of Greenwich, 23 minutes 20 seconds more than IST. A birth time written on a clock set to that meridian and entered today as IST is read 23 minutes 20 seconds late - about 0.2 degrees of Moon, and about six degrees of ascendant. The book does not explain its choice of zone for each chart, and this page does not guess at the history; the point is only that an old recorded time carries an implied zone, and two apps that assume different ones are computing different moments.

This app reads the time zone of the resolved birth place where it can, rather than a fixed offset. Whether that matches the clock your birth time was actually read from is a question about your own records, not about the software, and for a birth long ago it is worth asking.

The Moon's own position

This app does not run a full ephemeris. Its Moon comes from 35 principal terms of the lunar series in Meeus's Astronomical Algorithms, which the project's own documentation rates at about 0.1 to 0.3 degrees. An app on a full ephemeris can differ by that much with every other choice identical, and at the top of that range, 0.3 degrees or 18 arcminutes, roughly one Moon in eleven sits close enough to a pada line to land differently.

It is worth turning that into the units people actually care about. An error of 0.3 degrees in the Moon is 18 arcminutes. By the D/800 rule, that changes the nakshatra itself for about one birth in forty-four. And because each arcminute of Moon position is worth 2.7 to 9.1 days of dasha dates, 18 arcminutes is worth roughly seven weeks to five months on every later Mahadasha boundary. That is larger than many Pratyantardashas, which is why this app tells you to read its third dasha level as "which stretch, give or take weeks". An app on a full ephemeris can place these boundaries differently by up to that much even with every other choice identical.

Antardasha and Pratyantar has the full treatment of what that does to the sub-periods, and the methodology page describes the astronomy this app runs on.

The ascendant disagrees most of all

If the other app also shows a different ascendant nakshatra or pada, that is the least surprising difference of all. The ascendant moves through the whole zodiac in a day - about a degree every four minutes on average - so it crosses a pada line roughly every quarter of an hour. Two apps a few minutes apart in their reading of the birth instant, or using slightly different methods for the ascendant, will often disagree on its pada and sometimes on its nakshatra while agreeing about the Moon. For the ascendant's sub-lord, which this app computes in the sub-lord calculator, the window is a minute or two.

Rahu and Ketu: which node

Rahu and Ketu can be computed as the mean node, a smooth average, or the true node, which swings either side of it. This app uses the mean node everywhere, with Ketu exactly opposite. Measured with the correction this code carries for diagnostics, the gap between the two reached about 1.88 degrees between 1970 and 2015. Checked daily, Rahu's pada differed on 29% of days and its nakshatra on 7%.

That touches neither your birth nakshatra nor the Vimshottari Dasha, both of which come from the Moon. No text in this project's library says which node its rules were written for, so the choice is documented rather than settled. The one place the app reports the other node is the Kaal Sarp calculator, which says when its verdict would flip on it.

A case on the line

A birth on 19 October 1995 at 11:05 India time puts this app's Moon at 120.0995 degrees: 6 arcminutes into Leo and into Magha, pada 1, lord Ketu. Entered as 10:50, the same engine returns 119.9691: Cancer, Ashlesha, pada 4, lord Mercury. An ayanamsa 6 arcminutes larger than this app's would move the 11:05 birth across the same line.

The labels flip; the dates barely move. Each nakshatra's lord follows the previous one's in the dasha sequence, so the timeline runs straight through the line. On the shifted chart Mercury's period ends within hours of birth, Ketu's seven years follow, and every later Mahadasha boundary lands about 19 days later. Each arcminute of Moon position is worth 2.7 to 9.1 days of dasha dates, from the Sun's six years to Venus's twenty.

What else flips with it

The 1995 case is a particularly sharp one, because the line at 120 degrees is not only a nakshatra line. It is also the line between Cancer and Leo. Nakshatra lines and sign lines coincide at only three points in the zodiac - 0, 120 and 240 degrees - where the cycle of nine nakshatra lords starts again. At those three points a small shift changes the Moon sign as well as the nakshatra. Everywhere else, a nakshatra can change while the sign stays put.

Here is what changes for the 1995 birth between the two readings, using this app's own tables:

So a disagreement that looks like one label is several. The two readings agree on Gana and Nadi, but every factor of Guna Milan that reads the Moon sign - Varna, Vashya, Graha Maitri and Bhakoot - is computed from a different sign, and the Yoni changes from Cat to Rat, which the engine's own table lists as a natural-enemy pair. A match computed by two apps on either side of this line can come out several points apart without either app making an error. Moon sign and nakshatra sets out which matching factors read the sign and which the nakshatra.

The dates, as above, barely move: Mercury's seventeen-year period is almost over at birth on one side, and Ketu's seven-year period has barely begun on the other, so the boundary that follows lands in nearly the same place. The Ashlesha and Magha pages describe each nakshatra in its own right; for a Moon this close to the line, both are worth reading.

This app's Lahiri, checked against a published book

The earlier version of this page said that the engine's ayanamsa constant had been checked only against this project's own documents. That has now changed. One of the books in the project's library, Advanced Predictive Astrology (2000), prints a computer-cast chart for each of its case studies, and each chart header gives the date, time, zone and the ayanamsa used. The author states in the preface that the book follows the Lahiri ayanamsa only. That makes the headers an external table of Lahiri values, produced by someone else's software, against which this app's formula can be checked.

Across eight charts spanning 121 years, the two agree to within 18 arcseconds - three tenths of an arcminute. By the D/800 rule above, a gap that size changes the nakshatra on about 0.04 percent of births and the pada on about 0.15 percent. In other words: when this app and software that uses Lahiri disagree about your nakshatra, the ayanamsa formula is very unlikely to be the reason. The cause is much more likely to be the birth instant or the Moon's computed position.

The differences vary from chart to chart, between about 1 and 18 arcseconds, in both directions. A short-period wobble of the earth's axis of roughly that size exists, and the engine deliberately leaves it out, so it is a plausible explanation. But the book does not say how its software computed the values, so this page does not claim that is the cause. It reports the agreement and its size.

The book's values also check the formula's rate, not only its starting point. Between its 1856 chart and its 1977 chart - about 121.2 years apart - the printed ayanamsa grows by 1 degree 41 minutes 41 seconds, which is 6,101 arcseconds, or about 50.3 arcseconds a year. This app's formula adds 0.0139601 degrees a year, which is 50.26 arcseconds. Over a century the two rates differ by a few arcseconds, consistent with the scatter in the table.

The same rate gives a sense of scale for a single life. Over eighty years the ayanamsa grows by about 1.1 degrees. That matters for one practical reason: transits are computed with the ayanamsa of the transit date, not of the birth date, so a planet passing over your natal Moon in 2026 is measured against a zodiac that has slipped about half a degree since a 1990 birth. Every app that does this correctly will agree; an app that froze the ayanamsa at one value would drift. In a day, by contrast, the ayanamsa moves about a seventh of an arcsecond, which is why the hour of birth makes no difference to it.

What the same book says about KP

The same preface makes one more claim worth reporting. It says the Krishnamurthy (KP) ayanamsa is "very close" to Lahiri - close enough that, in its illustrations, a planet's Navamsa position did not change when one was replaced with the other. Since a Navamsa is the same 3 degrees 20 minutes slice as a pada, that is a statement that the difference was small enough not to move any pada in that book's charts. It is one author's observation about one set of charts, not a published offset, and this app still holds no value for KP. But it bounds the likely size of the gap: KP users and Lahiri users should mostly agree on the nakshatra and pada, and disagree more often only at the finer level of the sub-lord.

What this cannot tell you

Which ayanamsa is right. Arithmetic does not settle it, and neither does this project's translation of Brihat Parashara Hora Shastra: a full-text search of it finds the word ayanamsa zero times.

Nor can it tell you what a nakshatra "really" is for a Moon a few arcminutes from a line. Both readings in the 1995 case are the correct output of a reasonable method. What can be said is how close to the line you are, which choices would move you across it, and what changes downstream if they do. A reading that hides the closeness is less honest than one that shows it, even though it looks more certain.

A case far from the line

The 1995 Moon sat 6 arcminutes from a line. Most Moons do not, and it is worth working one that does not, because it shows how quickly the list of possible causes shrinks. Take the birth used in the Mahadasha essay: 15 August 1990, 10:30 IST, New Delhi. This app puts the Moon at 48.92 degrees - Rohini, pada 3.

Now suppose another app gives this birth pada 4, or Mrigashira. No ayanamsa difference anyone uses can explain it: the gap would have to be more than a degree for the pada and more than four degrees for the nakshatra, and, as the book comparison shows, two Lahiri implementations agree to arcseconds while the KP gap is small enough not to move padas in that book's charts. A different Moon model cannot explain it either, at 0.3 degrees at most. What can is the birth instant: the other app is reading a moment about two hours later, or eight hours later for the nakshatra. That points straight at a time-zone or AM/PM error, or at the wrong date.

This is the most useful habit the page can leave you with. Before arguing about systems, find how far your Moon is from its nearest line - the nakshatra calculator shows the degree - and compare it with the sizes on this page. Most of the time, one cause is the only one big enough.

A diagnostic order

When two apps disagree, the causes above differ enormously in size. Checking them largest first is the quickest way to find the real one.

The comparison is easiest with full longitudes. This app's birth chart calculator prints every planet's sign and degree, so you can line them up with the other app's chart planet by planet. If the Sun, Jupiter and Saturn agree to the arcminute and only the Moon and ascendant differ, the ayanamsa is not your problem - the birth instant is. If everything differs by the same amount, it is the ayanamsa, and the difference you read off is the gap between the two apps' values. The Lagna calculator is a useful second check, because the ascendant moves about a degree every four minutes and so exposes a time or zone error that the Moon alone might hide.

If the instant is the problem, check the zone first. A birth outside India entered without a place is read here as India Standard Time, and a birth in a country that uses summer time is an hour out if the summer offset is missed. Both are errors of the input, not of either app, and both are fixed by entering the place. What an unknown birth time costs has the per-hour figures.

Which app is right?

If both use Lahiri and the same birth instant, they should agree to well under an arcminute, as this app does with the book values above. If they use different ayanamsas, both are right by their own convention; no text in this project's library settles which convention to prefer.

Does BPHS say which ayanamsa to use?

No. A full-text search of both translations in the project's library finds the word ayanamsa zero times. The later case-study book names Lahiri and KP; the classical text names neither.

Can a nakshatra disagreement change my Mahadasha?

It changes the lord of the first period, but because each nakshatra's lord follows the previous one's in the dasha order, the dates of later periods move only by days per arcminute of difference.

The rashi charts match but the Navamsa charts do not. Why?

Because each Navamsa division is one pada wide, and a small shift - an ayanamsa gap or a few minutes of birth time - moves some planet or the ascendant across a pada line far more often than across a sign line. Navamsa, the second chart covers how sensitive that chart is.

Should I just pick the app whose reading I prefer?

Better to find the cause. If it is an input error, one app has the wrong birth moment and its whole chart is someone else's. If it is a convention, decide which convention you are following and use it consistently, rather than switching to whichever answer reads better.

Why does the other app show a different Rahu?

Almost certainly the true node instead of the mean node. The two differ by up to about 1.9 degrees; this app uses the mean node throughout and reports the other only in the Kaal Sarp calculator.

My Moon is far from any line, but the apps still disagree. Why?

Then the difference is larger than an ayanamsa or Moon-model gap can explain. Check that both apps received the same date, time, place and time zone - a day or an hour of input difference is far bigger than any of the method differences on this page.

Checking your own

The calculator below returns the ayanamsa this app subtracts for any date. Enter your birth date; the value moves about 0.14 arcseconds a day, so the hour does not matter. For the 1995 case above it returns 23.79 degrees, 23.7944 to four places.

If the other app prints its ayanamsa, subtract the two and convert the difference to arcminutes. Take your Moon's longitude from the Moon row of the sub-lord calculator, which takes your birth time; with no time, the nakshatra calculator gives the noon Moon. Pada lines fall on every multiple of 3.3333 degrees, every fourth a nakshatra line. If the other app's ayanamsa is larger, look at the line just below your Moon; if smaller, the one just above. If your Moon is closer to that line than the difference, that is your explanation. If not, check the birth time and time zone next.