How we calculate your chart

In BaZi, a year and a month don't follow the calendar. The year pillar changes at the exact moment of the Start of Spring (Lichun, 立春), and the month pillar changes at each of the twelve solar terms that open a month: the Start of Spring, the Awakening of Insects, Clear and Bright, and so on. Neither January 1 nor Lunar New Year is the dividing line. So if you were born in January or early February, your year pillar may belong to the previous year. In 2024, the Start of Spring fell at 17:27 Korean time on February 4, so someone born at 3 pm that day has the same year pillar as 2023: 癸卯.

A solar term begins when the Sun's apparent longitude reaches a multiple of 15°. We compute that moment ourselves, using the abridged VSOP87 series from Jean Meeus's Astronomical Algorithms. The Korea Astronomy and Space Science Institute (KASI) and the National Astronomical Observatory of Japan (NAOJ) publish these moments to the minute, rounding the seconds, and we round the same way. If you were born in the same minute as the published start of a term, you count as born in the new term.

We compared our times with 552 solar-term times for 2005–2027 from NAOJ's annual almanac. 516 matched to the minute. The other 36 were one minute apart. In 34 of them our seconds sat right around the 30-second rounding point (27.9 to 37.1 seconds); the other 2 were more than 11 seconds away from it, and we have not yet found out why. In practice, if you were born within a minute or so of a solar term, different published tables could give you a different year or month pillar.

For the year and month pillars we compare your actual moment of birth with the solar term. The true solar time adjustment described below is not used here.

Related term: Solar Term Boundary

The day pillar moves one step through the 60-pair cycle every day, with no gaps. We count it from the Julian Day Number, a continuous count of days, and check it against two fixed points: January 1, 2000 is a 戊午 day and October 1, 1949 is a 甲子 day. A lunar date is first converted to its solar date and then goes through the same count.

The hour pillar changes every two hours, and the hour branches turn over on the hour at 23:00, 01:00, 03:00 and so on. The hour pillar's Heavenly Stem follows from the day's stem.

The Zi hour (子時) runs from 23:00 to 01:00, so it straddles midnight. Schools disagree about whether a birth in the 23:00 hour belongs to that day or the next.

By default, AevoStar uses the single Zi hour: at 23:00 the next day has begun, and both the day pillar and the hour pillar follow the next day. If you choose Late and early Zi in the options, a birth in the 23:00 hour keeps that day's day pillar, while the hour pillar's stem follows the next day. If the adjusted time (the clock time when the solar time adjustment is off) falls between 00:00 and 01:00, both settings give the same chart. With the adjustment on and Late and early Zi chosen, though, a birth just after midnight on the clock can fall into the previous evening's 23:00 hour once adjusted, and then takes the previous day's day pillar.

We want to be upfront about the Late and early Zi option. We have not found a primary classical source saying that the hour stem for a 23:00 birth should follow the next day. The rule rests on a single principle: the 60-pair cycle of hour pillars should run without a break. Researchers disagree on when the split of the Zi hour began, and schools still reach different conclusions, so we don't present either option as the right answer.

One more detail: whether you were born in the Zi hour is judged after the true solar time adjustment. A birth in Seoul at 23:20 on May 12, 1990, for example, becomes 22:52 once adjusted. That is the Hai hour (亥時), not the Zi hour. The shift varies with the date, so on another day the same 23:20 can fall in the Zi hour.

Related term: Late and Early Zi Hour

Korea Standard Time is set to the 135°E meridian. Seoul lies to the west of it, so while that meridian is in use the Sun crosses Seoul's meridian on average about 32 minutes after clock noon. On top of that comes the equation of time: over the year, the Sun runs up to about 14 minutes behind or 16 minutes ahead of that average.

AevoStar applies both by default, and you can switch the adjustment off. The shift is (birthplace longitude − standard-time meridian) × 4 minutes plus the equation of time at the moment of birth, rounded once to the nearest minute. For Seoul that comes to about −47 to −16 minutes, depending on the date, while the 135°E meridian is in use, and about −16 to +14 minutes in 1908–1911 and 1954–1961, when it was 127.5°E. The meridian comes from standard time with daylight saving removed, so a birth during summer time isn't adjusted twice. The same formula works anywhere in the world, including half-hour and 45-minute zones such as India (UTC+05:30) and Nepal (UTC+05:45).

The adjustment is used only for the day and hour pillars. As described above, the year and month pillars use your actual moment of birth.

A limit worth knowing: your birthplace is read from a single reference point, such as a provincial capital or a chosen city, and the shift is rounded to the minute, so the far edges of one province can differ by a few minutes. Because we apply both the longitude difference and the equation of time, we call it a true solar time adjustment. Calculators that adjust for longitude only are also in use, and against them the hour pillar can differ for a birth near a block edge. Your calculation notes list the longitude part and the equation of time separately, so you can compare.

Related term: True Solar Time Adjustment

Korea's standard time has changed several times, and the country used daylight saving time in some years. For births in Korea we use our own table, checked against government records and historical sources. For everywhere else we rely on the history in the IANA time zone database.

Korea's 1948 daylight saving is confirmed by records of the National Institute of Korean History and a proclamation issued that May. We first used the IANA end date (00:00 on September 13), but the text of the decree sets the clocks back an hour at midnight on September 22, so in the October 2026 revision of our calculation basis we moved the end to 24:00 on September 22 (summer clock). Birth times recorded in Korea from September 13 to 22, 1948 are now read as summer clock time, an hour ahead of standard time. Comparing every minute of those ten days for a Seoul birth, the change moves the hour pillar for about half of them.

When clocks spring forward, an hour of clock time never happens; when they fall back, an hour happens twice. If you enter a time that never happened, we read it as standard time. If you enter a time that happened twice, we take the first one, which is daylight saving time. Either way we accept the time, and your result's calculation notes flag it as a daylight saving edge.

If you were born outside Korea, enter the clock time people actually used there when you were born. You pick your birthplace from 34,149 cities based on GeoNames data, and the IANA history tells us whether that time was daylight saving time and which standard time applied. Before a place adopted official standard time, we use the mean time it kept. France, for example, ran on Paris Mean Time nationwide from 1891 to 1911, so a birth in Lyon in those years is adjusted by about +10 minutes, while a birth in Paris is not adjusted.

A few dates vanished entirely when a time zone jumped across the International Date Line, such as December 30, 2011 in Samoa. Nobody there lived through that date, so we don't accept it and ask you to enter the day before or after.

You can enter a lunar birthday, marking whether it fell in a leap month, for years 1900–2100. The lunar date is converted to its solar date before anything else is calculated.

Our lunar tables are computed astronomically. A month begins on the day of the new moon, the month containing the winter solstice is the 11th month, and when 13 months fall between two winter solstices, the first month without a major solar term becomes the leap month. We keep two tables built this way, one for the Korean lunar calendar and one for the Chinese calendar. We compared all 1,867 Korean lunar months from 1900 to 2050 with KASI data, and every leap month fell in the same place.

That comparison turned up eight month starts between 1903 and 1911 that came out one day late. The cause: Korea then used the Qing dynasty's Shixian calendar, which placed new moons by UTC+08:00. We now use that reference for dates before 1912, and all eight match KASI. From 1912 on, the Korean table uses Korean time, so when a new moon or major solar term falls near midnight in China, a month can start one day apart from the Chinese calendar, or the leap month can fall in a different place.

From 2051 on, Korean lunar dates are beyond the range KASI provides, so they come from our astronomical calculation alone.

The Chinese table follows the same rules with a different reference time. New moons and major solar terms are placed by UTC+08:00, except from 1912 to 1928, when the almanacs of the day used Beijing local mean time (UTC+07:45:40). We compared it with the Hong Kong Observatory's published Gregorian–lunar conversion tables for 1901–2100, 2,473 lunar months in all. One month start, the 9th month of 2057, falls on a new moon just after midnight, within our calculation's margin, so we follow the published date. With that, every New Year and leap month matches, and 2,472 month lengths match. The remaining month, the 12th lunar month starting on December 31, 2100, can't be checked because the published table ends on its first day.

Lunar dates are accepted only for births in places whose calendar one of these two tables covers. Right now that means South Korea, North Korea, China, Taiwan, Hong Kong, Macao, Singapore and Malaysia. Births in South and North Korea are converted on the Korean lunar calendar, and births in China, Taiwan, Hong Kong, Macao, Singapore and Malaysia on the Chinese calendar used there. The two calendars usually agree, but across the full tables the leap month falls in a different place in 2012, 2017 and 2099 (in 2012, the Korean leap 3rd month against the Chinese leap 4th month), and in other years some months start a day apart. In some years, such as 2027 and 2028, Lunar New Year itself falls a day apart. When we convert on the Chinese calendar, your result's calculation notes say so.

We left out Vietnam: its lunar calendar is reckoned on UTC+07:00, which moves some month starts by a day against both calendars and can shift leap months, and we'd rather not convert it and pretend it fits.

Choose "I don't know the time" and we leave out the hour pillar and read the other three pillars, six characters in all.

To settle the year and month pillars and your Luck Pillar starting age, we use a reference moment: noon Korea Standard Time for births in Korea (daylight saving not considered), and noon on the local clock elsewhere (with daylight saving applied if it was in effect that day). With no hour pillar, there's no true solar time adjustment either.

If you happen to have been born on the day a solar term began, your month pillar depends on whether you were born before or after that moment, and so does your year pillar if the term was the Start of Spring. When that's the case, your result's calculation notes tell you. Your Luck Pillar starting age may also be off by about a year.

Luck Pillars start from your month pillar and change every ten years. They run forward, starting with the pillar after your month pillar, if your year stem is yang and you are male, or yin and you are female. Otherwise they run backward, starting with the pillar before it. Your result shows ten Luck Pillars.

The age your first Luck Pillar begins is the number of days from your birth to the next solar term (forward) or from the previous solar term to your birth (backward), divided by three and rounded. A result of 0 becomes 1, and it never goes above 10. The days are measured using solar-term times to the second.

The start year is your solar birth year plus that number, so these ages count the years since you were born: international age. Calculators that count ages the traditional East Asian way add a year, so the same Luck Pillar can appear one year later in age there. The classic Sanming Tonghui (三命通會) likewise describes this count as time actually elapsed after birth. Which Luck Pillar you're in now is decided by today's year in Korea time.

Related terms: Luck Pillars · Natal Chart

• One reference point per birthplace. We use one longitude per Korean province or city, or the city chosen abroad (a stand-in city if you choose only the country), and round the shift to the minute, so the far edges of a province can differ by a few minutes.
• Solar-term times after 2050 are less certain. We compute in Terrestrial Time and convert with ΔT, the difference between Terrestrial Time and Universal Time. ΔT for the distant future can only be estimated, so by around 2100 the times could be off by a few minutes.
• Mainland China uses China Standard Time (UTC+08:00) throughout from October 1, 1949. If you pick a western city such as Ürümqi from the list, we adjust from China Standard Time rather than the locally used Xinjiang time (UTC+06:00). The longitude part alone can reach −130 minutes, before the equation of time, which may move the hour pillar by one. Births before October 1, 1949 follow the time zone of the day.
• Ten symbolic stars only. Schools define symbolic stars differently, so we show ten representative ones, for reference.
• No overall compatibility score. Compatibility shows five aspects, each on a three-step scale: the Day Masters, the day branches, the Five Elements, cross-pairs of characters, and the current Luck Pillars.
• Some Korean reference coordinates are still being confirmed. The reference points for Daegu, North Gyeongsang, North Jeolla, Jeju and Sejong are still to be checked against official sources.

On your result, tap "See how this was calculated" to see which rules actually applied to your chart. Depending on your details, it lists the Zi hour setting, the true solar time adjustment, daylight saving, lunar conversion, the standard time in use, a birth on a solar-term day, a daylight saving edge, an unknown time on a solar-term day, local mean time before standard time, or a stand-in city. Change an option and we recalculate and tell you which pillars changed.