Home › Notes › Where manseryeok results diverge

Why manseryeok apps
give different results

Manseryeok · 2026.08.27 · 9 min read · Updated 2026.10.03 · Updated 2026-10-10

It's unsettling when you enter the same birth date and time and each app gives you a different Saju. You don't know which one to trust. To give you the bottom line first: it's mostly a difference in calculation assumptions. And some of those assumptions aren't a matter of taste but a matter of checking facts.

I've sorted the points of divergence into six, along with who each one affects.

1. Whose clock were the solar-term times written in?

The moment a solar term occurs is the same worldwide — it's the sun's position, after all. But which standard time it's written in varies from source to source.

Korea Standard Time is based on 135°E and China Standard Time on 120°E, so they're exactly 1 hour apart. If you take manseryeok data made in China and use it as-is, the entry times come out an hour early.

When reading the Saju of someone born in Korea, Korean time is the standard. Mutal Unnie calculates solar-term entry times in Korea Standard Time and verified them against Korean data. Entry times for each year are listed to the minute under 2027 and 2028: the 24 solar terms.

2. If you were born after 11 p.m.

It comes down to whether the day begins at midnight or at 11 p.m., when the Ja hour begins.

Under the jasi-ilsu view (the whole Ja hour belongs to the next day), the next day has already begun at 11 p.m., so the day pillar moves forward a day; under the late-Ja-hour (yajasi) view, the day pillar stays on that day and only the hour pillar is set to the Ja hour. For someone born at 11:30 p.m., both characters of the day pillar can be completely different depending on the app.

This isn't a matter of fact but a matter of interpretation. So for births in this hour, the Mutal Unnie calculator shows the results of both schools side by side. There's more in Born after 11 p.m.? — the late and early Ja hours.

3. If you were born during daylight saving time

Korea has had daylight saving time too — in twelve different years, no less.

During those periods, clocks were set an hour ahead. If your birth record says 3 p.m., the sun's actual position corresponded to 2 p.m. Without a correction, your hour pillar shifts by one slot.

Take someone born at 9 a.m. on June 10, 1958: correcting for both daylight saving time (−60 minutes) and the 127.5°E standard time (+30 minutes) gives 8:30 a.m. The Mutal Unnie app calculates with the clock time exactly as you entered it, and in the Manseryeok calculator you can switch the two corrections on and off to compare. This sits on the boundary where the hour branch changes from Sa (巳) to Jin (辰), so if the results split here, the whole hour pillar changes.

Which years, and from which day to which day, are laid out in a table in Twelve years of daylight saving time — people whose hour pillar shifts by one.

4. If you were born in 1954–1961

There was a period when Korea Standard Time differed from today's. From March 21, 1954 to August 9, 1961, Korea used 127.5°E as its standard meridian. That clock ran 30 minutes behind today's (135°E).

To convert the birth time of someone born in this period to today's standard, you need to add 30 minutes. At an hour-pillar boundary, 30 minutes is no small amount.

I've laid out how standard time has changed over the years in a timeline in The three changes to Korea Standard Time.

5. Is true solar time applied?

Our clocks are based on 135°E. But Seoul sits at 126.98°E. That's about 8 degrees to the west, so the sun actually reaches its highest point about 32 minutes later than the clock says.

On top of that comes an error of about ±16 minutes because the Earth's orbit is elliptical (the equation of time). Together, the difference can exceed 40 minutes.

Whether to correct for this depends on the school. Those who correct say “Saju is about the sun's position”; those who don't say “our ancestors used their local clocks, too.” Mutal Unnie calculates without the correction by default, and lets you switch it on and off in the calculator to see how the result changes.

6. Is the lunar calendar calculated for Korea?

This is where it's not the Eight Characters but lunar dates that diverge. It matters when you look at Son-eomneun-nal or lunar birthdays.

The lunar calendar takes the day on which the new moon (sak 朔) — the moment the moon lines up in the same direction as the sun — falls as the first day of the month. But the date depends on which country's clock you read that moment by. If the new moon occurs just after midnight Korean time, it's still the previous day in China.

How often do they diverge? Comparing everything from 1900 to 2050, the Son-eomneun-nal verdict differed on 427 days. 427 days over 150 years works out to about three days a year on average. Rare, but that day could just happen to be your moving day.

Mutal Unnie calculates the lunar calendar using new moons in Korea Standard Time and the no-zhongqi leap-month rule (a month without a mid-month solar term becomes the leap month). We checked Seollal dates and the list of leap months against public data.

So which one is right?

Points 1, 3, 4 and 6 are matters of fact. That daylight saving time was in effect, that standard time changed, and that solar-term and new-moon times are written differently in different countries are all verifiable facts. These should be reflected.

Points 2 and 5 are interpretation. It's an area where centuries of debate haven't been settled, so it's hard to say either side is wrong. That's why, instead of picking an answer, we chose to show both results.

How was the Mutal Unnie engine verified?

Claiming that our own calculations are correct means nothing. It only counts as verification when you state what you checked them against.

In the four-pillar cross-check, 3 cases didn't match, and on inspection all of them were within 60 minutes of a solar-term boundary. We use Korean time and the reference used Chinese time, so naturally the results split there. It's the difference described in point 1 above, showing up in practice.

Where we could be wrong, too

I'll be upfront about these.

• Lunar dates before 1912 may differ from public data. On October 3, 2026, we compared every date from 1900–2050 with a lunar–Gregorian library based on KASI data (korean-lunar-calendar): everything from 1912 on matched, but 240 days between 1903 and 1912 were off by one day in their lunar date. We're still looking into the cause, so if you were born in this period, please check your lunar date against public data as well.
• Births before 1900 aren't covered. We set the calculation range at 1900–2050. Outside it, the calculator won't give an answer.
• The equation of time used for true solar time is an approximation. The error is on the order of ten-odd seconds, so it's fine for practical use, but it may differ from a precise calculation where every second counts.
• Birth records themselves are often inaccurate. In the past, births were commonly registered days late, or with the time rounded off. However accurate the calculation, it's no use if the input is wrong. See also When you don't know your birth time.
• Interpretation isn't something we can verify. What we've said is verified here is the calculation, not the reading. How accurate the readings are, we publish in the Hit Ledger — misses included.

Check whether your Saju falls on a boundary

Enter your birth date and time in the calculator to see which corrections were applied and how the result comes out under the other school. If you're on a boundary, it tells you that first.

→ Open the Manseryeok calculator