Time Conversion Explained

Time conversion is trivial — until you reach "month" and "year"

Converting seconds to minutes to hours to days is the easiest unit conversion there is: every step up is a clean multiplication by 60, 60, and 24. Nobody argues about how many seconds are in an hour. The moment you climb past a day, though, the ground shifts. A "month" is anywhere from 28 to 31 days. A "year" is 365 days, except in a leap year when it is 366, and the long-run average is neither. Ask two calculators how many seconds are in a year and you can get three different answers, all defensible. The time converter on Calc Dragon settles this by fixing an explicit convention for every unit and telling you which one it used, so the number it returns is reproducible rather than a guess.

This article walks through the one operation every time conversion is built on, the units that are exact and the ones that are conventions, the difference between a Julian, common, Gregorian, tropical, and sidereal year, why leap seconds do not change any of the arithmetic, and the one situation where a time converter is the wrong tool entirely — measuring the gap between two real calendar dates.

The math: everything bridges through the second

Every conversion in the time converter passes through a single intermediate unit — the second. Each unit stores one number, its "seconds per unit" factor, and the conversion is two operations:

result = value × (seconds per source unit) ÷ (seconds per target unit)

So 2 hours in minutes is 2 × 3600 ÷ 60 = 120 minutes. The second-bridge means the converter never needs a separate factor for every pair of units — only one factor per unit. Adding a new unit is adding one number to the table, not a whole row and column. This is the same pattern the SI brochure and every serious units library use, and it costs one extra multiplication in exchange for being far easier to keep correct. The arithmetic runs in full IEEE-754 double precision; only the displayed answer is rounded.

The factors below the level of a day are exact by definition. A minute is 60 seconds, an hour is 3,600 seconds, a day is 86,400 seconds, a week is 604,800 seconds, a fortnight is 1,209,600 seconds. None of these depends on the calendar or the Earth's motion — they are pure definitions. It is only at "month" and "year" that a choice has to be made, because those units genuinely do not have a single fixed length.

Worked example: 90 minutes across the unit table

Take a 90-minute film and express its running time in every unit. The time converter gives:

• In seconds: 90 × 60 = 5,400 s. Everything else is derived from this bridge value.
• In hours: 5,400 ÷ 3,600 = 1.5 h. The obvious one, and a useful sanity check that the rest are right.
• In days: 5,400 ÷ 86,400 = 0.0625 d — exactly one sixteenth of a day.
• In weeks: 5,400 ÷ 604,800 ≈ 0.008929 wk.
• In months (mean Julian): 5,400 ÷ 2,629,800 ≈ 0.002053 mo, using the 30.4375-day mean month.
• In milliseconds: 5,400 × 1,000 = 5,400,000 ms. Sub- second units multiply rather than divide.

Reverse conversions are symmetric. One Julian year in days is 31,557,600 ÷ 86,400 = 365.25 days exactly; one billion seconds is 1,000,000,000 ÷ 31,557,600 ≈ 31.69 Julian years (a "gigasecond" birthday lands just before your 32nd). Because the bridge is the second, the source and target units can be chosen completely independently and any pair works.

The exact units: seconds to fortnights

Everything from the nanosecond up to the fortnight has a length that is true by definition and never varies:

• Second (s): the SI base unit of time, defined since 1967 by the caesium-133 atom — 9,192,631,770 oscillations of the radiation from a specific hyperfine transition. Every other time unit is ultimately expressed in these seconds.
• Minute (min): 60 s, exactly.
• Hour (h): 3,600 s (60 minutes), exactly.
• Day (d): 86,400 s. This is the "mean solar day" the SI treats as exact for unit purposes; the Earth's actual rotation wobbles by milliseconds, which is what leap seconds later mop up.
• Week (wk): 604,800 s (7 days), exactly.
• Fortnight: 1,209,600 s (14 days), exactly. The word survives mostly in British English — and in the physics-classroom joke unit "furlongs per fortnight" — but it is entirely unambiguous.

Because each of these is exact, converting between any two of them returns an exact answer with no approximation whatsoever. It is only when a month or a year enters the calculation that the result becomes convention-dependent.

The awkward units: how long is a month, really?

A calendar month is 28, 29, 30, or 31 days depending on which month and which year. To convert a month to seconds, the converter needs a single representative length, so it uses the mean Julian month: 365.25 ÷ 12 = 30.4375 days = 2,629,800 seconds. This matches calculator.net, Wolfram Alpha, and most reference tools. It is the right figure for questions like "roughly how many seconds in six months" but the wrong figure for "exactly how long from 12 March to 12 May", which is a calendar problem, not a unit problem — for that, use a date-difference tool (see the last section).

The year is where conventions really diverge. The time converter exposes several because different fields genuinely use different ones:

• Julian year (default): 365.25 days = 31,557,600 s. This is the ISO 80000-3, IUPAC, and IAU convention, given the symbol "a" (for annus). It averages a leap year in every four, and because it is a round definition it is exact. Astronomy uses it for the light-year, and it is the sensible default for any "seconds in a year" calculation.
• Common year: 365 days = 31,536,000 s. A single non-leap calendar year. This is the "how many seconds until this time next year" figure when next year is not a leap year — and the source of the famous "there are 31.5 million seconds in a year" factoid, and of the musical "525,600 minutes".
• Gregorian mean year: 365.2425 days = 31,556,952 s. The true long-run average of the calendar we actually use, which drops three leap days every 400 years (the years 1700, 1800, and 1900 were not leap years; 2000 was). This is the most accurate figure for multi-decade spans.

Two more year lengths turn up in astronomy and are worth knowing even though the calendar ignores them. The tropical year — the time from one spring equinox to the next, which is what actually governs the seasons — is about 365.2422 days. The Gregorian calendar's 365.2425-day average was engineered to track it to within about half a minute a year. The sidereal year — one full orbit measured against the fixed stars — is longer, about 365.2564 days. The roughly 20-minute gap between the tropical and sidereal year is caused by the precession of the equinoxes, the slow westward wobble of the Earth's axis. None of these is in the dropdown as a calendar unit, but they explain why "a year" has so many almost-equal values.

Leap seconds: why the converter is right to ignore them

A reasonable worry: if the Earth's rotation is irregular and clocks get a "leap second" added now and then, are these conversions actually correct? The answer is yes, because a leap second does not change the length of the unit "second". A second is always 9,192,631,770 caesium oscillations. Leap seconds are occasionally inserted into Coordinated Universal Time (UTC) to keep atomic clocks aligned with the Earth's gradually slowing spin — 27 have been added since 1972 — but they adjust the labelling of time, not the duration of the units. A day in the converter is 86,400 seconds; a day on the atomic clock during a leap-second event is 86,401 seconds for that one day only. The total drift across all of recorded history is a few dozen seconds, negligible against any span you would use a converter for.

This is about to become a non-issue anyway. In November 2022 the General Conference on Weights and Measures (CGPM) voted to abolish the leap second, with insertions to stop by 2035, after which UTC will be allowed to drift from astronomical time by more than a second and run as a continuous, discontinuity-free count. For unit conversion the practical rule is unchanged: treat the timeline as an idealised, leap-second-free sequence of exact SI seconds, which is exactly what the time converter does.

Sub-second units and the SI prefixes

Going the other way, below the second the converter follows the ISO 80000-3 SI prefixes: a millisecond is 10⁻³ s, a microsecond is 10⁻⁶ s, and a nanosecond is 10⁻⁹ s. These matter in fields where a second is an eternity. A single frame of 60 fps video is about 16.7 ms; human reaction time is 200–250 ms; a DDR5 memory cycle is on the order of nanoseconds; light travels almost exactly one foot in one nanosecond, a coincidence Grace Hopper made famous with a length of wire. Converting these is the same second-bridge multiplication — 250 ms is 250 × 10⁻³ = 0.25 s — but the direction flips: you multiply the source factor and divide by a tiny target factor, so the numbers balloon. Frequency is the reciprocal of a period, so if you are working the other way from a rate, the frequency converter is the matching tool: 60 Hz is a period of 1 ÷ 60 ≈ 16.67 ms.

Common mistakes

Assuming "a month" means 30 days

The 30-day month is a convenient fiction that drifts fast. The real average is 30.4375 days, so a "30-day month" model loses about five days over a year (12 × 0.4375 ≈ 5.25 days). Payroll, interest, and subscription systems that bill in fixed 30-day cycles quietly desync from the calendar for exactly this reason — twelve 30-day months are 360 days, not 365. If the answer has to line up with real calendar dates, do not convert through a "month" unit at all.

Using 365 days when you mean the long-run average

For a single year that is not a leap year, 365 days is correct. For anything spanning multiple years — an age in seconds, a countdown over several years, an interest projection — using a flat 365 accumulates a quarter-day of error per year, roughly a full day every four years. The Julian (365.25) or Gregorian (365.2425) year absorbs the leap days and keeps long spans honest.

Mixing "seconds in a year" figures between tools

One calculator says 31,536,000 seconds in a year, another says 31,557,600. Neither is wrong — the first is a common year, the second a Julian year. Problems only appear when a figure computed with one convention is fed into a step that assumes the other. Pick a convention for a workflow and stay with it, and prefer the tool that tells you which one it used.

Expecting the converter to know about calendars

A unit converter treats every month and year as the same fixed length. It has no idea that February is short or that 2024 was a leap year. That is a feature, not a bug — but it means the converter cannot answer "how many days until my birthday" or "how many working days in Q3". Those are calendar questions with a different set of tools.

When a time converter is the wrong tool

The single most common misuse is trying to measure the gap between two real dates by converting a "months" or "years" value. Because the converter uses an averaged month and year, it will be off by a day or more against the actual calendar, which has variable-length months and leap years. For "exactly how long from 12 March 2024 to 12 May 2026", the right tool is the time duration calculator or the date calculator, which count real calendar days. For counting only weekdays — project deadlines, delivery estimates — the workday calculator excludes weekends, and the day counter tallies the raw days between two dates. For an age expressed in different units, the age in days calculator works from an actual date of birth rather than an averaged year.

Use the time converter when you have a duration and want it in another unit — "how many hours is 5,400 seconds", "how many seconds in a year", "how many minutes in a fortnight". Use a date-based calculator when you have two points in time and want the span between them. The distinction is the whole game: durations are unit conversions, and dates are calendar arithmetic. Get that right and the rest is one multiplication with an honest, clearly-labelled constant. For the neighbouring unit families, the speed converter and distance converter follow the exact same second- and metre-bridge pattern, and the guide to the time duration calculator goes deeper on the calendar side.

Frequently asked questions

See the FAQ on the time converter page for direct answers on how many seconds are in an hour, why a year defaults to 365.25 days, how long a "month" is, what a fortnight is, and whether leap seconds are factored in. The combined calculator and FAQ cover both the quick reference and the deeper convention questions. For related conversions, the frequency converter handles Hz, kHz, and RPM as the reciprocal of a period; the speed converter covers mph, km/h, and knots; and the countdown calculator counts down to a future date rather than converting a fixed duration.