Stopwatch

What a stopwatch measures

A stopwatch measures an elapsed interval, the gap between two events. Press start at the first event, stop at the second, and the display shows the duration between them. That is a different question from a clock, which answers what time it is now. A kitchen timer counts down, and the countdown timer on this site covers that case. Here the count runs upward from zero.

The underlying unit is the SI second. Since 1967 the second has been defined by the cesium transition: 9,192,631,770 periods of the radiation between two hyperfine levels of cesium 133, as stated on the BIPM page for the SI second. Every stopwatch, from a mechanical windup model to a phone app, is in some sense a counter that tallies oscillator cycles against that definition. NIST describes the same idea in plain terms: we cannot measure time directly, only intervals between events, and a clock is a device that counts those intervals.

Two stopwatches can disagree by a few hundredths on the same event. Each one samples a different oscillator and each one reacts to a button press at a slightly different moment. That difference is not an arithmetic error. It is the gap between resolution, the digits a display can show, and accuracy, how close those digits sit to the true duration.

Reading the display: hours, minutes, seconds, hundredths

The display on this page follows the pattern hh:mm:ss.cc. The first field counts hours, the second minutes, the third whole seconds, and the last field after the period counts hundredths of a second. A reading of 00:01:23.45 means zero hours, 1 minute, 23 seconds and 45 hundredths. That is 60 + 23 + 0.45 = 83.45 seconds in total.

Hundredths are a sensible ceiling for hand timing because 0.01 second is 10 milliseconds, which is finer than human reaction can justify. A typical reaction to a visual signal is on the order of a quarter of a second, or 25 hundredths, so a stopwatch that resolved thousandths would display digits that carry no meaning. Laboratory systems reach microseconds with electronic triggers and calibrated oscillators, not with a thumb on a button.

Hours do not wrap at 24 because the display shows a duration, not a clock face. A timer left running for a day and a half reads 36:00:00.00, and one left for 100 hours reads 100:00:00.00. Leading zeros keep each field at a fixed width so the digits do not shift sideways as the count passes 9 seconds, 59 seconds or 59 minutes.

Each field is a fixed multiple of the one to its right, so converting between them is multiplication by 60 or by 100 and nothing else. The table gathers the values used most often.

Two habits keep conversions safe. First, carry the fractional part in seconds, the unit with the most natural decimals. Second, remember that 0.5 seconds is 50 hundredths, not 5, a slip that shows up whenever someone treats the last field like a decimal digit of minutes.

How the timer on this page works

Start records a timestamp and switches the display into live mode. Stop freezes the elapsed value into a stored base. The next start adds to that base instead of clearing it, so the pause and resume behavior matches what a coach or a cook expects from a stopwatch. Lap records the current elapsed time without stopping the count, and reset clears the base and the lap list.

The live display is painted from timestamps rather than by adding 0.01 to a stored number on every frame. That choice matters when a browser throttles animation in a background tab. The screen may freeze for a while, but the value drawn on return is computed from real elapsed time, so no duration is lost. The same timestamp approach drives the segment arithmetic behind the lap list.

Everything runs locally in your browser. The page does not upload a reading, and closing or reloading the tab clears the timer, so note a reading before you navigate away. The converter below the buttons accepts a formatted reading such as 00:01:23.45, a short form such as 1:23.45, or a plain number of seconds such as 83.45.

Converting a reading to seconds and minutes

The conversion is positional arithmetic with bases of 60 and 100. For 00:01:23.45: one minute is 60 seconds, add 23 seconds to reach 83, then add the 0.45 fraction to get 83.45 seconds. Multiply by 100 to get centiseconds: 83.45 x 100 = 8,345. Divide by 60 to express it in minutes: 83.45 / 60 = 1.3908, which rounds to 1.39 minutes. The widget returns exactly these values: format 00:01:23.45, 83.45 seconds, 8,345 centiseconds, 1.39 minutes.

The reverse direction is repeated division with remainders. Take 3,725.5 seconds. One hour holds 3,600 seconds, leaving 125.5. Two minutes hold 120 of that, leaving 5.5 seconds. The formatted reading is 01:02:05.50. A few common conversions:

The minutes column rounds to two places, which is why 3,599.99 seconds shows as 60.00 minutes. Keep that in mind when a report needs one consistent unit: total seconds avoid the rounding entirely.

Total seconds with two decimals preserve everything the display can hold: 83.45 keeps the hundredths, while 83.5 throws away information that a later comparison might need. Minutes with two decimals go the other way: 83.45 seconds is 1.39 minutes, a value that no longer rounds back to the original reading.

When a report mixes formats, label the unit next to every number. A column that holds 1.39, 62.09 and 150.00 is easy to read as minutes until someone assumes seconds. The time to decimal calculator handles the clock side of that conversion, and the table earlier in this page covers the common stopwatch cases.

Laps, splits and pace

A lap is read as a cumulative time: 00:00:40.00 means the first segment ended 40.00 seconds after the start. The final segment is the difference between the total and the last lap. If the finish reads 00:01:23.45 and the last lap was taken at 00:00:40.00, the segment is 83.45 - 40.00 = 43.45 seconds, which formats as 00:00:43.45. The widget prints both the formatted segment and its value in seconds.

Pace is a division on top of the same numbers. A 5 kilometer run finished in 25:00, which is 1,500 seconds. The pace is 1,500 / 5 = 300 seconds per kilometer, displayed as 00:05:00.00. Enter the same reading with a distance of 3.107 miles instead and the converter returns 482.78 seconds per mile, since 1,500 / 3.107 = 482.78.

Leave the lap or distance field empty and the matching output rows stay empty, so the same widget serves a plain timing job and a full split sheet. A practical rule for split work: record cumulative laps as you go, because reconstructing them from memory after the event is where most timing sheets lose their value.

A short split sheet shows how cumulative laps and segments relate. The total below is 00:01:23.45 with laps taken at 40.00 seconds.

The segment column is read by subtraction only: 83.45 minus 40.00 for the second row, and the first lap equals its own segment. When a segment grows from one lap to the next, the pace fell over that stretch, which is the whole point of recording laps instead of one final number.

Accuracy: what the digits do and do not promise

Resolution is not accuracy. A display that changes every hundredth of a second can still be off by a tenth if the start and stop presses arrive late, and hand timing adds the operator reaction time to both ends of the interval. The effect shrinks as events get longer. An error of 0.2 second on a 10 second sprint is 2 percent of the total. The same 0.2 second on a marathon is under 0.01 percent.

Oscillator quality sets the drift of the counting itself. NIST notes that commercially available cesium clocks keep time to within 1/3,000,000 of a second per year, roughly a third of a microsecond, while consumer devices run on quartz crystals with far looser tolerances that also shift with temperature. Over the minutes a stopwatch is usually held running, that drift is invisible. Over weeks of continuous counting it becomes the dominant error term.

Network time does not fix the problem inside a stopwatch. NIST's explanation of time.gov notes that the running clocks on that site are driven by the local oscillator after the first timestamp and should not be used to measure time intervals or to establish traceability. A stopwatch has the same boundary: it measures a local interval well, and it says nothing about absolute time. For a traceable interval you need a calibrated instrument and a documented procedure, not a browser tab.

Timestamps, drift and background tabs

Browsers offer two clocks. One returns the current wall clock time, which can jump when a device syncs or shifts time zones. The other is a monotonic timestamp that only moves forward, which is what an interval needs. This page uses the monotonic timestamp when the browser provides it and falls back otherwise.

Throttling is the other hazard. When a tab is hidden, many browsers slow or stop animation callbacks to save power. A timer that added 0.01 seconds per animation frame would lose that time entirely. Because the display here is recomputed from timestamps on every paint, a tab that returns after several seconds draws the correct total, including the time it was hidden.

The same logic drives the lap list. Each lap stores a timestamp, and the segments shown are differences, so nothing depends on how often the screen repainted during the run.

Stopwatch, countdown timer or clock time

Pick the tool that matches the question. A stopwatch answers how long something took. A countdown timer answers how long is left. A clock answers when something happens. Mixing them produces the classic errors: subtracting two clock times across midnight, or expecting a countdown to record how long an activity took.

When a result needs to be compared across sessions, record the raw reading in a fixed format rather than a rounded sentence. 00:01:23.45 can be re parsed, sorted and averaged. A note that says about a minute and a half cannot. If a spreadsheet needs a number, the time to decimal calculator converts clock style readings into decimal hours or minutes, and the hours calculator handles spans between clock times.

For interval training, a common setup pairs the two tools: a stopwatch for the total session and a countdown for each work interval. Record the stopwatch total at the end, and the laps along the way, and the session has a complete record with no extra equipment.