The model: the mathematics of compressed time

Every timelapse is a pact with time: the camera shoots at regular intervals, the video plays the shots back at frame speed — and the world runs. This calculator puts the pact into numbers, from the four that count: the interval between shots, the duration of the event, the frames per second of the final video and the weight per shot. Out come how many shots you need, how long the clip will run, how many gigabytes the card will be asked for — and the number that governs everything: the time compression factor.

The count is universal and free of national units: seconds, minutes, frames and megabytes — the same at every latitude.

The four knobs: interval, duration, frames and weight

The interval goes from 1 to 60 seconds: from flowing traffic to a cloud being born. The duration goes from 10 to 600 minutes: from a quick sunset to a full day on site. The frames per second are the real video standards: 24, 25, 30 or 60. The weight per shot goes from 5 to 50 megabytes: from compact JPEG to light RAW — the same honest photographic-weight convention used across the site. Four numbers, and the session is designed before the tripod is up.

The interval scale: from the river to the glacier

Two hours of event, video at 30 frames, 15 megabytes per shot; only the interval changes:

The middle row is the factory recipe: a five-second interval turns two hours into a 48-second clip with fourteen hundred shots. And the scale shows the fundamental law: doubling the interval halves everything — shots, clip and gigabytes — while the factor doubles. At the one-second interval the card pays 108 gigabytes for four minutes of video: the river is spectacular, but it drinks the card.

The compression factor: the doubly immobile number

The third result number has two rare properties. First: it is always an integer — the factor is the pure product of interval and frames, and it knows no decimals. Second: it depends neither on the duration nor on the weight — five seconds times 30 frames make ×150 for the ten-minute sunset and for the full day on site alike. It is the timelapse's separation of powers: interval and frames decide how fast time runs; duration only decides how far it goes. Whoever plans starts from the factor, not from the shots.

The three identities: the event that returns exact

The model rests on three identities verified at every version by the test bench. The first is the most elegant: factor times clip length equals the event duration in seconds — the ×150 times the 48 seconds give back exactly the 7,200 seconds of the two hours, to the second. The second: clip times frames equals shots. The third: shots times weight equals space, to the megabyte. Here compressed time loses no pieces along the way.

The four readings: from the near real to the glacial breath

The verdict reads the factor. Below sixty it is the near real time: the territory of living scenes, where a minute fits into a second. Between sixty and three hundred it is the textbook timelapse: sunsets, clouds, construction sites — the band of the manuals. Between three hundred and twelve hundred it is the day in a flash: shadows wheel, markets rise and fall. Beyond twelve hundred it is the glacial breath: tides, ice, seasons — where every second of clip is worth twenty minutes of world.

The weeping card: the five-thousand threshold

Above the readings, the model marks the shutter threshold: five thousand shots in one session — which the grid touches exactly at 500 minutes with a 6-second interval. From there up the session works as much as a month of normal photography: fast, roomy cards, external power, and a thought for the mechanical shutter, which has a shot count written into its destiny. It is not a ban: it is a contract — whoever wants the river pays in shots, and this count presents the invoice before the session.

The right interval: the rule of the subject

The real question of every timelapse is a single one: how fast does the subject run? Marching clouds live between 3 and 10 seconds; the sunset between 2 and 5; the construction site and markets between 10 and 30; the stars ask for 20-40 seconds with long exposures; the growing plant wants minutes, not seconds. The practical rule descends from the factor: choose how fast you want the world to run in the video, and divide by the frames — the table above does the rest.

The limits of the model: what the count does not cover

The perimeter is declared: theoretical shots from the interval, weight per shot as declared. The battery giving up mid-event is not in the count; the camera buffer slowing down below 2-second intervals neither; safety takes, bracketing and frames discarded in editing are a separate chapter. And the real weight sways with the scene: a uniform sky compresses better than a crowded market. The number that comes out is the base of planning: the right order of magnitude, before the card fills mid-event.

The duration scale: the factor still, the card growing

The classic 5-second interval, 30 frames, 15 megabytes; only the event length changes:

The column that is not there is the most eloquent: the factor stays ×150 across the whole scale — duration never touches the speed of time, only how far time goes. Everything else is honestly linear: six times the event, six times the shots, the clip and the gigabytes. The full ten-hour day pays 108 gigabytes and four minutes of clip: duration is the knob of quantity, never of the quality of motion.

The frames scale: the same matter, another breath

The same 1,440 shots of the factory recipe; only the video standard changes:

Shots and gigabytes do not move: the raw matter is the same, only the breath with which the video consumes it changes. From cinema to fluidity the clip halves and the factor doubles — with everything that was photographed staying equal. It is the timelapse's free knob: it turns in editing, after the session, and it allows second thoughts: the shots are forever, the frames are a choice.

The weight scale: the column that touches only the card

The fourth knob is the quietest: the weight per shot moves a single row of the result. The recipe's 1,440 shots are worth 7.20 gigabytes at the compact 5-megabyte JPEG, 21.60 at the full-quality 15, 72.00 at the light 50-megabyte RAW. Shots, clip and factor know nothing of the format: the weight only decides what the very same clip costs the card. It is the twin of the freezer's litres knob: the format does not change the film — it changes the invoice.

The extremes of the grid: from ten shots to thirty-six thousand

The model's minimum is the miniature: ten minutes of event at the one-minute interval — 10 shots, barely half a second of clip at 24 frames, a ×1440 factor and 0.05 gigabytes: the glacial breath in stamp format. The maximum is the long river: ten hours at the one-second interval, 60 frames, light RAW — 36,000 shots, ten minutes of clip and 1,800 gigabytes: almost two terabytes for a single session. Between the two extremes runs a factor of 3,600 on shots and 36,000 on space: the whole grid fits inside four knobs — and that is why the count comes before the tripod.

The duel with sped-up video: why not just film

The question always comes: why not record video and speed it up in editing? The answer is the card: ten hours of 4K weigh 300-500 gigabytes, all to be written and read, with the camera on and hot for the whole duration. The timelapse is compression at the source: it photographs only the frames the final video will actually use — the day's 7,200 shots against the million frames of continuous video. Sped-up video wins only below factor ×30, where full fluidity is needed: from there up, the timelapse is mathematics on the card's side.

The session strategy: five moves before the tripod

The checklist descends entirely from the count. First: decide the clip — 15-60 seconds for the web. Second: choose the factor from the subject, with the readings rule above. Third: derive the interval — factor divided by frames — and check shots and gigabytes with the calculator. Fourth: a five-minute test — sixty shots are enough to unmask exposure and framing. Fifth: plan the power before the space — because a full card gets swapped, a dead battery stops time. The successful timelapse is the one designed standing still.

Frequently asked questions

How many shots does a timelapse need?

The event duration in seconds divided by the interval: two hours at 5 seconds are worth 1,440 shots, at 1 second 7,200. The calculator does the exact count and adds what the shots do not say: the clip length, the compression factor and the gigabytes.

What interval do I use for clouds?

Between 3 and 10 seconds: at 3 the clouds run fluid, at 10 they fly. The sky changes character with the wind — with strong wind shorten the interval, in a calm lengthen it. The true rule is the factor: convincing clouds live between ×100 and ×300.

What interval do I use for a sunset?

Between 2 and 5 seconds: the sunset is fast — half an hour of good light — and deserves fluidity. At a 2-second interval, thirty minutes are worth 900 shots and 30 seconds of clip at 30 frames: the textbook classic, with the light gliding without harsh jumps.

How long is the final clip?

Shots divided by frames per second: 1,440 shots at 30 fps make 48 seconds. It is the number to decide BEFORE the session: a publishable timelapse lives well between 15 and 60 seconds — and from there you climb back to the right interval with the calculator.

What is the compression factor?

How many times the video runs faster than the world: interval times frames, always an integer. Five seconds at 30 fps make ×150 — an hour in 24 seconds. It is the model's doubly immobile number: it knows nothing of the event duration or the file weight.

Better 24, 25, 30 or 60 frames?

24 and 25 for the cinematic look, 30 for the web, 60 for absolute fluidity or for slowing down in editing. With equal shots, more frames shorten the clip: the same 1,440 shots last 60 seconds at 24 fps and 24 at 60 fps. The factor rises accordingly.

How much space does a timelapse take?

Shots times weight: the classic 1,440-shot session at 15 megabytes asks for 21.6 gigabytes. In RAW the weight per shot doubles or triples — and the 64-gig card stops being enough. Do the count first: the card filling mid-event is the most expensive mistake in timelapse.

JPEG or RAW for timelapse?

RAW if the light changes a lot (sunsets, night): the recovery in post is worth the gigabytes. JPEG for construction sites and very long sessions, where space and battery win. The honest middle way: compressed RAW or top-quality JPEG — the calculator's 15-25 megabytes per shot.

Does the battery survive a long session?

It is the true limit, outside the count's perimeter: an internal battery is worth 1-3 hours of interval shooting, and cold halves it. Above two hours you need external power or a USB power bank. The professionals' rule: plan the power before the space.

Does the shutter wear out with timelapses?

The mechanical one does: it has a life cycle of 100-400 thousand shots, and a five-thousand-shot session consumes one to five per thousand of it. The modern solution is the electronic shutter: silent, wear-free, perfect for intervals — where available, it is the default choice.

Can I shoot timelapses with a phone?

Yes, and the mathematics is identical: phones pick the interval by themselves in automatic modes, but manual apps expose interval and resolution — and this page's count applies again. The phone's limit is not the software: it is battery and overheating in long sessions.

Why are the shots theoretical?

The model declares the convention: duration times sixty divided by the interval, without rounding down. In practice the camera may lose a shot to the buffer or add one at start-up: single-unit noise, irrelevant to the project. Planning lives on the exact order of magnitude, not on the single shot.

Updated July 2026