Microseconds to Seconds

Setting Up a Shot Where Everything Is Measured in Microseconds

A high-speed rig runs on two numbers that both live in microseconds: how often the sensor takes a picture, and how long each pixel is allowed to collect light. Camera software and trigger boxes usually accept those in microseconds, while the physics you are shooting — a splash, a bullet, an airbag inflating, a blade passing a strain gauge — is described in seconds and metres per second. Getting between the two scales quickly is most of the setup work at the bench.

Conversion factor: 1 µs = 0.000001 s, so divide microseconds by 1 000 000. Running at 10 000 fps the sensor takes a frame every 100 µs, which is 0.0001 s — the same figure a spec sheet writes as a 1/10 000 s cadence.

The Four Numbers That Fight Each Other

Exposure is not the frame interval

The interval is how often a picture starts; the exposure is how much of that interval light is being collected. The rest goes to reading the sensor out and resetting it, so exposure is always the shorter of the two.

Blur is speed multiplied by exposure

A subject travelling 300 m/s smears 1 mm across the frame in 3.33 µs. Decide the smear you can live with first, and the exposure falls out of that division rather than out of taste.

Light is the binding constraint

Dropping from a 1 000 µs exposure to 10 µs throws away 99 % of the photons, about 6.6 stops. That single step is why high-speed work is dominated by lighting rather than by the camera.

Memory decides the clip length

Internal RAM holds a fixed number of frames, not a fixed number of seconds. A buffer good for 20 000 frames gives 20 s at 1 000 fps and 0.2 s at 100 000 fps.

Turning a Shot Plan into Camera Settings

Work from the event you are trying to resolve back towards the two boxes on the acquisition screen.

Type the exposure you calculated

Enter 20 in the left field and 0.00002 s appears on the right, which is the 1/50 000 s a lighting technician will recognise. A comma is accepted in place of the decimal point, so 3,33 and 3.33 both work.

Check it against the frame interval

An exposure only fits if it is comfortably shorter than the interval at your chosen rate. At 25 000 fps the interval is 40 µs, so an 8 µs exposure sits well inside it while a 35 µs one leaves the sensor nothing for readout.

Reverse when the spec arrives in seconds

Press the swap button (↔) for s → µs and a shutter written as 0.000125 s becomes 125 µs, ready to be typed into a camera that refuses anything but microseconds.

Copy the digits into the acquisition form

The copy control on each field hands over the number alone, with no unit and no spaces, which is what a trigger-delay or exposure box expects. Ctrl + C inside a field works the same way.

Judge blur in pixels, not millimetres: 1 mm of smear sounds harmless until you know the field of view. Across a 300 mm field on a 1 280-pixel-wide sensor that millimetre is 4.3 pixels, which is visibly soft. Halve the field of view and the same exposure doubles the damage.

Frame Rate, Frame Interval and a Workable Exposure

The interval is fixed by the frame rate. The exposure column shows a value that leaves the sensor room to read out — a fifth of the interval here — with the shutter fraction a stills photographer would recognise alongside it.

Frame rate	Interval (µs)	Interval (s)	Exposure (µs)	Exposure (s)	As a shutter fraction
1 000 fps	1 000 µs	0.001 s	200 µs	0.0002 s	1/5 000 s
2 000 fps	500 µs	0.0005 s	100 µs	0.0001 s	1/10 000 s
5 000 fps	200 µs	0.0002 s	40 µs	0.00004 s	1/25 000 s
10 000 fps	100 µs	0.0001 s	20 µs	0.00002 s	1/50 000 s
25 000 fps	40 µs	0.00004 s	8 µs	0.000008 s	1/125 000 s
50 000 fps	20 µs	0.00002 s	4 µs	0.000004 s	1/250 000 s
100 000 fps	10 µs	0.00001 s	2 µs	0.000002 s	1/500 000 s
1 000 000 fps	1 µs	0.000001 s	0.5 µs	0.0000005 s	1/2 000 000 s

Note what happens to the resolution as you go down the list: sensors reach their headline rates by reading fewer rows, so the bottom two lines usually mean a letterboxed strip rather than a full frame. The exposure column collapses faster than most lighting rigs can follow, which is why the last row belongs to specialist equipment rather than to a general-purpose bench camera.

What Earns Its Place During Setup

Walk a rate ladder in one pass

Both boxes take input and update each other live, so stepping 500, 200, 100, 40 microseconds down the exposure ladder never means clearing a field first.

Go back the other way for the report

Swapping the direction turns a seconds figure from a test report or a lighting spec straight back into the microseconds the camera wants.

Sub-microsecond values move to exponent form

Anything under a millionth of a second is printed with an exponent instead of a run of leading zeros, so a 0.4 µs exposure reads cleanly rather than as 0.0000004.

Milliseconds are one dropdown away

Pre-trigger windows and total record length are quoted in milliseconds and seconds, and the searchable unit lists on both sides cover the whole range without leaving the page.

Questions From the High-Speed Bench

What exposure freezes a 300 m/s projectile inside 1 mm of blur?

Divide the blur you will accept by the speed: 0.001 m ÷ 300 m/s = 3.33 µs, or 0.00000333 s. Tighten the target to a tenth of a millimetre and you need 0.333 µs, which almost no continuous light source can feed. That is the moment most ballistics work stops using the camera shutter and starts using a flash short enough to act as one.

If a camera runs at 10 000 fps, can I use the whole 100 µs interval as exposure?

Not quite. The sensor has to transfer charge and reset between frames, so the maximum exposure a camera offers is a little short of the interval and the menu will refuse anything longer. More to the point, you rarely want it: at 100 µs a subject moving 10 m/s smears a full millimetre, so the frame rate would be resolving the event in time while the exposure quietly blurred it away.

Why does my recording time collapse when I raise the frame rate?

Because the buffer is counted in frames. A camera holding 20 000 frames records 20 s at 1 000 fps, 2 s at 10 000 fps and 0.2 s at 100 000 fps — the same memory, filled ten times faster each step. Dropping resolution buys frames back, since a smaller region of interest means smaller frames. It is also why a circular buffer with an end-trigger is standard practice: you let it overwrite itself until the event fires, then keep the last fraction of a second.

How much light does a microsecond-scale exposure really demand?

Every halving of the exposure costs a stop, and the jump from a comfortable 1 000 µs to 10 µs is a hundredfold cut — 6.6 stops that have to be paid back in illumination, aperture or sensor gain. In practice that means banks of continuous LED heads or focused fibre-optic light guides rather than ordinary studio lamps, and it explains why high-speed footage is so often shot against a dark background with the subject lit from close range.

How is flash duration different from shutter time?

In a darkened room the flash becomes the shutter: if nothing else is illuminating the subject, the sharpness is set by how long the lamp burns, not by how long the sensor is open. A xenon strobe typically fires for something between 5 and 30 µs, a small speedlight at its lowest power for a few tens of microseconds, and specialist air-gap units for under a microsecond. Manufacturers quote this two ways — t0.5, measured between the half-power points, and the stricter t0.1 — and the t0.1 figure is often two to three times the t0.5, so compare like with like before trusting a number.