For high speed photography the exposure time is very important, however it´s not possible to synchronize a flash with the camera shutter for time < 1/160th of second for most modern DSLR cameras. This is due to the shutter which is composed out of two curtains.
The shutter works like this:
One curtain is closed the other is opened. When the shutter is opened, the closed curtain moves out of the way (for example to the top) exposing the sensor to the incoming light. Afterwards the shutter is “closed” by a second curtain which closes from the bottom to the top. This mechanism ensures, that every part of the sensor gets the same exposure time. For really short exposure times aka < 1/160 the second curtain starts closing before the first curtain has fully moved out of the way! So there is no point in time where the whole sensor is exposed at once!
It´s therfore impossible to capture events < 1/160 with the shutter set to < 1/160, because only the a part of the sensor is exposed when the event happens.
To circumvent this problem, it’s possible to have a much longer exposure, so the shutter stays open and only illuminating the scene while the event is happening. This requires the scene to be black or almost black when the event does not happen and the shutter is open. This is of cource only practicable for scenes not emitting light.
As a illumination source typically i photo flash is used, because it generates a short pulse of light. “Short pulse of light” is pretty undefined. To better understand this, we should know how a photo flash works:
Basically it charges a capacitor to a very high voltage and dumps that load in a xenon flash-bulb as soon as it is triggered. To avoid destruction of the bulb the current to the bulb is limited by some means. To measure the “light-flash” i connected a light sensor to an oscilloscope and just plotted the sensors output. You can see the result in the following pictures. The flash has some “intensity setting” the first picture shows the light-flash with the flash set to the minimum setting, the last picture shows the flash set to the maximum setting.
As you can see in the first picture we see a capacitor discharge curve. Most of the energy is dumped within 1/2000th of a second. However in the 4th picture you can see the current limiting circut to the xenon-flash bulb. This makes the flash stay at maximum brightness for a short amount of time before entering the normal capacitor discharge curve. In the last picture this effect is even more obvious.
So the best way to capture a very fast event would be to have a pitch black scene a exposure longer than the event (1 second or so) and a photo-flash set to the minimal setting to get the shortest possible light pulse.