I started to experiment with flash synchronized high-speed photography some time ago and the most obvious starting point was of course to capture falling water drops ( https://pbase.com/jakobe/drops ). The principle is straight-forward, where the shutter of the camera is opened where the subject is in total darkness and the illumination is done by a standard flash unit – in my case a Canon Speedlite 550EX.

Interestingly enough, the most critical part of the setup is the flash and not the camera itself (although a good camera typically doesn’t hurt).The key is a short flash burst, yet powerful. I’ve been searching the net for some accurate timing for different flash units, and I don’t know how Canon's flash guns compare to others in this respect. As there is no link between the camera and the flashgun, any flash unit would do.

What is important is the duration of the flash burst, which in turn depends of the output power of the unit. Most flash units seems to change the rated output power by simply changing the duration of the flash burst, which means that the key is to set it as short as possible, yet giving enough light to illuminate the subject.

I have two Elinchrom studio strobes which I hoped should work, but the duration of the flash is simply too long to make this kind of high speed photography meaningful. The peak of the burst is somewhere in the 500 microsecond range (i.e.about 1/2000 sec), even at the lowest output power. The 550EX is better in this respect (well, the output power is by all means much lower) with around 100 microseconds peak time at 1/128 output power. The actual pulse time is longer, as there is some rise- and fall times involved. Some 200 microseconds is a more practical value to base calculations upon.

I decided to borrow a CO2 pistol from a friend to make some practical experiments. Compared with a traditional air pistol which needs to be removed and loaded by hand force, this was simply a better option as the gun could be kept in the fixture between shots. A eight-bullet cartridge was of great help.

The gun I borrowed has a rated muzzle velocity of some 120 m/s, which means that the bullet would travel some 12 millimetres in 100 microseconds. Giving that the bullet is a 4.5mm calibre one, this means that it would travel almost 3x its own length during the peak duration of the flash burst. Again, as the flash burst is a bit smeared out, some serious motion blur must be considered.

Tethered operation surely helps to quickly evaluate exposure and find out appropriate timing settings. I assume it really goes without saying that proper measures must be taken to handle glass debris that scatters around.

Well – I’ll try to describe how it was all setup and what I’ve found out during these experiments...