Although this project has already been done before (Hobby Robotics), I was thinking of some of the difficulties that may arise from using this device. My first concern is trying to position the laser beam exactly where the drop will be located. The waterdrop position could vary between all the droplets, meaning you could miss many of the drops.
Line Laser Idea
The first solution that came to my mind was to expand the laser beam so that you have a larger region to capture the falling drop. Unfortunately, this idea involved lenses which are generally expensive. Secondly, you would like the laser beam to be thick in one direction but thin in the fall direction. The laser would look like a ribbon or plane of light. This would give the best time resolution of the falling drop. Again, this requires special lenses that could be expensive. For example, you would need a miniature cylindrical lens (or Powell lens if you want to get fancy) as well as two larger spherical singlet lenses. This starts adding significantly to the cost. The basic lens probably costs ($10-20). The cylindrical lens is a little more exotic. It’s possible that a cylinder plastic rod could be used instead of an optical quality lens. So it’s possible, and I may try it at some point, but I’d like to minimize the overall cost of this starter project.
Custom Dropper Idea
My second solution is to make a custom dropper housing with an integrated light trigger. This solution is nice because the light beam can be positioned immediately at the exit of the dropper so you have addressed the problem with light / droplet positioning. As your trigger is near the dropper and far from the surface, the device would allow the user to set a delay between when the droplet is sensed and when the flash is fired. I purchased some infrared sources and sensors that have a thin profile so they might work better in this setup than the laser and photoresistor. The Arduino board would be located close by. If this actually works, I might buy the Arduino nano board to incorporate the electronics into the dropper housing.
My main concern with this idea is how well you can set the delay between the incoming light beam trigger and the outgoing flash trigger. The Arduino reference suggests the analog input samples around 10kHz. This suggests you know the drops position within 100usec as it crosses the light beam. The rise and fall time of the infrared detector is tens of usec. So if we have our dropper 50 cm (~1.5 ft) above the surface, the total time the drop is falling (ignoring drag) is
x = ½ g t2
t = sqrt( 2(0.5 m) / 9.81 m/sec2 ) = 319.275 msec
This is a fairly long delay so there will probably be some experimentation to see how accurately I can delay the flash signal from the drop signal.