The 'Ice Balloon' setup consisted of popping water balloons over an aquarium.  A tall plastic beaker was placed upside down with another beaker on top, where the liquid nitrogen was contained.  The flash unit was attached to the side of the aquarium.

We set up a splash box, bowl, cylinder, and bottle cap. The bottle cap rested on top of the cylinder, which stood at the center of the bowl, which rested directly beneath the photo gate so that water would drip into the bottle cap. The flash unit was placed directly behind the bowl, facing slightly downward, and connected to the time control box, which was connected to the photo gate. The camera was placed on a tripod and adjusted so that it faced down, inside the bottle cap and then focused on just in front of the bottle cap, where the water would splash up. The tongs were placed next to the container of liquid nitrogen, which was placed next to the splash box to lessen the time between filling the cap and replacing it.

At the start of the experiment we planned to burst water balloons over liquid nitrogen.  We did this by setting up a sound trigger, filling balloons half with water and half with air, and then popping them over a beaker.  Although this resulted in interesting pictures, our desired result was to catch the water freezing in mid-air- which proved to be impossible under the conditions, as the water only froze after being suspended in liquid nitrogen for several seconds.  After realizing this we switched to dripping water into liquid nitrogen.

• Final Method

For this experiment, we planned to take photographs of the splash that resulted from dropping water into liquid nitrogen. We did this by setting up the camera and focusing on the bottle cap, then filling the pipette with water that was colored red with food coloring on a splash box so that the water would drip through a photo gate, causing the flash unit to flash. We then dripped the water into the bottle cap, photographing the splashes. Between taking these test photos, we adjusted the delay unit and flow of the water to get the desired timing.

After we settled on timing and settings that looked to work best, we filled the bottle cap with liquid nitrogen and used tongs to place it back on the cylinder. We had left the water dripping while doing this, and as soon as the cap was in place, started taking pictures by turning the shutter speed on manual and holding the button down until we saw the flash, taking as many pictures as possible before the nitrogen had completely boiled. After the nitrogen had boiled, we reviewed the photographs and adjusted the delay unit and drip rate according to what we felt needed to be changed before refilling the cap with nitrogen and repeating the process.  Eventually, we switched to a watch glass to provide the best photograph.

The first attempt involved a sound trigger position near the water balloon.  We had to fill the water balloons with air to create a loud enough noise so that the sound trigger could signal the flash.

The flash unit was triggered by water dripping through a photo gate. The length between the drop passing through the photo gate and the flash triggering was delayed using a Time Control Box that was easily adjustable, and the actual delay time was changed many times during the experiment according to how the last set of images had turned out and where in the splash we wanted to try and get a picture.

Because the water actually fell in a very random pattern, we had to use a manual shutter speed to be sure to catch the drops. We turned the lights off and then held the shutter open until the light flashed, released it, and pressed it down until the light flashed again.

We found that the f-stop was set to 22 for most of the images to increase the depth of field and also because many of them became over expoused, and this let in as little light as possible without changing ISO from 400.