Ernst Chladni figured out if you bow a metal plate with sand on it in just the right way, the sand forms precise mathematical patterns. These are now known as Chladni patterns on Chladni plates. The patterns themselves show areas of high and low vibration.
Where the sand congregates are nodes and the opposite are antinodes. Nodes are caused by the cancellation of the waveform; antinodes are where waves reinforce each other. It’s a visual representation of what we described in the pages on room acoustics. Waveforms in the air bounce between walls and cancel/reinforce each other at various points in the room, with every level of change in between.
There are many ways to create a Chladni plate. The modern version uses a speaker and an oscillator instead of a bow, but the simple bow version is much easier and cheaper, assuming you have a spare bow lying around. Instructions are readily available online (we’ll add ours here once we make it; see below for what we did do).
What we did was create a Chladni-esque plate that’s more malleable. Here’s what our experiment looked like:
While we did add sand and got some patterns similar to a round Chladni plate, what you’re looking at is an analog audio visualizer. A laser is aimed at a mirror, which is glued to a tympanic membrane (in this case, a garbage bag stretched over a bucket). The bucket has no bottom and is placed over a subwoofer. The end result is a reflected light pattern that represents the vibrations on one spot of the membrane. This pattern is known as a Lissajous curve. If you want to go into the technical aspects of such curves, definitely go to this link (thanks to Nick Day for pointing out what they’re called).
The experiment can be done with a speaker of any size, but lower frequencies = more movement (resulting in a larger pattern). We used an 8″, 100-watt, JBL subwoofer and a fairly powerful but inexpensive purple laser. The bucket was a 3.5-gal foodservice pail (found free at many stores and restaurants). The mirror was a shard of CD. We purposefully chose a garbage bag that was thinner to increase the transient response.
Any laser works for this project, though I’d recommend one that takes AAA or AA batteries since it’ll be sitting for months between uses. The one we used can be found on eBay for less than $5 (<$2 if you wait for it to arrive from China), though I’m sure there are quality control issues. If I were to recommend a speaker, I suggest one like this Yamaha subwoofer (link goes to Amazon). It’s the closest to what we used. The diameter should be great for fitting a normal bucket over (remove the grille/fabric of course; it should be built to pop on and off easily).
The audio was played through the subwoofer and a nice, 2.1 computer speaker set. The visualizations are only from the audio going through the subwoofer, which only reaches up to about 150 Hz. A lot can be learned about how a song was mixed and the nature of high vs. low frequencies from observing these visualizations. For example, an 8kHz tone won’t move the laser beam much, since it’s vibrating the plastic at 8,000 times per second. Related to that fact is the high amount of energy required to reproduce low frequencies—it takes more power to move the speaker diaphragm farther. Also, if a song is mixed well, you’ll hear crisp lows and see clear patterns; alternating with these will be tight patterns and a single dot when the bass drops out.
Playing with a simple online oscillator is especially enlightening. Pull up a couple in separate computer tabs. Start one and you’ll likely see an oval. Start the second one at the same frequency or a perfect harmonic and you’ll see the same oval. Change the second oscillator and you’ll see what it looks like when waves combine. It gets complex real quick. Have fun with it!
You can also compare what you see on the wall to what registers in an oscilloscope (like the one at this link). Once you let it hear your computer’s microphone, it’ll show the waveform of whatever it hears. Raise the gain if it’s too small. Raise the “Seconds / div” if you want to see a longer section of the waveform at a time. This type of tool is an amazing addition to a discussion of frequency—just whistle from low to high and see how much closer together each peak is as you go up.
We first did the experiment with a portable speaker in a medium-size cup, a balloon and a dollar store laser. The results were good enough for a proof of concept but were definitely underwhelming. If the speaker had been Bluetooth it would have been a little better, but it was still very little movement (mostly due to the lack of bass response in the speaker). Before we could set up the camera, the dollar laser ran out of juice. You get what you pay for.