I’d been wanting to try this experiment for quite some time, after seeing it in a YouTube video several years ago.
The explanation of the phenomenon is pretty simple:
- The camera is recording an image 24 times per second (23.976 to be exact).
- There is a speaker which vibrates the water at the same frequency, causing it to drip in a near-perfectly repeating pattern.
- The camera images the stream of water in exactly the same position every time, thus it appears not to move.
When the audio frequency is changed to be slightly faster than the camera’s sample rate, the drops appear to fall very slowly. Likewise, when it is made slightly slower, the drops appear to move slowly in reverse.