Change of direction at moment of impact
Next we tried eliminating the horizontal motion after rebound such that after the impact the dot moved straight up.
We also tried inverting the videos so that they went up and then down instead of down and then up.
Finally, we made a "through the bar" condition where the dot does not rebound at all, going straight through the impact point and mirroring its normal trajectory.
We hypothesized that a rebound would be needed for the effect to occur, the horizontal motion would not be necessary, and inverted videos would behave in the same way as normal videos. This would show that a rebound is sufficient to make the sound "go with" the impact motion and that horizontal motion is not needed, allowing us to focus on the vertical motion.
The results came out as predicted.