Movement of sound wave through air molecules
First time forum user (though lurker for awhile). I am trying to figure out the best way to create this effect of a sound wave moving through air molecules in this NPR video. The effect starts at timecode 1:48. I have also included an image to help you locate the right time.
When closely observed, the dots are oscillating between two positions, with the offset creating the illusion of sound waves moving out. It's really great. I considered creating maybe six nulls with the oscillating movement and parenting the dots to them, then positioning everything and timing everything to create the effect, but thought I would check with this group if there was a faster/better way.
Thanks in advance for any help you can provide.
So first, I'm not sure that depiction is actually correct. Smarter people than me can pile on after this but I'm pretty sure the movement of molecules that sound waves produce in a medium is like ocean waves - the molecules of air or water don't actually translate with the wave, they just move up and down "in" the wave and it's the wave that does the moving, or rather, the compression structure is passed from molecule to molecule. The molecules in the depiction that wander back and forth between crests is probably right (that's the compression squeezing and relaxing or you can think of it as the molecules sliding up and down the trough hills) but to have rings of molecules traveling outwards... I'm not so sure that's right.
There is outward movement of air for sure, but that's not from the sound, its from the sound-producing object displacing the air: whether that's the air that's being pushed out from between the hands or the expanding air from an explosion.
Anyway I just wanted to be sure you were depicting the right thing. And I could be wrong there.
There are sooooo many ways to do this. Let me touch on a few and then you can pick one and if you need more help we can dive down into that method.
The easiest is probably Particular - you can keyframe the particle per sec of the emitter in "disk" mode to turn on and off for a frame with a period (the wavelength) in between those two keyframes- use hold keyframes. Use a very low variability in the velocity and the particles will stick to "rings" while still being not quite all the same. See attached. You can then add wanderers that wiggle around with a wiggle expression.
You could use a wave gradient (either generated by the radiowaves effect or by a cycling and repeating wave gradient) and use that as a displacement map for a bunch of dots (which is probably closer to what actually happens)
If you have Element3D you could use it's cloning features to have dots stick to the high and low parts of a 3D wave and another set of dots for the wanderers be displaced by the 3D wave object.
Or you could set this up in a linear fashion and animate a few molecules moving down the screen the way you want them. If you use equations you can use sin math to get them to cluster properly then build in a random factor that uses the index number as a seed and as the X position offset. Duplicate a ton of these and that will create a field of dots across the screen moving down with crest and trough concentrations. Then apply a polar coordinates (in distort) effect to the whole comp to wrap it all up in a circle. (see attached. I've cheated here by using particular again for the downward flow just so you can see the polar effect in action because I use it again in another example below.)
Another way to get the oscillations is to create a 3D "drum" of molecules (think: oil drum). Rotate that drum axially and view from the curved side of the drum. Place a few of these drums side by side with space between and it will look like what you are after (then do the polar coord trick). And this should be closer to what actually happens in real life. For more accuracy, duplicate each drum in position and have it spin the other way. The drums should have lots of molecules in the middle (along the axis) and very few out near the edge. You can then choose how far apart the pairs of counterrotating drums are from the next pair.
You could use particular for the "drum" shape with a velocity of 0, or you can place them by hand or with expressions. Or you could make a null (but with the 3D box checked) with parented children dots scattered about (concentrated in the center along the y axis) and have all those dots always face camera. Now duplicate that null and it's children and rotate it about its y axis, taking the children with it. Repeat and rotate to get a "fan" of planes around a central axis. There's one drum of dots. duplicate comp and move over as needed for wave crests. If using a 3D camera for this us a long lens so you don't see the arc of the dot's travel.
I just realized and I'm too lazy to retype all this that the in the drums example the drum's dot concentration and rotational axis will have to be in the X for the polar coord effect to work. All else is the same.
Here's a really good graphic of what actually happens (but in linear mode - break out the polar effect!)
Wow, thank you so much. I think you're right...the animation may be technically wrong. When I compare it to the one you posted, the latter seems far more correct. Thank you for taking so much time to answer this. :)