[Steve Bentley]

Depends on your version of AE. The new text tools allow you to tie the text value to something that can change, like time or frame number or even a text doc outside the program or just simple math.
For these we usually do a bunch of lines of code so they fit within the area the camera can see, precomp that. Then use the offset filter (under distort) to shift the whole thing up or down and wrap the lines of text back on the incoming side. This way you have an endless set of lines. As long as they keep changing you never know there were just 10 lines or so.

[Steve Bentley]

Wanna upload the AE project? (I don’t need the footage) and I’ll see if there is something obvious that’s not working right. Depending on the version of AE and C4D the export and import can work a little different. For some stuff I even have to import files into AE5.5 and then convert it to 2019. I currently have 3 different exporters for C4D inside of AE that I use depending on the case.
See if there’s a version of C4D under help ( i know its lite but I don’t want to send back a file thats too new for your version. )

[Steve Bentley]

Thats an hour and half video – any clues as to the time stamp on the frames you are interested in?

[Steve Bentley]

hard to see which null is the tracker null but it looks like it might be 2D only? If that’s the case thats why they aren’t showing up. You can only export 3D nulls and solids to cinema.
Could that be it?? We can make that null 3D and not loose the data which would fit it. But it still is not moving in depth, just in x and y.
Just thought I’d blast that off quick and it might help – going to watch the rest of the vid now

[Steve Bentley]

Try also setting the caps lock key to on before you open the file- this will keep it from trying to render.

[Steve Bentley]

Once you have everything working in C4D you can either import the C4D file as a layer into AE and have AE render the whole thing for you or render out the C4D file (through the imported camera from AE) and bring those frames in as a layer into AE.

You can also bring the C4D project (via the render settings Compositing Project File under Save) into AE as you work to check progress and how things are lining up inside AE You must put an external compositing tag on any object in C4D that you want to export and these will show up as nulls in AE (assuming you just want to check animation and position and not actually render) only cameras and lights will come in as actual 3D objects. You can color each exported object in C4D with the external compositing tag so you know what object is what in AE. With this you are just checking the anchor points and how they move, no objects actually get exported.
The “project” you open in AE from this process shows up as a comp and you have to dive into the comp, pull the items out of the comp and place them in your master comp so everything lives together in the same “space” You won’t need the camera or lights.

[Steve Bentley]

There’s now an export function in AE. (file/export/C4D (different versions have the list menu item listed differently).The objects have to be 3D solids or nulls or lights or cameras. Make sure to label them and having different colored solids helps you see them in C4D too.
That creates a project that you can open in C4D and the exported objects will dance around in time inside C4D.
One thing of note. For things like expressions you sometimes have to collapse the expression into keyframes to get the export to give out the animation info to C4D.
Make sure to use the camera from AE in C4D too so you are looking through the same lens. And the AE comp size has to match render size in C4D.

[Steve Bentley]

I think it would be easier to generate trace splines from a particle emitted from node 76 (or at least node 76’s null).
The traces can be associated with sweeps or lofts so that they in turn generate geometry. You can also use plugins to render just the splines themselves. (white lights is one of my favorites). The traces grow (as would the sweep) as the particles move about.
I’m not sure about the simultaneity. The issue (as I see it) is going to be distance. Each node will be a different distance away from the others so the distance will have to be traversed at different speeds for them all to arrive at receivers all over the place. You could measure the distance between the pair of nodes that are picked in the random sort and then do the math for how much velocity should be imparted so each particle arrives at the given time, the farther away the more initial oomph needed to send the trace on its way. I think you can achieve all of this with expresso.
Or you could use a really big velocity everywhere and then have the sweeps grow over time along all the traces (the traces would span the distances almost instantly).

But there’s a problem here. You want randomness, but only the first volly is random. After that its always the receiver that sends out a new beam to a new random node, but can it be one of the ones that has either already received or already fired?
Instead of using the random function in the link list, you might have to make an array the number of objects available and when picking a new “random” value make sure its not one that’s been used before. So the array would get bigger or smaller each iteration either adding to the list of ones not to use or removing ones that have been used. Then use that array to pick choose the index of the node in the link list to fire next.
Keep in mind there is no “memory” in expresso for whats come before so either have to calculate the whole thing for every frame right back to the beginning of time, or make a change for each node that is tested on each frame. So if node 1 has been used, it gets slightly bigger lets say. When you choose a new value to fire you check to see if its already gotten bigger, and if so pick the next index in the array and test that one (and so one). Coffee would work well for this as would python. Having the array grow or shrink might look after this.

[Steve Bentley]

Are we talking neurons firing? That kind of thing?
You could use the link list in expresso, populate the list with all sender nodes and randomly select which node fires a connection and which node (a second link list of receiver nodes) receives the connection. You could fire a particle along the vector between the two and use the trace effect to make a spline out of that particle’s travel. From there you can sweep that trace path to get real geometry that can be textured.
You can use the collision test to determine when the particle reaches the receiver node and that can trigger a specific texture.
Does that help? Do you have a file to share so we have visual aids?

[Steve Bentley]

Is it as simple as mapping a movie or image of a grid and projecting that onto the cheek and moving the texture over time? Or does the “scan” effect have to be dimensional and extend off the cheek surface?