CC Particle World - Origin and 3D Camera tracker strange issue

[ShiveringCactus]

Now here's a puzzle I'm curious about and maybe it's just pushing the limits of the effect too far.

I was playing around with CC Particle World and footage of a woman dancing.  I'd roto-brushed the figure and used CC Particle World's Color Map's Origin Constant setting.  This produces a cute effect, similar to Particle Systems II's but in 3D space.

The footage is from Pexels and the camera is moving, so I added a camera tracker.  This is where it gets odd.  Without the 3D camera, the particles appear on the figure.  With the 3D camera, they are displaced and small in frame.  Using Workbench's Normalize Track script to reduce the tracking distances, the particles emit at the correct size of emitter, but are still not on the figure.

It's not that big a deal, but the overall effect looks better when a floor is added, as it makes the particles look like they are in real space more.

No camera:

With Camera:

With normalized script and camera:

Original footage visible:

Normalised but Origin off ( a similar issue occurs without normalising the tracking data)

(I can't replicate again, but yesterday when trying this, I ended up with mini figures reflecting(?) on the particle layer.)

I think Particular can do this, but as I make tutorials, I try to force myself to not promote pay-for solutions.  My feeling is if you are a watching a tutorial because you need to learn how to do something, then you probably don't have the resources to spend even more money.

I've used CC Particle World with a 3D cameras loads of times, but I'm not sure if I've ever used it with a 3D Tracker Camera.  Missing from my screenshots is that these were in Advanced 3D, but I've also tried with the other 3D renderers.

So that's the head scratcher.  Any ideas?  

(AE 24.6.2 on Windows 11)

[Rick Gerard]

Have you applied Normalize Track to the camera tracker before applying CC Particle World? CC Particle World will expect the camera's origin to point to the comp center instead of 0, 0, 0, as it does when you run Camera Tracker.

Normalize Track - Workbench

[ShiveringCactus]

Thanks for replying.  Unfortunately trying that approach didn't work.  Removing Normalize Track from the workflow, so setting a ground plane and origin, creating a camera then applying CC Particle World results in the reflected mini-figures I mentioned earlier:

Running Normalize track then adding CC Particle World returns the same issue:

You must be right about the camera origin being connected somehow.  I wonder if I can figure out an expression to move one to the other?

[ShiveringCactus]

I'll just add I went right back to the start, did everything, applying CC Particle World last in a new comp and the same result, although when not using the origin setting for color map, the particles are at least in the right 3D space for the tracker camera. 

[Rick Gerard]

I forgot that CC Particle World does not use the comp camera, and I cannot figure out how to change the emitter position, which is based on a percentage of the width with the comp center being zero, and the camera position and orientation, which constantly change, when you use the Camera Tracker.

The math may be possible if you could convert the One-node camera to the Two-node camera, but that doesn't work because the point of interest never changes. 

You'll have to use a particle system that uses the comp camera and then normalize the camera track.

[ShiveringCactus]

It's very confusing, because it does respond to the comp camera normally.  I think there just seems to be this quirk where when using Origin it can't quite cope when there's a 3D tracker camera.  I wonder if "origin" in the tracker is accidentally using the same term that CC Particle World uses.  And I think you might be right, if I can deactivate the tracker camera, but translate its movements to the producer/effect camera, will it work.   That's a challenge.