[Steve Bentley]

To create a better link between C4D and AE so that objects can track or influence objects in the other program, we are constantly passing nulls back and forth and then using those objects (since they are in 3D in both programs) to send positional and rotational data to other objects, or parent them in the time line or the object manager.

Personally (forgive me Andrew) I feel handcuffed by the E3D interface after using 3D apps for the last (gulp) 40 years, So I prefer the flexibility of C4D for getting my 3D just right.

That being said, Adobe has added 3D capabilities to Pshop so its possible its not far behind for AE. I realize they could make the buttons even smaller in AE next time as an upgrade “feature”, but you’re right, adding 3D would be a much better upgrade enticement.
(but then, I’m still waiting for a robust paint in AE – if only adobe owned a paint program they could steal from!)

As for adding obj’s to AE, that’s just the tip of the iceberg. They would also have to revamp the lighting system, invent a texturing system, and implement true 3D math and add 3D renderer – and that’s going to slow the entire render pipeline down. In essence: create an entire 3D app.
Element3D is truly a huge achievement and spectacularly fills a gaping hole, and does a great job at a screaming pace. And I don’t think the limitations are Andrew’s fault; he’s had to work within the frame work of what AE lets you do as third party developer.

[Steve Bentley]

Ah I see, its an approach difference. (at least it appears to be). When we use real flow we make a batch of .bin files
in RF (basically OBJ meshes for every frame). These remain outside C4D (keeping the file size down) and we use the plug in that comes with RF to load them into c4D. This lets us change out the mesh after the fact without upsetting the project.

In your case (I think) you are loading the fluids right into C4d (which is fine). But here’s what i’m wondering. Because the meshes from RF aren’t actually moving in time – each frame is a snap shot of where the particles were in RF – I don’t think they are going to react properly to the dynamics engine. That engine is looking around for things your earrings need to collide with and suddenly, on the next frame, out of now where, the earrings are suddenly touching or embedded inside a blob of liquid. The engine will say “where the heck did that come from?” The mesh from the previous frame wasn’t approaching the earnings and wasn’t in danger (on that previous frame) of running into the earrings so the engine would have been oblivious to it.

So I’m not sure the dynamics approach is going to work.

What you could do is make a blobby object that approximates the fluid movement and you hand animate it to be a collision proxy (or as we call them around here – stunt objects) or make a particle/mograph system inside C4D that travels like the one from RF does, but it doesn’t need to be high a resolution and it wont be used in the render. This particle system is used as a proxy to trigger the crown splash when the earnings intersect the particle stream.

[Steve Bentley]

If you are using AE 2018 R15.03 or above, it could also be that the codec the video is in is no longer fully supported. AE 15.00 still supports every codec back to the stoneage I think.
It could also be (if its an mp4 or similar) that what are seeing is the frames in the codec that don’t actually exist. Some compressions “make up” frames where the data isn’t changing enough to warrant one. AE wants discreet frames to play with. Michael’s suggestion of transcoding it is a good one. Make it something that AE can play with properly.

[Steve Bentley]

You can also use that spill as a key to pass through the color that is coming from the screen in your final comp. The actor would naturally have a color spill that emanates from the monitor and it will only appear where the green is spilling onto their face.

[Steve Bentley]

And remember that because the actor is going to be composited on top of, what amounts to, the background he was already on (minus the offending objects of course) your rotos don’t have to be perfect like they would be if the actor was shot on a green screen or a completely different background. Any errors in the roto mask will only reveal what was already there.

[Steve Bentley]

Good point M. I forgot about the 8 light limit. There are cheats around it but then you feel like you’re adding a cheat on top of a cheat.
Of course these days (those young whippersnappers!) no one wants to light anything, they just want to hit the GI button.

[Steve Bentley]

Radius doesn’t affect the distance the “in focus” area extends to; the radius number is basically the amount of blur. Yes you can reduce the blur or radius amount and that will seem to extend the in-focus sections but its only because the difference between in focus and out of focus is less. Kind of like when you have very little out of focus its hard to see where the roll off is from in focus to out of focus.

What “radius” does from a technical aspect is it determines how many pixels to include in the blur. The more pixels (the radius number) the more will be included in the averaging of that pixel group. If only two pixels are included, only that pixel and its immediate neighbors will be averaged, so you can still see lots of detail. But if you choose 10, a square of 10 pixels by 10 pixels is going to be averaged together, so any details smaller than 10 pixels are going to be smudged together with 100 other pixels.

[Steve Bentley]

You’re right I got those backwards. But remember that the position is where the anchor point is. So if you have a 501 pixel wide solid and you put it on an full integer pixel for position, the edges are going be half way across a pixel and you will get antialiasing.

Keep in mind too, if you are going to video, sharp edges aren’t always desired – they can ring or vibrate.
Also the higher the contrast between the shape and the background the more noticeable this “problem” is going to be as the antialiasing has to go through more steps to transition from the solid color to the background color.

[Steve Bentley]

You are correct. The link between AE and cinema is so that you can put 2d renders of 3D objects into a 2d or AE’s version of a 3D scene and have those objects follow along with a camera that has come from AE or C4d. So in the end everything reacts as if the footage from C4D “feels” like its a 3d Object. But if you change something like a camera angle in AE, you have rerender the C4d file with the new camera position and replace the rendered footage in AE.

It also lets you bring objects from AE into C4D. But those objects being exported in this direction are cameras, lights, nulls and 3D planes.
Remember AE isn’t a 3D program in any way. Its 3D renderer is cheating. Element manages this trick because its a 3D program on its own even though it lives inside of AE. It renders the object in 3D within its own interface and then passes each frame to AE’s stage.

Yes element 3D is great way to go to keep all elements live and in the same work environment. There are a few other plug ins out there that will also accept a 3D mesh from an OBJ and have it be a live 3D object in AE – however these other plug ins (excluding E3D of course) don’t have very good texturing routines (or none at all) and you get simple spherical or cubic mapping at best and simple gray shading at worst.

Once you have a pipline set up for rendering files in C4d and then using that footage inside AE it’s pretty slick. While we have both full C4D and Element 3D here, I still feel, compared to E3D, I have more control inside of C4D – better navigation, a more standardized work environment (as 3D apps go), a more robust texturing system and animation control/timeline, all the attributes for each object are attached to it and you don’t have to pretend that each object is a particle and edit their attributes “outside” of the program where you texture and place the objects (as you do in E3D).
If you have both C4D and AE running together, you end up just exporting cameras back and forth, then importing/replacing the rendered footage. Jumping from AE to C4D is not that different or any more handcuffing than jumping from the comp in AE to the inner interface of E3D.

[Steve Bentley]

There also seems to be a bit of morphing or “reshaping” going on – look at the nose and when the mouth smiles – its more than just a displace or dissolve, the structure actually changes shape through the transition. I don’t think its a morph on the entire head though.
What a great look that gives!