[Steve Bentley]

I guess another thing we need to know is how are you rendering. With GI? With another render engine?
Should the gel be luminous or not?
For the tendrils you could do that with particles or a cloner or the hair engine.

[Steve Bentley]

Keep in mind too that AE is just soooo flexible and there are sooo many ways of doing any particular task that not everyone will work the same way or use the same techniques. (just have a look at the differences in opinion over in the thread about the graph editor). As a 26 year vet of AE I’m sill learning things as I look over someone else’s shoulder who has maybe only l or 2 years on the books. It’s also the reason I’m active in the forums. I can’t tell you how many times I’ve read a post and thought, “what a great solution to that particular problem”.

When we are looking for new hires, the things that stand out are not how good they are with the software but how well they handle the concepts of animation, fx and motion graphics: things like understanding mattes and track mattes, easing and the natural dynamics of motion, lighting and depth, and on the FX side how well they look at the world so they can replicate something and make it look real.
I find the software mechanics are far easier to teach than all the other skills you get from a good artistic background.

The biggest difference between a Pro and a dabbler is the ability to deliver what you said you would in the time and budget agreed.

[Steve Bentley]

Walter brings up a good point.
As much as you shouldn’t design for the latest neat plugin, I’m trying to remember the last time one of the Trapcode suite of plug ins wasn’t part of a project we had done. They have become so ubiquitous and so well rounded I think people might look at you funny if you said “I don’t have Trapcode.”, or at the very least Particular. They also fill a much exploited niche (at least in 2.5D) that no other plugin really measures up to. To get equivalent effects you have to go to a 3D package.

[Steve Bentley]

On the PC side there was the first version of AE2018 called R15.00 or .01. If you can get the Mac equivalent it still allowed all the old codecs to work properly.

[Steve Bentley]

For those who have wandered in to this post but don’t use Mocha, you can apply tracking data of a shot to its anchor point and voila, its now stabilized around the object you tracked(this only works if the thing you tracked is not moving). You may have to nest the shot to a null to then move the footage someplace where you can see what you want in the shot, and even reduce the comp size or scale the footage, as stabilization can move the edges significantly outside the frame. Or you could put an offset on the data with an expression.

[Steve Bentley]

So like a Jello mold with the core being deep red and that fades to the outer color of green? (how festive!)

I haven’t tried any of these but…

1) What if you apply a fresnel effect to the transparency of the inner object so that its edge just sort of fades away. If you make both the outer object and inner object the same RI you shouldn’t sere that transition (I haven’t tried this but it would work in a physical sense in the real world)

2) What if instead of doing it with the color of the material you do it with light. Put a light in the center of the object that has a distance fall off that is not quite the edge of the object and have that light red, and then have lights surrounding the object that are green with fall offs that don’t reach the core. You could make all these non specular lights (so you don’t get a bunch of highlights) and then do some studio lighting with just spec hilight lights where you want them. These studio lights would not add to the brightness of the object, just the highlight.
If this creates an object that is too evenly lit you could use additional white lights with negative values to reduce light in some areas.

3) I haven’t tried using the subsurface scattering effect with a transparent object but that might work as well using a modified version of idea #2

4) You could try a 3D noise with a really big scale and a blob of one color of the noise being the size of the core and the rest of the noise being the other color. So mapped to a flat the noise would look like a grad from one color to the other.

5) What about a pyrocluster shader but invert the puffy gradient so edges are sharp. And use two different colors within that “puff”.

6) Depending on shape of object, you can often put a card behind the object that is a grad of the colors you want and have the gel object refract the card through it. The card would always face camera and always be on the back side of the object. We use this trick to get just the right look to full coke bottles. Sometimes you can put a compositing tag on the card and have it not be seen by the camera and only be the refraction and transparency.

[Steve Bentley]

I couldn’t see the image (password required for the googledrive) but are you running the Dell at full resolution? Some of the newer software (AE especially) assumes you want to run 1920 or higher and some of the menus won’t populate if they don’t feel there is enough resolution to do that in (usually its a Y resolution limitation).

I doubt this is the issue but I thought I’d mention it.

[Steve Bentley]

Good point Angie. Like a man with a hammer I was only looking for some nails to hit. I was thinking purely about true 3D objects that you could rotate around and slide existing 2D assets between multiple 3D objects.

To be clear for the OP: using Cineware creates a 2D layer of your 3D C4D scene. This is a huge time saver over sending data back and forth between the two packages – make a change in C4D, and it updates immediately in AE.

However, there is a catch for all this elegant integration:
Lets say you are doing a solar system in AE, and lets say you do the planets in C4d but you want to do the sun in AE. And you want your camera to view the planet orbit “rings” obliquely (the way Saturn is usually depicted) – the orbits become ellipses.
If the Cineware layer is on top of your sun layer (an AE layer remember), when the planets pass in front of the sun, all will be well, but when the planets need to pass behind the sun, they will still render in front because the Cineware layer is on top of the Sun layer – even if your sun layer is a 3D layer.

This is true of Element 3D as well. Even though you can move your objects around in what feels like 3D space, the layer that Element 3D is on as an effect, still renders as a 2D layer by itself before it is stacked with the other 2D AE layers.

You can however use a separate layer for every 3D object in your scene (with both Element and Cineware) and stack those layers in the correct order so that they overlap as they should, with each other and with other AE3D layer “flats” that are interspersed in depth.

[Steve Bentley]

Hey Angie,
I wanted to make sure I wasn’t missing something. Your phrase:
Also, any subsequent scaling or rotating will happen around the anchor point, which is now somewhere unsuitable.
outlines a problem but seems to be a pro for doing it this way.

If I have kept the anchor point closer to stage-center and then I want to scale that object, the distance between the anchor and the object will also scale, thereby moving the object from where I placed it. Similarly, rotating it will inscribe the object along an arc instead of just spinning it in place. And as you know, clients always wanna change stuff after you’ve gotten it just right.

This also shifts the problem I described in the OP to the camera view, which I wasn’t expecting. Using your method I can now see all the anchors (you’re right of course; the axis widget is defining where the anchor point is) in the ortho views no matter where the object is, but with the camera venturing out far enough to pass by objects in the extreme foreground, objects in the background, now viewed through the camera, visually loose their axis. So again it seems like there is camera clipping going on for te axis. But its odd that bounding boxes don’t get clipped in either method and the actual pixels of the object still render.

[Steve Bentley]

I’m with Roel., I much prefer the Value Graph. You end up with a shape as pictured below.
As the red (X) curve falls its moves more and more to the left (smaller x values are left) and the as the curve flattens to horizontal the layer’s movement slows until the curve is flat and the layer stops moving (in other words it’s values stop changing). Then as the curve start to fall again (therefore moving in the same direction) it moves farther and farther left.

Its very hard to visualize an “ease” in a speed graph, especially when that speed changes after the keyframe – you end up with lots of infinite peaks and valleys in the curve. Its also hard to visualize acceleration – the key to any good animation.

There are so few places to use “speed” as an attribute in AE, that I find it makes more sense to use position values, so why not graph those values?
The Speed graph in AE is actually an anomaly in the industry. Pretty much all of the 3D packages (where this graphing thing began) use a value, or “F” curve graph (F because the curve is a Function of position – the value – over time. and isn’t position/ time just “speed” in the end?)
I also see in the forums so many people getting into trouble when they are trying to change the speed of something, especially with expressions. AE has no memory of what was happening on the frame before the one it’s on, so you can’t increase the speed of something if you can’t find out how fast it was going before.