[Steve Bentley]

Boujou has a vector component for all of its tracking points. You can link your object’s orientation to the vector (the direction it points)

[Steve Bentley]

Also if you are supplying a “bug” to a broadcaster make sure to get and understand their specifications for how they need it as an element.

[Steve Bentley]

Joe’s right. CT3 is pretty impressive. Its more like a traditional animator’s dope sheet set up and animation package:
But we also love Duik.
You could get where you need to go with either. We use both. So here’s the pros and cons.

Duik – Its not so much software as a bunch of expressions tied to a script which gives you some really powerful IK solutions inside of AE. But in the end it’s just expressions with the puppet tool in tow. AE is doing the heavy lifting.
Pros:
You get to use it inside AE. So if you have complicated layering to be done, you’re home free.
All of AE’s power is at your fingertips and everything is in one package.
Its free!
Its easy to use and quick to learn
If you are already used to the puppet tool its easy to get great results
Its elegant.

Cons:
You have to divide up the content (arms legs etc) inside AE (which can be cumbersome) or in pshop (better) but to make changes you often have to go back to the source art (outside of AE)
If you want to update the character after the animation has been done it can cause issues with the puppet mesh deformations.
More planning with character art is required before the animation stage.
Duik can’t do anything you couldn’t do yourself with expressions so its not really a new way of doing anything. A limitation of AE will limit Duik.

CT3 – this is a complete stand alone animator software. Once you have created your animations they can be exported to After effects and you can do your composites there.
Pros:
CT3 uses bones instead of pins to warp the mesh. This might be better for 3D animators to get the hang of. It should technically give you better results than puppet pins but if you are a veteran of puppet pins you can achieve identical results to bones. But the bones in CT3 mean you have fewer nodes to animate to get those results.
Changing the art after the animation is done tends to be less problematic than Duik
Tons and tons and tons of presets – like walk cycles and 4 legged movements
Tons and tons if preset motions – like squashes and stretches
Massive library of characters and art available
Art is “live” while you are animating so you aren’t stuck with a pshop file and you can make/change the art inside of CT3 (if it was made there).
Vectors tend to be less problematic than with Duik (its not really a duik issue – any duik issue is usually an AE limitation)
You won’t be pooched when adobe changes something to make the plugin obsolete or unuseable.

Cons:
It costs a bit more ($50 to 200) depending on your usage.
You can’t do a complex comp within the software like you can with AE. You don’t have all the effects and plug ins that AE has (yet). But you can make elements and bring them into AE.
To my knowledge you can’t “paint” weights yet. Instead you use anchor points like one would use in the puppet tool. Coming from a 3D animation background I miss this feature

One place we have a better pipeline with Duik/AE is when we have precomped animated characters, then in the outer comps we’re adding things like smoke or light effects that track with the character, and time remapping on the animation etc; we can use the adobe expression system to link things inside the precomps to track particle effects (or other things) in the outer comps. You would have to make additional elements and do some tracking to get the same effect in a CT3/AE pipeline.
Why would we want to?
Lets say you have character with a torch in a precomp. You can speed up and slow down and bounce around the animation timeline to get the performance you want AFTER the animation has been done. If you attached a particle smoke system to the torch in the precomp of the character animation the smoke would change speed (or reverse) or scale or just not work right as the animations timeline is remapped. But if you do the smoke in the outer comp and link its position to the torch in the inner comp with expressions, speed ramps are no longer an issue and the smoke just keeps chugging at the correct speed even though the animation might be jumping all over its timeline.

[Steve Bentley]

Based on your code, when the value falls below 35 the y position will instantly pop from 100 to -100, similarly when the value hits 0 the value will instantly pop from -100 to 0. Is that the intention? And when you go below 0 the value will pop to -100 again.

Or do you need something like a range mapper? Where you have a set of values (say 0-35) and you want to linearly map another set of values to that (say -100 to 100).
If you call the set of values 0-35 values set1 and call 100 to 100 value set2 here’s the equation for that. Assuming you have a value from set1 (valueFromSet1) that you want to map to the values in set2.

newSet2value = lowSet2 + (valueFromSet1 – lowSet1) * (highSet2 – lowSet2) / (highSet1 – lowSet1)

[Steve Bentley]

Cassius is correct. I had used this trick with a precomp that had some time warps on it (in the pre comp) and it works great. But with the time warp (or time remap) in the outer comp you would have to use valueAtTime method.
You could get that value by looking at the time remap value in the outer comp- this will be the frame number the inner precomp is on. Remember to convert frames to time as the valueAtTime method uses seconds not frames.

[Steve Bentley]

Sure, precomp your body animation (but not the time remap) and then you can put a null in the character chain inside the precomp that has the time remapping on it. Retime the precomp layer in the outer comp.
Inside the precomp, put the anchor point of a null where the mouth belongs and make it part of your bone chain on warping system so that it tracks along with the head.
Then in the outer comp link your animated mouth element to the inner null with toWorld positioning expressions. This way where ever the inner comp animation moves the outer one tracks to the same position, regardless of the time warp on the body animation. You get to run your body back and forth from the outer comp and animate the mouth in the outer comp, but the two will be positionally joined.
The expression goes like this: (and it goes on the position attribute of the mouth layer in the outer comp) – everything after // is a remark.

preComp=thisComp.layer(“preComp”); // preComp is the precomped animated bodylayer. If you have named it something else then just replace the “preComp” – only the one in quotes – with the name of that layer or use the pickwhip
object=comp(“preComp”).layer(“Object1″); //”Object1” is the name of the null that represents the mouth in the precomped body animation. Just replace “preComp” with the name of your precomped animated layer and “Object1” with the name of your null layer in the precomp, or use the pickwhip.
p1=object.toComp(object.anchorPoint); // all of these things have been defined in the code so far. So p1 will hold the position of the anchor point of the null that is standing in for the mouth in the precomp
preComp.toWorld[p1[0],p1[1],p1[2]]; // this line calculates the position of p1 but where it would be in the outer comp for x,y and z – you can drop the ,p1[2] if you don’t need the Z.
// this is the last line of the code so whatever is calculated here gets put into the attribute the code is on.

[Steve Bentley]

The best way is to precompose your sources. A good rule of thumb with any warp effect as there can be distortions and squeezes you might not think of before hand. With a procomped source you can move scale or in your case flip the content so that when its viewed in the outer comp (with the distortion effect on), it looks correct.

[Steve Bentley]

You can always make a copy of your key layer and crush it and blur it and then use it as a garbage matte so the main key doesn’t add noise way beyond the head on top of the text.

[Steve Bentley]

Ya thats a problem with Mixamo – they use four different (as far as i’ve counted) skeletal paradigms that don’t mix well.
What we’ve done is duplicate that skeleton/mesh, fix the second and then retarget the first mesh to the second skeleton.

You can also take the rotated “bone” and shift the keyframe values of just that object in the timeline fcurves. Select just the single rotation direction and all its keyframes in the time line and drag them up or down until they are rotated the right way.

This way you can alter the rotation of that object but keep all the animation intact.

You could also do an expresso with a math node (and a degree to rad and visa versa nodes on either end- C4D does its expresso math in radians) and just pass the original value in and get a new value out.

[Steve Bentley]

Here are two ways.
If you can give the position value a formula and have the sliders produce values to use within that formula.
So lets say the position is a sin wave type shape. You could set the position with a formula of x=sin(time)… this is oversimplified but I”m hoping this might help. Then you bring in your slider and you change the formula to x=sin(time)+sliderValue. This way the base motion is there and you are just modifying the whole thing with a slider “offset”.

or with keyframes already on your position attribute:

xVal=value[0]+effect(“Slider Control”)(“Slider”);
[xVal,540];
this way you can have your keyframes x value amplitude be controlled by the slider but keep your initial keyframes value. And you can keyframe your slider to have that change factor change over time. (the 540 number is just a dummy number for the y value so it stays the same.
value[0] is the first component of the position (the x) of the keyframes already defined in your timeline
[xVal,yVal] is how your write the position value when you want to break them apart.