[Steve Bentley]

Even the though the dolphin will porpoise up and down (I couldn’t resist), the wake shouldn’t. So you could link the x motion of the wake to the dolphin and then the wake will follow but not rise and fall. Then you could put a slight animation on the y motion of the wake so it’s not exactly linear.
You can break apart the x and y position attribute in the curve graph editor (the button is at the bottom of the graph). Or you could use an expression.
On the position attribute enter: [use Pickwhip here to link the x of the dolphin, some Y value];
The notation for separate x and y in expression is [x,y];
You could link the Y to a keyframed slider effect or use keyframes on the position. Expressions can pick up keyframes from the source layer with the Value method [this.Comp.layer(“Dolphin”).transform.position[0],value[1]];
value[1] means Y, value[0] means X. Only using the y value will ignore the keyframed x value and use the dolphin x position for the x of the wake.

[Steve Bentley]

Oh I hate coincidences like that – 72 points to the inch and the difference is 1/72 between the two programs.

I get 72.223 when I do the same. I thought the dpi (or PPI) settings in AI might affect the output but they don’t appear to. I can’t however find a reason for the small discrepancy. It seems but a small difference but that error is equivalent to 5mm – a gaping chasm in 3D printing land.

[Steve Bentley]

Thinking about a better pipeline for you – you can now 3D print from Pshop. Perhaps there is a better scale for scale system set up between the two apps.
I would have them rolling in the aisles in the Pshop or AE forums with this suggestion. We’ve been waiting for some Pshop features in AE for about 30 years now. But its worth a look.

[Steve Bentley]

And then when outputting to an STl for your service bureau, set the scale to 100 and the units to mm. This won’t change the size but will give the printer the units it’s expecting.

[Steve Bentley]

Changing to points (or to mm for that matter) doesn’t change the size of the thing, it just changes the numbers. 1″=2.54mm=72 points.
You’ve got the scaling factor I mentioned. Keep changing the scale factor on import to C4D until the size of the object it brings in is the same as the illustrator file. Then always use that scaling factor.

[Steve Bentley]

Illustrator is a vector program and all its “sizes” (apart from being arbitrary and not really related to anything) are based on the printing paradigm of dots per inch.
3D is also arbitrary – you can change the scale in a way where the object isn’t any bigger really, but in fact it is.
Neither package has a system like photoshop where a pixel is a pixel and that’s gospel. (so many errors have been propagated in both AI and Pshop with the use of inches – the next question should always be “at what DPI? – there is no question required with pixels – at least these days).
The only way to ensure accuracy in 3D printing is in the 3D package. AI and C4D don’t share a measurement system. You can set one up but it means finding out how big something is when it comes in to C4D and then changing your art board and resolution in AI to compensate. At a different DPI in AI you can have your rulers be accurate to final outline size in C4d.
Also we’ve found that outputting in inches in C4d can cause issues at 3rd party service bureaus for 3D printing. Most want mm – even in the USA. You can end up with a spinal tap object.
But with that said, when we create a 15mm cube in C4d it prints at 15mm +/- .01mm, so C4D is on board with 3D printing scales.

One other note – if you extrude those outlines in C4d and then chamfer or round the caps, you must use the constrain button in the extrude parameters to make sure the outline doesn’t expand to compensate for the chamfer.

[Steve Bentley]

Even with an unchanging object moving about in frame, the way that object is rendered in pixels changes its shape in the file ever so slightly, Add camera noise to that and no two frames of the exact same moving object are going to be the same. So that’s one use, to set a tolerance for an object that won’t be quite the same on the next frame. So when AE goes looking in the next frame for that thing, it says, we’ll that thing there is within a 20% or 80% tolerance of the previous frame so that’s probably what I should lock on to.

On the other end of the scale, you might have an object that is rotating or the camera is rotating around an object, (or worse, changing size) and that object may not be the same shape or color on all of its “sides”.

So in effect you are telling AE ,with that setting, to match that object either more perfectly or less perfectly so that there is some forgiveness in looking for that “exact” shape, because it will NEVER be exactly the same from frame to frame.

[Steve Bentley]

Can you upload the file and we’ll have a look?

[Steve Bentley]

I had this problem a while ago and did solve it. I’ll try and dig up the file. But in the mean time, I remember fiddling with the matte object check box in the compositing tag or including a background plane that the camera couldn’t see (or both)

[Steve Bentley]

With the precomping and time remapping solution that Walter suggests, the great thing is that because the animation is being done by AE, there are no missing frames or “mask” positions. Even though you may not have a actual keyframe that covers half of an “O” lets say, when you remap the time in the outer comp AE will interpolate this time dilation properly between the keyframes you do have. Its kind of the ultimate way to stretch or shrink time because there are never duplicate frames like you get when you just do a time stretch on running footage.