[Mike Clasby]

Oh, yeah, if your using ai files don’t forget the Continuously Raterize switch (like a little sun) and Advanced 3D for the Comp (in comp settings>Advanced).

[Mike Clasby]

Actually there are two AE cheats, one with Shatter and one by stacking layers in Z space.

1) Shatter for depth tut here:

Click on Roland Kahlenberg’s head above and scroll down to, Adding Z-Depth to Your Logo.

2) You can do it pretty well with just dupping the layers and separating them in 3D z space. Pretty good refers to not lingering in the dead edge-on because the edge is a little fuzzy there, it’s not as bad as it sounds and can be made Really Pretty Good by tinting the middle layers Black or another solid color , easy to do, quick to see if you like.

Make your layer 3D.

Add this expression to the position of your ai layer:

position+[0,0,(index-1)*1]

This just tell the layer to separate itself 1 pixel in the z direction from the layer above, so when we dup them they arrange themselves like a real thick sandwich.

Dup it three times.

If you want a different color for the side, apply the Render > Tint to it (choose your color).

Now dup the layer as many times as you want it pixels thick.

So 10 layers = 10 pixels thick.

Add a camera (at the bottom) and fly around or rotate your letter in 3D, just don’t linger too long on that edge, not as bid a problem as it sounds.

If you want to get fancy you can tie the “1” (last one in the expression) to a slider and bunch them up closer than 1 pixel.

[Mike Clasby]

Actually there are two AE cheats, one with Shatter and one by stacking layers in Z space.

1) Shatter for depth tut here:

Click on Roland Kahlenberg’s head above and scroll down to, Adding Z-Depth to Your Logo.

2) You can do it pretty well with just dupping the layers and separating them in 3D z space. Pretty good refers to not lingering in the dead edge-on because the edge is a little fuzzy there, it’s not as bad as it sounds and can be made Really Pretty Good by tinting the middle layers Black or another solid color , easy to do, quick to see if you like.

Make your layer 3D.

Add this expression to the position of your ai layer:

position+[0,0,(index-1)*1]

This just tell the layer to separate itself 1 pixel in the z direction from the layer above, so when we dup them they arrange themselves like a real thick sandwich.

Dup it three times.

If you want a different color for the side, apply the Render > Tint to it (choose your color).

Now dup the layer as many times as you want it pixels thick.

So 10 layers = 10 pixels thick.

Add a camera (at the bottom) and fly around or rotate your letter in 3D, just don’t linger too long on that edge, not as bid a problem as it sounds.

If you want to get fancy you can tie the “1” (last one in the expression) to a slider and bunch them up closer than 1 pixel.

[Mike Clasby]

You’ve got the right idea with the tweaks. These two posts from Chris Smith give detail to you question. The original thread had about 20 entries in case you want to read all.
An Advanced search for camera null and Chris Smith reveals all.

Name: Chris Smith
Date: May 11, 2006 at 9:25:45 pm
Subject: Re: Complex Camera Moves

I shut off the camera’s POI. Then layer groups of animation to nulls. One null is position only, next null is rotation only, next null is shake/vibrate only, then lastly the cam. All of these in a parent/child chain.

This way everything is in it’s own layer of motion and not all mixed together. Also each null works relative to 0,0,0 so any expression writing is easy.

Chris Smith
https://www.sugarfilmproduction.com

Name: Chris Smith
Date: May 12, 2006 at 6:43:58 am
Subject: Re: Complex Camera Moves

I have no secrets 😉

1. null_CamPosition
2. null_CamRotation (Parent = null_CamPosition)
3. null_CamShake (Parent = null_CamRotation)
4. Camera (Parent = null_CamShake)

I do this so I am never commited to any keyframing that is done, I can still offset things how I wish. For example for the cam to vibrate, it uses up the cam’s position parameter. But if the vibrate is a parent of another null, it moves first from the first null, then it adds vibration to it.

Also if you have rotation on it’s own null, for example, the rotation null can be the camera’s general aiming direction as it moves through space (from the PosNull). But then you still want to override the camera angle to look at something for a moment or to offset it temporarily, then you can still key the camera’s actual rotation and it won’t effect the rotation you already have from the RotNull. Or if your PosNull had a good animation path, but you just think it should be tweaked in space because it’s not quite the right spot, instead of moving all your keys, simply drag the camera layer around (Since it doesn’t have keys itself, it just rides along).

Like when I do a motion track, I paste the track data to a null and not to the video layer itself. Then I parent the video layer to the null. This way dealing with offsetting the layer is as easy as grabbing it in the comp window and dragging it to a new position. Yet it still tracks along because it is parented to the null with the track data.

Chris Smith
https://www.sugarfilmproduction.com

[Mike Clasby]

You’ve got the right idea with the tweaks. These two posts from Chris Smith give detail to you question. The original thread had about 20 entries in case you want to read all.
An Advanced search for camera null and Chris Smith reveals all.

Name: Chris Smith
Date: May 11, 2006 at 9:25:45 pm
Subject: Re: Complex Camera Moves

I shut off the camera’s POI. Then layer groups of animation to nulls. One null is position only, next null is rotation only, next null is shake/vibrate only, then lastly the cam. All of these in a parent/child chain.

This way everything is in it’s own layer of motion and not all mixed together. Also each null works relative to 0,0,0 so any expression writing is easy.

Chris Smith
https://www.sugarfilmproduction.com

Name: Chris Smith
Date: May 12, 2006 at 6:43:58 am
Subject: Re: Complex Camera Moves

I have no secrets 😉

1. null_CamPosition
2. null_CamRotation (Parent = null_CamPosition)
3. null_CamShake (Parent = null_CamRotation)
4. Camera (Parent = null_CamShake)

I do this so I am never commited to any keyframing that is done, I can still offset things how I wish. For example for the cam to vibrate, it uses up the cam’s position parameter. But if the vibrate is a parent of another null, it moves first from the first null, then it adds vibration to it.

Also if you have rotation on it’s own null, for example, the rotation null can be the camera’s general aiming direction as it moves through space (from the PosNull). But then you still want to override the camera angle to look at something for a moment or to offset it temporarily, then you can still key the camera’s actual rotation and it won’t effect the rotation you already have from the RotNull. Or if your PosNull had a good animation path, but you just think it should be tweaked in space because it’s not quite the right spot, instead of moving all your keys, simply drag the camera layer around (Since it doesn’t have keys itself, it just rides along).

Like when I do a motion track, I paste the track data to a null and not to the video layer itself. Then I parent the video layer to the null. This way dealing with offsetting the layer is as easy as grabbing it in the comp window and dragging it to a new position. Yet it still tracks along because it is parented to the null with the track data.

Chris Smith
https://www.sugarfilmproduction.com

[Mike Clasby]

As for understanding expressions, Methinks the two best places are here:

Dan Ebbert’s site (also, see reference books at bottom of home page)

https://www.motionscript.com/

and Gifford’s (very basic but not simple stuff, probably a good first stop)

https://jjgifford.com/expressions/basics/index.html

[Mike Clasby]

As for understanding expressions, Methinks the two best places are here:

Dan Ebbert’s site (also, see reference books at bottom of home page)

https://www.motionscript.com/

and Gifford’s (very basic but not simple stuff, probably a good first stop)

https://jjgifford.com/expressions/basics/index.html

[Mike Clasby]

Hey, I dunno how it works mathematically, but practically the expression on the camera tells it to focus on the layer number that corresponds to the keyframe number set by the slider in the Null (layer 7), So you can tell it to change focus to another layer by changing the Slider keyframes in the Null layer. So if you want to focus on another layer, say layer 3 (third layer down), set a keyframe for 3 and the camera will shift to layer 3. Transition time is hardwired into the camera expression and I think Dan talks about that in the tut. It looks like to me it take 1/3 to 1//2 second (10 to 15 frames) for that camera to track down the new layer.

Hope that helps.

[Mike Clasby]

Hey, I dunno how it works mathematically, but practically the expression on the camera tells it to focus on the layer number that corresponds to the keyframe number set by the slider in the Null (layer 7), So you can tell it to change focus to another layer by changing the Slider keyframes in the Null layer. So if you want to focus on another layer, say layer 3 (third layer down), set a keyframe for 3 and the camera will shift to layer 3. Transition time is hardwired into the camera expression and I think Dan talks about that in the tut. It looks like to me it take 1/3 to 1//2 second (10 to 15 frames) for that camera to track down the new layer.

Hope that helps.

[Mike Clasby]

There are several Converters at cnet.com