[Darby Edelen]

[Ben Waflik] “Well, I’m rendering on a black background into 32bit *.exr format. In the output settings, I choose zip compression and RGBA. There’s no premultiply checkbox or something like that.”

If you’re using the EXtractoR effect to extract your passes from the EXR then the premultiplied/straight interpret footage options in AE have no effect.

Instead you’ll want to use the ‘Unmult’ check box in the EXtractoR effect to interpret the footage as premultiplied, otherwise the assumption made by the effect is that the EXR has a straight alpha.

Also, if possible I’d recommend rendering to half float 16-bit EXRs as the file size is half as large and the quality is generally not much different.

Darby Edelen

[Darby Edelen]

What are you currently using for your environment map? Are you working in 32bpc?

Darby Edelen

[Darby Edelen]

Alternatively you could make the keyframe from your step (2) into a hold keyframe after you’ve extended the frame.

Darby Edelen

[Darby Edelen]

[Jared Gardner] “1) Add time remapping
2) Go to a frame I want to pause and create new keyframe
3) Move over one single keyframe ahead and create another keyframe
4) Drag the keyframe from (3) and the very last keyframe at the end in unison to extend it”

You do not want to move one frame ahead and create a keyframe as in your step (3). You should:

1) Add time remapping
2) Go to a frame you want to pause and create new keyframe
3) Copy the keyframe created in step (2)
4) Drag the keyframe from (2) and the very last keyframe at the end in unison without moving the current time indicator.
5) Paste the keyframe from step (3) at the current time indicator so that it is on the same frame that the original keyframe from step (2) was.

The values of the first and last keyframe of a paused section should be exactly the same, not 1 frame different.

Darby Edelen

[Darby Edelen]

[Jared Gardner] “Something that I’m still wondering though is that for the extended frame, it slows to 0.01 second/sec. Is that the default?”

No, 0 seconds/sec is correct for a frozen frame.

Darby Edelen

[Darby Edelen]

It sounds like the values of your keyframes could be incorrect. Alternatively they may appear to be correct but actually be incorrect due to a difference between the frame rate of the composition and the frame rate of the footage.

For example, if you had a 120 frame clip that runs at 30fps and you wanted exactly 1.0 seconds/second you’d need to verify that the time remap keyframe on frame 0 of a 30fps timeline has a value of 0 and the time remap keyframe on frame 120 (the last keyframe) has a value of 120 (all keyframe values are in frames in this example). This works because the frame rate of the footage and the frame rate of the timeline are the same.

If, however, that 120 frame clip runs at 25fps and you have it in a 30fps composition then the last time remap keyframe with a value of 120 should be on frame 144 of the timeline. If you were viewing time in timecode instead of frames this means that at frame 4:24 of your 30fps timeline your keyframe should have a value of 4:20 which may appear to be wrong but is in fact correct.

Intuition might lead you to adjust the the timeremapping keyframes to match the corresponding time/frame values in the timeline (changing the keyframe to 4:24 at 4:24 for example), but this will result in an incorrect speed if the frame rate of the footage and the comp are not exactly the same. I chose 25fps and 30fps as my examples because the difference is more extreme, but if you had 29.97fps footage in a 30fps comp or vice versa you could run into the same problem; the error would just be very small.

For the other problem you mention, there are three ways to adjust the speed of an animation:

1) Adjust the incoming/outgoing velocities of the keyframes.

2) Move the keyframes in time.

3) Change the keyframe values.

It sounds like you’re doing (1) when you should probably be doing (2). If you have a 1 second/second speed between two keyframes and you increase the velocity of the first keyframe without moving the keyframes then the speed must dip below 1 second/second before the second keyframe in order to keep the area below the speed graph constant. The area below the speed graph (the integral of the speed) between two keyframes corresponds to the total change in value from the first keyframe to the second keyframe. Since the total change in value hasn’t been altered (you haven’t (3) changed the keyframe values) and the time over which it’s changing hasn’t been altered (you haven’t (2) moved the kefyrames in time) the graph must run slower later in the animation to make up for running faster earlier.

Darby Edelen

[Darby Edelen]

[Shaun Carollo] “when placed into the comp, the alpha layer is not lined up as seen above”

If these are image sequences make sure that they are interpreted with the same frame rate. You can check this by selecting them in the project panel and looking at the footage information at the top of the panel. You can change the frame rate interpretation by pressing ctrl-alt-g (cmd-opt-g on mac) with the footage item selected.

Otherwise make sure that the layers in points are, in fact, aligned on the timeline. One way to do this would be to duplicate the RGB footage and the duplicate selected holding alt and dragging the alpha footage from the project panel to the timeline. This will replace the selected layer with the new footage source while keeping all other layer properties (in point, out point, transforms, all effects and properties) the same.

Also, since these are 3D renders it would generally be much cleaner to import the camera from your 3D package into After Effects instead of using the Camera Tracker. The workflow for doing this varies depending on your 3D package but there’s always a way to get that data in there.

Darby Edelen

[Darby Edelen]

Okay.

You’re describing scaling. It’s not moving forward at all. It’s just scaling. The important piece of information in your description is that it’s not actually moving at all in the top view.

Most likely the problem you’re having has more to do with the anchor point (the point from which a layer scales). Anchor Points default to the center of a layer which causes them to grow along all axes (X,Y & Z) when scaled, which sounds exactly like what you’ve described.

If you move your anchor point (or the anchor point of the control null) to the bottom of the helicopter (to its landing gear) before scaling then your helicopter will scale away from its bottom and appear larger while still “landing” in the same spot.

Another thing to consider is the camera you took the photo with may have a different field of view (zoom) than the camera you have in your After Effects composition. This can effect the relative sense of perspective.

Darby Edelen

[Darby Edelen]

The Control Null is not itself a limitation, it’s the work around of a limitation.

Any effect that renders 3D internally (the effect renders a 3D object/system) including many of the Trapcode effects (Particular, Form, etc.) and Element 3D needs to be applied to a 2D layer. The effect is responsible for rendering in 3D perspective based on the composition camera’s view and the effect must assume that it is on a 2D layer to do so.

If you apply these effects to a 3D layer they will not render properly.

The Control Null is a way to ‘place’ the object in AE’s 3D space. This way you can apply the effect to a 2D layer and still have an After Effects based 3D control to place the object in perspective. You only use effect controls or a 3D Control Null to place objects this way, you never move the 2D layer.

If you are still confused you’ll have to be more specific or upload pictures.

Darby Edelen

[Darby Edelen]

Or cheat even more and do it in compositing 🙂

Darby Edelen