[Steve Roberts]

I’ll bet you were trying to make an SWF and AE ignored the Unsupported Features.

[Steve Roberts]

Couldn’t have said it better myself!

But regarding roto, I have to be the broken record here:

1) study the footage carefully before you crack open the pen tool. Make note of the points in time where the to-be-masked thing’s shape changes direction or speed. For example, if something moves to the left, then up, mark where it changes direction. If it gets bigger, then smaller, mark that change. A marker on the layer should do it.

2) If something (like a head) just changes position, not shape, draw a mask for the head alone. Set a mask shape/path key at one of the points I mentioned. Move to the next point and move the mask — check the manual for that. Note: if things change shape/direction at different times, you may need to set up null layers (one for each thing to be masked) and mark those lest you get confused with a whole buncha markers on one layer.

3) Rotoing with a paintbrush sucks. Too much jitter. It’s a last resort. The pen tool and masks are better because AE automatically interpolates the position of mask vertices for you. Compare this: moving a mask over 30 frames, versus painting every frame for 30 frames. Yecch. I just masked a tree branch in a tracking shot. Drew a mask once, then moved it. Piece of cake.

4) Note: if you rotate a mask, the vertices move in a straight line. It’s just the way it works. Sometimes Smart masking helps.

5) Use different colors for each mask outline. Check the manual.

6) Feather the mask a pixel or two.

Hope that helps.

[Steve Roberts]

Apply the lens flare to a black solid, set to Add mode, on top of the layer stack?

[Steve Roberts]

Just to clarify, AE imports images rendered from 3D apps, not 3D scene files, not 3D model files. So MAX files? No. The Adobe doc could be clearer on this issue.

In some cases, you can also import lights and a camera to work with AE’s 3D layers and that rendered image(s) of which I spoke.

However, through the Zaxwerks Invigorator plugin, you can import some 3D models into AE, but you’d have to confirm that on the Zaxwerks site since I haven’t used Invig in a while. C4D for me, see?

[Steve Roberts]

Can we make this thread a sticky?

[Steve Roberts]

[Dave LaRonde] “… simply drag your clip onto the Make Comp icon at the bottom of the project window, and you get a comp in you’r clip’s resolution and length.
“

I always do that when working with a comp that needs to match footage. Avoids human error.

[Steve Roberts]

The contents of the timeline do not affect the playback of the final rendered file.

My guess is that your rendered file is just too big (too much data to push through) for your computer to play back smoothly. Big files are meant to be played back on computers with fast hard drives, such as those configured to actually edit HD video.

Consider the delivery platform of the file. If you intended the file to be sent to a video editor, your only concern is whether he/she can play back that file. If you intended the file to be played back on a hard drive, compress the file to H.264, ProRes HQ or maybe Photo-JPEG until you can play it back smoothly. If you just want to check how the file looks, compress it until you can play it back on your system.

[Steve Roberts]

Hmm … shouldn’t DVCPro be 720×480? Not 486?

[Steve Roberts]

… and we don’t know what the codec is, do we?

If it’s 320×240, I doubt it’s in the Animation codec.

So, extending Bartek’s analogy, if it’s the Sorenson codec, you’d be taking a rotten scrambled egg and making a fresh, whole one out of it.

So I’d be prepared to say you can’t do it and make it look good.

[Steve Roberts]

Not stupid. 🙂

I’d add the time at the end of each precomp, then go into each precomp and select-all-drag the layers in each precomp just as you did in the main comp.

Anybody else?