[Steve Roberts]

Thanks, Todd.

[Steve Roberts]

Is there any advantage to doing it “old school”, that is, manually launching new instances and rendering to a TGA sequence using “skip existing frames”?

I say that, given that I’ve read that it’s sometimes advantageous to switch off multiprocessing.

[Steve Roberts]

Cool.

I’d distort the text a bit to follow the curve of the road.

[Steve Roberts]

Vid 2 may be looking for a video that’s local on your machine …?

[Steve Roberts]

Posterize time?

Or: Apply time remapping and make keyframes at the frames you want to keep where they are. For freeze frames, make those hold keyframes. Make the graph editor visible. Mess with the graph, adding new keys, but making sure you don’t touch the keyframes you made at the “keep” locations.

Note that after a hold frame, the value graph would be flat, then just before a “keep” keyframe, the graph would either slope up steeply (fast catch-up motion) or jump up immediately to show the next frame.

I hope that makes sense.
Just realize that the value of a time remapping graph is the frame being shown. If you imagine the movie as a series of stills that can be shown however you want, it might be easier.

[Steve Roberts]

You could wrap a green fabric around the talent’s torso.
Or use the second “body” greenscreen on a sandwich board and use the shadows on it as shadows for the building …?
Hey, do a test.

[Steve Roberts]

Nice link, thanks. I’d check to see if there are any changes in syntax between CS3 and CS4.

I forget the programming term for this, but I created a text layer and tied its source to the time of the keyframes as indicated in your expression to make a thing that displays the value of a line in the expression. When it didn’t return the time of the keyframe, I realized that something was missing, so I added “time” to your keyframe time-finders, first at the beginning, then at the end (duh).

After that, it was the “else” and brackets that needed to be fixed.

[Steve Roberts]

I beg to differ, O Long-Lost Brother of Mine. 🙂

Check the motion of the floater screen: mostly up and down.
The handheld device moves from side to side and rotates.
The floater misses this motion.

[Steve Roberts]

Try this, to wiggle between keys 1 and 3:

//get the specific keyframes
timeA = transform.position.key(1).time;
timeB = transform.position.key(3).time;

//wiggle properties
a = wiggle(40, 40);
b = wiggle(20, 30);

//condition
if ((time > timeA) && (time < timeB)){ [a[0], b[1], value[2]] } else { [value [0],value [1],value [2]]; } (I had to search for if/then and time (comp methods) examples in the help, and insert the else to return the "normal" values. Been a while ...)

[Steve Roberts]

Blah: rotation is the key. The handheld device rotates. You need to track it with rotation. It should move as if it were a big clipboard bolted to the handheld device.

If you can’t track it, you’ll have to do it by eye. It would be good training for your eye. Make sure the big layer’s anchor point is where the handheld device rotates … his wrist maybe?

Blah 2: video didn’t load.