Steve Roberts
Forum Replies Created
-
Try Leave Color or Change To Color.
-
Is motion blur switched on for both the layer and the comp?
-
In AE7 at least, select the layer(s) you want to see in wireframe, then hit composition>preview>wireframe preview. Make sure your layer with audio has its little speaker switched on.
-
Try searching the COW for “track vertices”. I believe there is a script that will allow you to attach tracked points to mask vertices.
-
Yep, you need 4 different comps.
-
Ah — I see. Then continuity between the sections isn’t an issue … just seamless wrapping around each section.
And we have the same number of LEDs at the top of a section as at the bottom, but they will be spaced farther apart at the bottom of each section, no? Hmm … how to map that …? If you don’t mind distorting things so they’re wider at the bottom, then simply placing animations on the ziggurat map would work.
Otherwise, to make rectangles look like rectangles on the tree, they should look like top-heavy trapezoids on the ziggurat, I’d think. You could do that with corner pin, I guess. Maybe then, you could just use the offset (or motion tile) effect to mirror-tile the animation at the ends …?
-
So it’s sort of like cylinder-mapping a discrete number of cylinders? Like a wedding cake?
If so, I can see a definite “stepping” effect when an animated object moves up to the next cyclinder … so I see an issue not just with seamless mapping around the cone, but being seamless within the cone.
Since the tree will be a cone, ideally you’d need a cylinder that’s the height of one string of LEDs, if we imagine a tree made up of all LEDs. But that’s not the case, apparently.
If I were you, I’d have no animations crossing the boundaries between the levels. Then you’d just have to worry about the distortion involved within each level.
Actually, I’m concerned that whoever designed this system hasn’t seen how it maps to a cone. In my opinion, the diagram should not look like a ziggurat, but like a series of warped, concave-upward surface developments. Sort of like this (https://www.hsecc.co.za/posters/inter2.pdf ) but sideways, u-shaped. Imagine each level, with the ends warped upward. But then that probably could not be mapped cylindrically.
Anyway, I definitely think you’ll need to “rehearse” the mapping with a 3D cone.
-
Corner Pin.
-
Of course, that makes perfect sense.
It may be possible to test the image-to-LED mapping projection process by mapping test images to a cone in a 3D app. Flat projection? Which axis? That will determine what the movie looks like: a circle? A wedge?
So I echo Kevin’s question: how are you mapping the image to the LEDs?
-
Do itashimashite!
(^_^)