Thehardmenpath
Forum Replies Created
-
The shot with the spinal column you mean?
I always believed it was made with a bunch of dots put in 3d and a rotation of the camera with the center placed at the column sketch more or less. Only the dots are in 3d.
When the camera stops rotating with an ease in, the dots become a bit alive because the ones more near the camera seem to end a bit later. I believe no further effect is needed there.
It’s a nice tricky opening shot with a lot of fake 3d. I think the rotating brain consists of two 2d layers with opposite x movement. Clever and sweet.
-
It’s not necessary to translate every expression for every project. If you are going to work with large imported project with expressions in other languages, open the “Languages” folder in the AE 7 directory and run it in the language you want.
-
It’s not necessary to translate every expression for every project. If you are going to work with large imported project with expressions in other languages, open the “Languages” folder in the AE 7 directory and run it in the language you want.
-
Yep, you were right. I just thought it was shot from the top and the movement was irregular (that would make a curious effect by the way)
If the movement is (almost) constant, you just have to play with the rotation value until you get it right, it’s easy to make even if you don’t know the original rotation speed.
-
Yep, you were right. I just thought it was shot from the top and the movement was irregular (that would make a curious effect by the way)
If the movement is (almost) constant, you just have to play with the rotation value until you get it right, it’s easy to make even if you don’t know the original rotation speed.
-
If I saw the video probably that would help, but as I understand what you say… They probably tracked the rotation of the girl with two trackers (front and behind? left and right shoulders? must see the video) into a null object and then attached the camera to it.
-
If I saw the video probably that would help, but as I understand what you say… They probably tracked the rotation of the girl with two trackers (front and behind? left and right shoulders? must see the video) into a null object and then attached the camera to it.
-
Not really. It seems to use several particle cannons from the letters, but the letters don’t become particles, I think they dissappear behind the bubble maddness.
For a fast experiment, tweak your particle filter to make the bubbles from a point, then put that point on a letter that size with a mask. At the moment when most particles are created, decrease mask expansion.
-
That’s probably easy to do with a script, you can ask in aenhancers.com to try it.
Still, you probably don’t need to put them together for what you want. Close both masks (right clic on a vertex) and then set them both to add mode (will be by default). That will hopefully be enough in your case.
Another workaround comes to my mind, yet it’s not exactly what you would like to do: With both masks closed, menu/layer/autotrace
Uncheck apply to new layer and put the tolerance value as small as possible.Hope that works.
-
Each frame of interlaced video footage contains two different images taken in different moments. The reason for this comes from an old necessity for solving tv flicker. That’s why, when you see footage you may find that there are kind of horizontal lines where objects move, because even and odd lines belong to the first or second halfframe (field). Depending on the system, NTSC, PAL, etc, the odd lines may be before or after the even lines, that’s why you must sometimes tell the program whick the first field is, the “lower” or “upper”, or your exported move will flick annoyingly, regardless of having effects on it or not.
Additionally, you can choose “deinterlace fields”, also called “motion detect”. That will merge the two fields so that they will contain just one image each frame. Data of still objects merge and data of moving objects are vertically blured by one pixel. (there are more complex ways to do that). What interests you right now is the fact that if you rotoscope an object that has two different positions in the two fields, the mask won’t cut the object properly, as there will be only as much as a single mask for each two different fields. So the best choice for you probably is to choose that deinterlace option in the “interpret footage” window. That way, what you see when you put the mask there is what you get when you export.
There could be talked much more about this, but I think that will probably be enough for what you need to do.