Kalleheikki Kannisto
1
COW Coin
Forum Replies Created
-
Hi Pär,
You can find the common factors calculator here: https://www.calculatorsoup.com/calculators/math/commonfactors.php
Those are the only integers that divide exactly, and 60 is the largest one that goes through the center point in both directions. You can do 120 with 2 subdivisions, but the end result is virtually the same: just every other line is thinner.
Kalleheikki Kannisto
Senior Graphic Designer -
There might be an easier solution, but all I can think of right now is a “lookAt” solution. You would have a third null in the point where you wanted the brush handle to point towards and add a lookAt() to the rotation of second null (the one that moves in elliptical fashion). Unfortunately it’s somewhat tedious with 2D layers in After Effects…
I’m sure there’s a simpler solution involving the rotation value of the center null and a sine function, but can’t figure out the code in my head. Need more coffee.
Kalleheikki Kannisto
Senior Graphic Designer -
Any of these grid spacings will divide evenly to those two dimensions (common factors of 1920 and 1080):
1, 2, 3, 4, 5, 6, 8, 10, 12, 15, 20, 24, 30, 40, 60, 120
60 sounds the most useful to you, both vertical and horizontal grid lines go through the center point.
120 with 4 subdivisions isn’t too bad either.
Kalleheikki Kannisto
Senior Graphic Designer -
See the attached dummy project.
12867_rotatingtoothbrush.aep.zip
Kalleheikki Kannisto
Senior Graphic Designer -
Then there’s this. If I recall correctly, it uses Particular for the lines.
Kalleheikki Kannisto
Senior Graphic Designer -
It is actually very close to this problem here:
As long as it is square, you can get the scale from the difference of the bounding box and the original box directly, if it is a rectangle and you need to constrain proportions, it’s a bit more tricky. I think you’ll have to check the result in x and y direction for any given angle and pick the larger of the two.
Kalleheikki Kannisto
Senior Graphic Designer -
I managed to get this far. This is a trigonometric problem. We know where points O and P are, the don’t move. There are two lines that can be defined by being at a known angle and going through a known point. Alpha is the angle of rotation. The blue line goes through point P and is at angle alpha from the vertical (or 90-alpha from the horizontal). The red line goes through O and is at the angle of 45+alpha from the horizontal. Their common point R2 tells us where the corner of the scaled and rotated square is. The division of the distance of O to R1 (same as distance of O to P) and O to R2 is the scale.
So the basic problem is finding the coordinates where the two lines meet. Does someone want to take it the rest of the way?
Kalleheikki Kannisto
Senior Graphic Designer -
I would think adding a blue ambient light should do it.
Kalleheikki Kannisto
Senior Graphic Designer -
The muffled sound comes from losing the high frequencies. Interestingly enough, you can copy and paste parts of the spectral display. So you could try selecting the most similar-looking top section of the spectral display and copy and paste the top frequencies back in to the affected area. It may work. Since it will not necessarily be identical, it might sound a bit off. But worth a try nonetheless. Perhaps you can remove the offending frequencies afterwards. You’re trying to rebuild the full higher range without introducing new frequencies (notes).
Kalleheikki Kannisto
Senior Graphic Designer -
If you add a camera to the comp, you don’t get this issue. At least not in the inner comp. I guess the “default” camera tries to adjust the view to accommodate for the change in comp dimensions, while an actual camera layer retains its viewpoint.
Kalleheikki Kannisto
Senior Graphic Designer
