To me, the top image looks the most interesting. The images are more attractive when there is not too much intersection of the lines.
You seem to be using periodic functions to adjust the direction. The regularity is quite visible, which is probably not what is desired. The "shock" idea was to set the direction randomly after a random time interval had elapsed. Alternatively, a randomly oriented impulse could be applied to the particle (resulting in a sudden change in direction but the new direction would be distributed near the old direction).
Yet another idea is to have different rotations of the vector field for different particles. E.g., most particles could follow the vector field exactly, but some fraction (say 20%) would move at a fixed 45 degrees to the vector field.
To me, the top image looks the most interesting. The images are more attractive when there is not too much intersection of the lines.
ReplyDeleteYou seem to be using periodic functions to adjust the direction. The regularity is quite visible, which is probably not what is desired. The "shock" idea was to set the direction randomly after a random time interval had elapsed. Alternatively, a randomly oriented impulse could be applied to the particle (resulting in a sudden change in direction but the new direction would be distributed near the old direction).
Yet another idea is to have different rotations of the vector field for different particles. E.g., most particles could follow the vector field exactly, but some fraction (say 20%) would move at a fixed 45 degrees to the vector field.