W.I.N.D test Week 6

TESTS

W.I.N.D Tests.
(Wave Implementation =
New Discoveries)

The bulk of my testing this semester revolved around
more practical implementations of the particles
from semester 1 into the larger scene. These were:
getting the particles to respond to the waves, testing
how each particle system works together to make
a different dynamic scene based on the Beaufort
Scale, and trying to get these to implement with
weather. These tests, although mainly practical,
gave me insights into how the implementations
might create novel breaks in the work. By testing
the particles in the simulation itself, it revealed that
they responded to the wave movements in ways I
wouldn’t have anticipated. They created movements
which were unexpected and exaggerated invertedly,
achieving results similar to what I had originally
intended to explore in my W.A.V.E chart.
Later in the semester I became disillusioned with
using only foam motion to achieve what I had set out
to answer: how to create a dynamic real-time longform
experience through motion. I turned to looking
at macro scenes and time dilation to see if they
could achieve this.

Test 1.1: Floating Niagara Particles

Aim: Particles responding to real-time wave movement.

Method: I followed a tutorial by Ionized Games
to learn how to create a Niagara particle system that
followed water movement. The system utilised a custom
Module script which allowed the waves to manipulate
the height and velocity of the particles in the scene so
they would change height with the waves, but also be
pushed around by the wind direction that effects the
swell direction.

Reflection: The ability of this system to respond to wind
speed direction through the waves velocity would have
been ideal for the simulation, giving it not only the wave
responsiveness but also the wind direction. This wasn’t
possible to implement as it would cause Unreal to fail
every time I played the simulation. I never uncovered
why; I think it was too expensive an operation for my
computer to process, forcing Unreal to fail.

Test 1.2 Buoyant object – particle system

Aim: Particles responding to real-time wave movement.

Method: In this test I attempted to hack a method to
create a particle system which could move around the
scene moved by wave motion. I had created a buoyant
object which floats around the scene moved by wave
height and direction, and then attached a particle
system to it hoping this would emulate wave motion.

Reflection:
This worked well until the object was shot into the air by
waves causing the system to fire out of the water. This
was very unexpected – maybe an accidental test for my
non-normative motion, in that it moved in an unexpected
way. But I think due to the nature of the work the motion
didn’t look intentionally unexpected or controlled, but
more a glitch or a failure of the system. This made me
realise that I wanted the implementation of exaggerated
motion in the scene to be convincingly articulated. I
also realized that before implementing the foam into
the wave simulation, it was easier to judge exaggerated
foam motion as it was de-contextualised, but within
the simulation I felt if it didn’t move as realistically and it
failed to elucidate any surprise; it just felt poorly done.

Test 1.3 Material distortion using Gerstner Waves

Aim: Particles responding to real-time wave movement.

Method: I learnt this method from Sjoerd de Jong,
an Unreal technical artist and educator. In the tutorial
he explains a method of creating a custom material
function which distorts the world offset position of any
material to the Gerstner waves which generate the
waves in the simulation. I used this technique on the
foam to distort the mesh to the Gerstner waves.

Reflection:This technique was the best solution to have
the splashes distort to the wave height. I ended up using
this same technique to animate anything that needed to
conform to the ocean mesh. This is the same technique
used by John Gerrard in World Flag, and I was pleased
with this realisation as it meant I could play with similar
scenes later on.

Test 1.4 Material distort + buoyant object

Aim: Particles responding to real-time wave movement.

Method: In this method I combined the buoyant object
with the wave offset material by adding the material
function to all the foam materials in my scene then
adding the particle system to the buoyant object.

Reflection: This technique also worked well, creating
interesting results where foam appeared to float
between waves in a more realistic way. It failed when
a particle system couldn’t be rotated, such as when
using a flat plane it would rotate perpendicular to
the water. I think this could be controlled by locking
the object to the z rotation, not x and y. This was a
accidently successful non-normative particle motion,
but controlling this is would be difficult.
Revisiting this as I write this document it’s actually the
most succesful mapping of particles to the wave heights,
it maybe just a matter of timing the particle correctly and
it would work well.
As I update the wave heights live it causes the object
to spin making the foam float in these crazy serpentine
patterns above the waves, I initally wrote this off but its
worth revisitng for sure.