Bubbles and flotsam : https://youtu.be/xMcZFqszD3E
Random flotsam:https://youtu.be/sVH_2X2CEs8
Aim: To test surface foam implementation and bubbles
as a macro scene.
Method: I created a sphere in Blender (initially it was
a full dome, but found a half dome bubble worked
better through testing) and then created a bubble
material following tutorials online by Coreb Games.
When spawning the particles I noticed they spawned
on top of each other, creating glitches with the particles
translucency. I then followed a tutorial by CGHOW on
physically-based dynamics in Niagara which allows
the particles to have rigid body dynamics - they can
touch but overlapping is unlikely. This worked well, but I
wanted the bubbles to clump together and deform with
soft-body dynamics when they touched. I tried to do
this through creating a material which distorted to the
distance field of other objects, but realised this wouldn’t
work because particles don’t have distance fields, so I
ended up abandoning this and animating fewer bubble
particles. I then animated the bubbles using a vortex and
attraction force to keep them rotating.
Reflection:
I’m happy with the result of the bubbles test, and I’d love
to implement the bubbles clumping together as this
could really bring another element into the scene. I think
they create a good sense of focus. Having this smaller
scale element in the world adds to a realism in the scene
as a whole and I think it allows the viewer to imagine a
larger world as it contains large-scale objects like clouds
down to small bubbles. The shift in focus in scale creates
a much better break than a surprise splash or foam
movement can. I still need to tweak the motion of the
bubbles so that they pop and float to the surface, which
I will do for Surface.
Aim: testing spawning random flotsam to sustain
interest and surprise.
Method: Using the previous particle system spawning
mesh objects, I added multiple objects and bubble
particle systems. I then gave them different likelihoods of
spawning and some randomization on scale and spawn
placement. Each time it spawns a different arrangement
of particles would be visible. This test was inspired by
Mountain by David Oreilly in which new objects hit the
mountain at random times as you play the game.
Reflection: I realised from an overnight live stream test
I did of the simulation (link)(link day) that the objects
over time also lose their novelty and ability to sustain
interest for the viewer - they just loop their behavior, not
changing or updating enough to retain interest.
I decided to test adding new models instead of giving
the old models different forces to act upon them as
I felt a new object might provide more novelty than a
new movement for an old object. The iteration was
successful in adding new interest with new objects. I
think now is the time I would need to figure out how they
would update, how their movement is changed over
time, or how they sink and are replaced by new objects.