The animation tests that I intend to practice, which I did not cover in the previsualisation, include the following:

1. The umbrella drop using dynamics
   • The prop, held by the character, will be hidden when he lets go of it. Then a duplicate will be made visible with gravity rigid bodies and passive colliders applied.
2. The disappearing apple pie.
   • I intend to make the apple pie disappear (perhaps even more items on the kitchen table too), in a mirage/ dream-like effect. This will occur when the apple pie is replaced with the vase.
   • When completing my previs, I did research and experiment somewhat with what the options to complete this effect were, and I concluded that I may have to render this section in layers with transparent backgrounds. Then in post-production, I would add the effect to only the apple pie layer.
   • In opposition to these experiments, I have researched further and found other options for this animation using nparticles dissolve.

1. Umbrella Prop Drop

The method I used to create these tests were through fields and solvers> active and passive rigid bodies. An active rigid body for the umbrella and a passive rigid body for the colliding surface (the floor and walls).

I edited the initial velocity and spin to replicate the character throwing the umbrella. However, this sometimes-produced strange effects and the active rigid body kept spinning infinitely, without any gravity. Meaning, that the initial parameters that were applied to the active rigid body did not cease after the initial release. This is incorrect, and I could not solve this issue until I restarted the test.

I managed to find methods and settings that somewhat worked, as you can see in the video. Though there is plenty of room for refinement and improvements.

2. Dissapearing apple pie

My intended effect for the apple pie’s disappearance is a mirage or being blown away in the wind effect. The method I used to create this effect (at first) was with nparticles, to dissolve the object with particles that blow away in the wind. The video below contains the method I followed.

I chose to complete this experiment quickly, in order to understand the technical method and produce a working outcome. Considering I have used a representative model for now, I will have to complete this test again. Therefore, while this practical experiment was quicker, the final outcome will be much more refined now that I have gained greater understanding of the method.

The turbulence seems to affect the model movements, unlike in the tutorial. While I would like some turbulence in my particles, I do not want the model to move at all.

Much later in my workflow, I encountered issues with this method. I repeated the nparticles method and left the keyframing of the timing until later so that I could align it properly with the vase placement animation on the table. However, when this point came about, the alpha animation that made the texture opacity disappear refused to work for some reason. After much deliberation and experimentation, I abandoned this method. Whilst attempting to fix this issue I tried:

• Restarting the method in a new Maya file.
• Opening an earlier version.
• Breaking some/or all connections with the texture and reconnecting them.
• Assigning a new texture and making the texture connections again.
• Using remote access to complete the tests previously listed instead of my PC.

None of these worked, therefore the new method I began experimenting which involved Hypershade nodes in the texture graph network. ‘Ainoise’ to create the disturbance in the texture, which connects to the Aistandardsurface geometry opacity. And ‘Airange’ which controlled its application in terms of amount that could be key framed. Airange allowed me to control the intensity of the dissolving effect, wherein I could keyframe a gradual disappearing effect completely up to my own will.

The below video consists of the outcome from my application of this method onto my chosen objects.

The hypershade graph network should look like this: