This past week, I worked on the main smoke simulation and started testing smaller particle simulations.

New model

One critique I received last week was that the phoenix's animation quality wasn't great. I found a different version of the same phoenix with different animations, so now there are a variety of animations without me having to change the animation myself.

Smoke tests

I spent most of this week on getting a general shape for my smoke. I watched a video called Intro to Houdini Pyro - Full Beginner Course by Voxyde VFX on YouTube. This video helped me better understand pyro and what the different parameters do. I also had guidance from our TI Sean Fitzgerald and Professor Fowler. This is what my setup looked like after watching the video and getting other assistance:

One struggle I had was getting the smoke to be dense enough and keep the shape of the phoenix. Since the original geometry was flat, I used a polyextrude to give the geometry thickness. This improved the density of the smoke.

Before polyextrude

After polyextrude

Once I had a good starting shape, I started adjusting other parameters such as the dissipation and turbulence. My reference leaves behind a long trail as it flies, so I tested different dissipation values. A dissipation of 0.375 leaves behind a shorter trail but keeps the phoenix defined. A dissipation value of 0.1 leaves behind a long trail, but the phoenix shape gets lost. I may need to combine smoke simulations to get the long trail and keep the phoenix defined.

Frame from reference

Flying

Dissipation: 0.375

Flying

Dissipation: 0.1

Gliding

Dissipation: 0.375

Gliding

Dissipation: 0.1

The smoke in my reference is also very swirly. I've begun a tutorial on cigarette smoke by Alessandro Pepe from Professor Fowler's website. Using my file cached smoke, I merged the converted cache to a VDB and merged the VDB velocities. I also scattered points along the phoenix geometry and used sin functions in the scale so that the geometry transforms over time to change the shape of the emission. My result so far doesn't look great because the particles stop, so I will continue to troubleshoot and follow the tutorial.

From Alessandro Pepe's tutorial

Current result from Alessandro Pepe's tutorial

I also tested additional particle simulations that will trail behind the phoenix. So far, I have a popnet with a curl noise in a popvop. 

Particle trail from popnet

Flipbook of combined smoke and particle results

My goals for this week are to get swirls in my smoke as well as the additional particle simulations. I will continue with the tutorial from Alessandro Pepe.