First Renders

Here is a first pass at comping. We are rendering the dust. The ship is a quick, low detail render, but we wanted to throw everything we had so far together to start our compositing.

Wire Sim

Hey guys, here is a flip book for the wire sims.

Next I will be finishing up the texturing so that we can light and render the ship. The thrusters have been completely rendered as well as the environment and will make it to the blog as soon as we can get the ship rendered.

Sand simulation

So last week was a bit slow. Joey and I were preparing for teaching a class at Dreamworks. Today, we got started again in earnest. I spent the day finalizing the sand simulation. With my initial sand simulation, it only required bringing in the proxy collision geometry and tweaking the scale.


It should be noted that in this flipbook there is no environment geometry, so the particles behind the ground are visible. It's kind of interesting to see the FLIP fluids fall of in sheets, but in the render they will not be visible. Joey did some great sims for the wires hanging off the ship. Next is going to be a small particles sim for the small rocks getting tossed around by the engines.

I'd like to do the smoke simulations, but unfortunately we don't have enough space yet. After finishing the smoke sims, I need to light the ship and do shadow passes for the ship. Luckily we can just bake out the depth-map shadows for the key light and use it in all of our smoke, sand and rock renders.

First pass dust simulation

Now that our landing animation is final, I've worked in the new animation with the dust simulation and dust sourcing. In this simulation I'm using some very low resolution proxy geometry for the dust to collide with. Each dust simulation in the above video represents the dust kick-up from an individual engine. These are low resolution simulations, just so I could kick them all out today. When I increase the resolution for the final sims I might also add in some vorticles. I haven't decided if they're necessary yet.

In the final comp we're just going to layer these dust simulations together, in conjunction with a FLIP sand simulation and a particle sim for some small pebbles and rocks.

We're also lacking space right now. Each of these simulations will require a significant amount of data, as the shot is 600 frames long. My estimation of our necessary disk space is somewhere in the range of 500GB to 1 terabyte between the two shots. The next step for me is simulate the wires of the hanging cables attached to the ship.

Engine Test

We have gotten our final animation done, besides some small details on the ship, and I have thrown in the thrusters tool that we created to see how it would fit. I am  very pleased with the initial test and am excited to see it with the completed animation.

Sand RnD

Today I started working on a FLIP fluid simulation of the sand. The main goal was to get rivulets of sand that slide across the ground in interesting ways. This is just a test with test terrain, but ultimately the sand will interact with with the rocks and ground.

The set-up is a simple point cloud with a lot of outward velocity, combined with a field force DOP that grabs some swirling vectors from SOPs.

Updated Concept Art

Here is an updated concept painting that illustrates some of the new elements we will be implementing into our project.

Dust RnD

I've been working on the simulation for the dust that gets kicked up by our engines. Joey posted one of the earlier tests and since then I've gotten a working version of our ship asset into the scene.



The dust sourcing system is created by raying a point onto the surface and then generating a series of concentric rings that radiate outwards from that point. This is accomplished with some simple vector math in a VOP SOP and a voronoi noise VOP.



The next step was to create a couple point clouds with vectors on them that could be used to control the dust simulation. I used one in the shape of a torus to create a rolling effect as the smoke travels outwards. I also attached a cone of vectors that point down and out to simulate the actual thrust of the engines.



Finally this information was piped into a pyro solver, along with a primitive collision proxy for the ship. The end result is nice for a first pass. More updates as we incorporate the final ship animation and final environments.

Dust kick up from thrusters

Travis has been working on the dust simulation from the force that the thrusters emit.

Color Corrected Concept

Laser Scanner Test

Here is a test of the laser scanning system that will be attached to the ship as it is landing. Glow and opacity will be added in comp to achieve realism.

Engine Test

Engine thrust achieved using a particle system that has metaballs copied onto it. The metaballs are then converted to a volume. Noise will be added to the particle velocity to give it a flickering effect on the tail of the blast.

Here is a shot with the matte painting in the background.

Landing Test

Here is a test to show off the ship and the ground terrain a bit.

Beginning

Hello,

This blog has been set up to document the creation of our visual effects shot Icarus. There are three of us working together on this project (Freek Hoekstra, Travis Harkleroad, and Joey kowalewicz). All of whom are current interns at Side Effects Software. The goal of this project is to not only complete a photo real shot, but to also bring what we have into the gaming environment. That being said, Icarus will consist of two shots where a spaceship is landing on some sort of foreign terrain that is made up of mainly volcanic rock and sand. There will be loads of effects to make these shots seem believable which we will be using Houdini to achieve. With Houdini's dop network and advanced particle systems we will be able to accomplish our effects in a timely manor. We will continue to update the blog as we push onward with the project.