MPM Solver effect Houdini
Solaris & Karma
This project was created using the new MPM solver introduced in Houdini 20.5. It was developed in SOPs, look dev built in LOPs, and rendered using Karma on GridMarkets.
​
Explore the in-depth explanations below to uncover the details of the process or DOWNLOAD THE SCENE and try it yourself.
by Manuela Cárdenas - Nov 2024
Breakdown
The goal of this project is to simulate various objects with different physical behaviors and achieve a striking visual effect:
-
A gelatinous behavior (the dice)
-
Particles
-
Secondary smoke
There are various ways to achieve this effect in Houdini, but the features of the new MPM Solver were key in selecting it for the project. Its ability to process different materials with the MPM source, along with its refined handling of collisions between elements, were crucial factors.
​
Share this case study
MPM in Houdini 20.5
​This solver can use multiphysics and is an extension of the FLIP Solver to solid mechanics.​
The MPM solver can be configured using three additional nodes: MPM Source, MPM Collider, and MPM Container.​
The range of possibilities it offers for configuration and project planning is impressive, and this case study was designed to explore those capabilities.
Initial MPM Test Flipbook - Frame of the first test with the MPM solver.
​
Slow Motion
​To highlight the collision and simulation capabilities of the MPM solver, a slow-motion effect was applied. This approach allowed all elements to be observed in detail while enhancing the visual impact.
To achieve this effect, the entire project was simulated at the required slowest speed, and the initial frames were later accelerated to create the desired transition.
​
MPM Test Flipbook - Test frame generated using the simulation performed with the MPM solver.
​
Secondary Smoke Simulation
To enhance the level of detail, a secondary smoke simulation was added. This simulation is directly influenced by the primary MPM simulation and the surrounding environment.
​
Secondary Smoke Simulation
Frame from the flipbook showcasing the secondary smoke simulation.
​
Secondary Smoke and MPM Simulation Flipbook
Frame from the flipbook showcasing both the secondary smoke simulation and the MPM simulation.
​
Karma and Solaris
The effect was done in SOPs, the look dev was in LOPs, to leverage Solaris lighting and was rendered using Karma CPU.
​
Viewport Scene Setup - Screenshot of the scene setup displayed in the viewport.
​
Optimization
Upon completing the project, we explored several strategies to optimize both rendering time and quality, aiming to achieve the best possible results in the most efficient and cost-effective way.
​​
​After conducting several tests and trials, the render farm proved to be a strong advantage in terms of both time and cost.
Render Layers - The three layers into which the final render was divided, along with their details.
​
Several strategies were implemented for optimization, including:
-
The slow-motion effect was incorporated directly into the Houdini file to render only the necessary frames, eliminating the need to discard frames later due to velocity changes.
-
The render was divided into three layers, each with its own AOVs. This allowed for precise adjustment of render settings and frame counts for each pass, distinguishing between dynamic and static elements.
-
Multiple Render Geometry Settings nodes were used to efficiently fine-tune render parameters.
​​
Compositing
Compositing was performed using Nuke. The three layers were combined to replicate the effect that would have been achieved if rendered in a single pass.
As an additional step, an extra sequence was created to streamline this process.
​
Auxiliary Compositing Sequence - An additional sequence created to streamline the compositing process.
​
Benefits of using a render farm
Utilizing a render farm for the final rendering significantly impacted both the time and quality of the project.
​
The difference between the estimated rendering time on a local machine and the time required on the GM 4800 Standard plan is striking. Not only is the rendering time significantly reduced, but the use of a render farm also allows the local machine to remain available for other tasks while rendering occurs. This ensures that work continues uninterrupted, enabling multitasking without any issues.
Additionally, the ability to calculate the estimated cost and time directly on the web platform enables efficient planning.
For this project, the estimation was based on rendering two frames: an initial frame (18) and a final frame (470). Since the camera followed a linear path and the on-screen percentage of elements in the frame remained consistent, we were able to estimate the rendering time and credits with a high degree of accuracy.
​
Render Comparison Table - Table comparing rendering times between a local computer and a render farm machine
​
​Manuela's interest in render farms grew while preparing her second submission to The Rookies Awards, seeking solutions for more advanced Houdini simulations. She chose GridMarkets for its reliable services and positive feedback from other artists.
​
GridMarkets, initially impressed by her talent, invited Manuela to join a larger collaborative project after seeing the quality of her work, guided by Adam Ferestad, Senior TD at GridMarkets: "Acrylic Paint Effect Houdini: 85M particles".​
Manuela is excited to take on new projects and explore exciting career opportunities. She remains committed to refining her skills through personal work and is passionate about advancing her professional journey. To connect with her or view her past projects, you can visit her LinkedIn profile.​​
Manuela Cárdenas and GridMarkets
Manuela Elena Cárdenas de la Miyar is a Junior FX Artist based in Seville, Spain.
In 2023, she completed a Master’s in VFX and Compositing at Animum Creativity Advanced School in Málaga, Spain. She also holds a degree in Fine Arts from the University of Granada. Manuela was selected for The Rookies Awards' Drawn Selection in both 2023 and 2024, earning an Industry Rank A for her exceptional skills.
By: GridMarkets marketing
​