Applications of Houdini for Cinematic Scientific Visualisations
Abstract
SideFX’s Houdini is a software used mainly by artists in the visual effects (VFX) industry, it has been picked up by many scientists for creating visualisations for scientific communication and educational purposes. This form of visualisation is also known as cinematic scientific visualisation. This essay discusses and outlines some of the primary reasons for the extensive use of Houdini in academia. Three main reasons are highlighted for the reason why Houdini’s use is justified for cinematic scientific visualisation: (i) Houdini is a node-based software package where it promotes a non-destructive workflow for users. (ii) Data-driven software means it can handle large datasets, which is often the case for scientific data. (iii) Lighting, camera and rendering features that traditional scientific visualization tools lack. The essay also notes some of the potential drawbacks of current definitions of Houdini in scientific visualisation.
Research question
Critically evaluating the applications of Houdini outside of Film/TV, especially in the field of scientific visualisation.
Keywords
Cinematic Scientific Visualisation, Data Visualisation, Simulation
Introduction
This essay attempts to postulate that Houdini holds a significant role for cinematic scientific visualisation. As there are different definitions for visualisations in the science community, this essay will only focus on the role of Houdini in the field of cinematic scientific visualisation. This essay highlights three justifications where Houdini shows promising viability in communicating research analysis. Even though, Houdini excels at visual effects, it is noted that there are still some challenges for Houdini to be fully adapted into the production process of creating cinematic scientific visualisations
SideFx Houdini Software
Houdini was developed by SideFx and it is well known for its capabilities in high-end visual effects. In the film, gaming and advertising industry, it is one of the most extensively used computer graphics software. Due to the continued development by SideFx, Houdini excels at simulating rigid bodies, fluid dynamics and particle systems (Aguilera and Johansson, 2019).
Another feature of Houdini that sets it apart from other 3D computer graphics software is its node-based workflow. This workflow facilitates flexibility and efficiency in altering geometries and simulations. This flexibility allows for the creation of custom tools or node networks which can be easily redeployed to different scenarios without the need for building the setup from scratch.
Literature Review
Scientific visualization, also known as data visualization, involves the use of graphical techniques to present scientific data in a way that enhances comprehension, interpretation, and insight. It is widely used in various scientific domains such as astrophysics (Burkhart et al., 2020), biology (Sener et al., 2021) and remote sensing (Wróżyński, Pyszny and Wróżyńska, 2024). However, it is important to distinguish between scientific visualisation and cinematic scientific visualisation. As visualisation caters to different groups of audiences, the former generally refers to graphics presented to other fellow academics. Traditional science visualisation is usually done with science visualisation tools such as VMD which cantered around accurately visualizing molecular data for scientific purposes, with a focus on clarity, real-time performance, and scientific accuracy (Humphrey, Dalke and Schulten, 1996). Cinematic scientific visualisation’s target audience is commonly made for the general public with no prior background knowledge of the data presented to them (Borkiewicz, 2020, 1:24). Hence traditional scientific visualisation might not be suitable to effectively communicate the information to its audience.
The role of Houdini is crucial in cinematic scientific visualisation as part of the pipeline for producing content for science communication. (Sener et al., no date) stated that Houdini has become the main software application that helps to drive the production process for their film Birth of Planet Earth. Besides the field of astrophysics, Houdini is also often employed in the nature sciences. In Harrap et al. (2019) research, the estimated rockfall path is simulated on Houdini with the use of the Rigid Body Dynamics (RDB) simulation tools.
From the few of the aforementioned examples, it is hence important to understand the importance of Houdini and discover what other features which Houdini might lack which could possibly pose challenges for cinematic scientific visualisation.
Complex Lighting and Rendering Features
As Houdini’s main target audience is for VFX artists to create photorealistic content, its lighting and rendering focus on high-fidelity outputs. Lighting system inside Houdini allows the user to have much freedom and flexibility to art-direct the scene. For example, many different light types are included in the software and lighting properties can be adjusted to a large extent to achieve the required look (Borkiewicz, Naiman and Lai, 2019). Besides lighting, Houdini also supports high-end render engines like Karma or third-party renderers like Redshift or Octane. Overall, the workflow of lighting, shading, rendering and post-production produces high quality realistic imagery. With the use of datasets from scientific research, visuals can be improved with Houdini following a similar pipeline.
Specialised Simulation Tools
Apart from lighting and rendering, simulation tools in Houdini works very well for visualisation as well. As scientific simulations often deal with phenomena similar to visual effects in the movies. Houdini’s simulation features are adept at handling the complexities required by academics. For instance, Houdini contains fluid simulation solvers using the Fluid-Implicit-Particles (FLIP) method to model the behaviour of liquids with high precision (SideFX, no date). It is able to capture the nuances of fluids such as ripples and viscosity variations. For example, researchers are able to utilise Houdini to calculate and model ship’s motions based on the ship wake patterns simulated with Houdini (Ahmadibebi, Jones and Shirkhodaie, 2022). Furthermore, the non-destructive workflow in Houdini allows researchers to easily build visualisation variations by adjusting and fine-tuning the initial starting conditions. These unique features of Houdini are crucial for scientists to iterate more efficiently and produce quality results to be communicated to their respective audiences.
Handling Large Datasets
Data handling is another area in which Houdini excels for cinematic scientific visualisation. Houdini is able to efficiently manage complex and large data sets with ease. Furthermore, the software supports various different data formats. It can handle simulations with point counts in the millions and billions (Agrotis, 2016). Houdini also allows for customisation through the use of scripting and programming plugins to interface with existing other tools and libraries for better connectivity within a larger scientific data ecosystem. For instance, Borkiewicz et al. (2018)’s work contains custom C++ plugins for Houdini for their cinematic scientific visualisation work. Hence, these capabilities make Houdini an invaluable tool for managing, analysing, and visualizing scientific data.
Drawbacks and Challenges
Floating Point Precision
Although Houdini is suitable for cinematic scientific visualisation, there are still some drawbacks of the software to be considered. Houdini operates with 64-bit floating-point precision to handle calculations and simulations in conventional Cartesian space. Floating-point arithmetic introduces rounding errors due to the finite precision of floating-point numbers up to 64 decimal places. The implication of float point clamping in Houdini will ultimately cause inaccuracies in simulations. In the field of astrophysics, where simulation involves values both very large and very small, rounding errors can accumulate temporally leading to drastically different results (Izquierdo and Polhill, 2006). The resulting simulations will diverge from the expected behaviour or sometimes cause visual artefacts that significantly impacts the visual quality of the visualisation.
Non-Uniform Data Resolution
Another caveat would be the incompatibility of cross-software data resolutions. For instance, meteorological spatial data is often recorded in non-uniform resolution. This could be the case where only the significant areas are given priority in terms of resolution while lowering the level of detail (LOD) in arbitrary areas (Roesler et al., 2020). Volume data from such cases would not produce good results in Houdini as volume grids in Houdini expect voxels to be uniform. Even solutions to overcome such challenges can be problematic as secondary complications can arise. For example, matching the resolution to the finest data resolution comes with a high cost of resources needed to effectively simulate and render the data. Taking mean data resolution throughout the dataset can help reduce the computational cost but at the expense of inaccuracies from data loss.
Conclusion
In summary, this essay has covered briefly why it is at the forefront of the visual effects industry and is also growing in popularity as a visualisation tool for the scientific community. Three main reasons are also discussed as to why Houdini is feasible for scientific visualisation. Lastly, the essay cautions against some of the potential drawbacks of Houdini for scientific visualisation as it is still lacking in certain features. After all, each software has its pros and cons. It is important then for the user to balance between clarity vis-à-vis accuracy.
Bibliography
Agrotis, A. (2016) ‘A Fluid Implicit Particle (FLIP) Solver Built in Houdini’, Bournemouth University, NCCA [Preprint].
Aguilera, G. and Johansson, J. (2019) ‘Avengers: Endgame, a new approach for combustion simulations’, in ACM SIGGRAPH 2019 Talks. SIGGRAPH ’19: Special Interest Group on Computer Graphics and Interactive Techniques Conference, Los Angeles California: ACM, pp. 1–2. Available at: https://doi.org/10.1145/3306307.3328203.
Ahmadibebi, A., Jones, B. and Shirkhodaie, A. (2022) ‘Physics-based wake modeling for marine vehicles activity recognition based on simulated synthetic aperture radar’, in W. ‘Will’ Hou and L.J. Mullen (eds) Ocean Sensing and Monitoring XIV. Ocean Sensing and Monitoring XIV, Orlando, United States: SPIE, p. 25. Available at: https://doi.org/10.1117/12.2629032.
Aleo, P.D. et al. (2020) ‘Clustering-informed cinematic astrophysical data visualization with application to the Moon-forming terrestrial synestia’, Astronomy and Computing, 33, p. 100424. Available at: https://doi.org/10.1016/j.ascom.2020.100424.
Borkiewicz, K. et al. (2018) ‘Scientific and visual effects software integration for the visualization of a chromatophore’, in SIGGRAPH Asia 2018 Posters. SA ’18: SIGGRAPH Asia 2018, Tokyo Japan: ACM, pp. 1–2. Available at: https://doi.org/10.1145/3283289.3283324.
Borkiewicz, K. et al. (2020) ‘Introduction to cinematic scientific visualization’, in ACM SIGGRAPH 2020 Courses. SIGGRAPH ’20: Special Interest Group on Computer Graphics and Interactive Techniques Conference, Virtual Event USA: ACM, pp. 1–267. Available at: https://doi.org/10.1145/3388769.3407502.
Borkiewicz, K., Christensen, A.J. and Stone, J.E. (2017) ‘Communicating science through visualization in an age of alternative facts’, in SIGGRAPH Asia 2017 Courses. SA ’17: SIGGRAPH Asia 2017, Bangkok Thailand: ACM, pp. 1–122. Available at: https://doi.org/10.1145/3134472.3134488.
Borkiewicz, K., Naiman, J.P. and Lai, H. (2019) ‘Cinematic Visualization of Multiresolution Data: Ytini for Adaptive Mesh Refinement in Houdini’, The Astronomical Journal, 158(1), p. 10. Available at: https://doi.org/10.3847/1538-3881/ab1f6f.
Burkhart, B. et al. (2020) ‘The Catalogue for Astrophysical Turbulence Simulations (CATS)’, The Astrophysical Journal, 905(1), p. 14. Available at: https://doi.org/10.3847/1538-4357/abc484.
Chen, M. and Golan, A. (2015) ‘What may visualization processes optimize?’, IEEE transactions on visualization and computer graphics, 22(12), pp. 2619–2632.
Cinematic Scientific Visualization: Where Science Meets Hollywood Visual Effects (2020). Available at: https://www.youtube.com/watch?v=z4W6jnyPDyg.
Cox, D.J. (1988) ‘Using the supercomputer to visualize higher dimensions: An artist’s contribution to scientific visualization’, Leonardo, 21(3), pp. 233–242.
Cox, D.J. (2004) ‘The Art and Science of Visualization: Metaphorical Maps and Cultural Models’, Technoetic Arts, 2(2), pp. 71–80. Available at: https://doi.org/10.1386/tear.2.2.71/0.
Cox, D.J. (2006) ‘Metaphoric Mappings: The Art of Visualization’, in P.A. Fishwick (ed.) Aesthetic Computing. The MIT Press. Available at: https://doi.org/10.7551/mitpress/1135.001.0001.
Elkins, E.B. (2020) Simulating destruction effects in SideFX Houdini. East Tennessee State University. Available at: https://dc.etsu.edu/honors/524/.
Hakimshafaei, M. (no date) ‘An agent-based simulation of the wildfire deploying cinematic visualization’. Available at: https://users.soe.ucsc.edu/~pang/261/f21/projects/milad/CSE261-Milad.pdf.
Hansen, C.D. and Johnson, C.R. (2011) Visualization handbook. Elsevier.
Harrap, R. et al. (2019) ‘Our GIS is a game engine: Bringing Unity to spatial simulation of rockfalls’, GeoComputation 2019, pp. 11–14.
Horsley, D. and Stuart, P. (2019) ‘The rigid body and fluid dynamics of LAIKA’s “Missing Link”’, in ACM SIGGRAPH 2019 Talks. SIGGRAPH ’19: Special Interest Group on Computer Graphics and Interactive Techniques Conference, Los Angeles California: ACM, pp. 1–2. Available at: https://doi.org/10.1145/3306307.3328143.
Humphrey, W., Dalke, A. and Schulten, K. (1996) ‘VMD: visual molecular dynamics’, Journal of molecular graphics, 14(1), pp. 33–38.
Izquierdo, L.R. and Polhill, J.G. (2006) ‘Is your model susceptible to floating-point errors?’, Journal of Artificial Societies and Social Simulation, 9(4).
James, O. et al. (2015) ‘Gravitational lensing by spinning black holes in astrophysics, and in the movie Interstellar’, Classical and Quantum Gravity, 32(6), p. 065001. Available at: https://doi.org/10.1088/0264-9381/32/6/065001.
Jensen, E.A. et al. (2023) ‘Evidence-Based Methods of Communicating Science to the Public through Data Visualization’, Sustainability, 15(8), p. 6845. Available at: https://doi.org/10.3390/su15086845.
Kwatra, N. et al. (2010) ‘Fluid Simulation with Articulated Bodies’, IEEE Transactions on Visualization and Computer Graphics, 16(1), pp. 70–80. Available at: https://doi.org/10.1109/TVCG.2009.66.
Marshall-Colon, A. et al. (2017) ‘Crops In Silico: Generating Virtual Crops Using an Integrative and Multi-scale Modeling Platform’, Frontiers in Plant Science, 8, p. 786. Available at: https://doi.org/10.3389/fpls.2017.00786.
Maupu, D. et al. (2017) ‘Artist-driven crowd authoring tools’, in ACM SIGGRAPH 2017 Talks. SIGGRAPH ’17: Special Interest Group on Computer Graphics and Interactive Techniques Conference, Los Angeles California: ACM, pp. 1–2. Available at: https://doi.org/10.1145/3084363.3085035.
McCormick, B.H. (1988) ‘Visualization in scientific computing’, Acm Sigbio Newsletter, 10(1), pp. 15–21.
Medina, J.J. et al. (2020) ‘A rapid and cost-effective pipeline for digitization of museum specimens with 3D photogrammetry’, PLOS ONE. Edited by L.B. Porro, 15(8), p. e0236417. Available at: https://doi.org/10.1371/journal.pone.0236417.
Munzner, T. (2014) Visualization analysis and design. CRC press.
Naiman, J.P., Borkiewicz, K. and Christensen, A.J. (2017) ‘Houdini for Astrophysical Visualization’, Publications of the Astronomical Society of the Pacific, 129(975), p. 058008. Available at: https://doi.org/10.1088/1538-3873/aa51b3.
Qin, X., Rodriguez, F.C., et al. (2023) ‘A Single-cell Perturbation Landscape of Colonic Stem Cell Polarisation’. Available at: https://doi.org/10.1101/2023.02.15.528008.
Qin, X., Cardoso Rodriguez, F., et al. (2023) ‘An oncogenic phenoscape of colonic stem cell polarization’, Cell, 186(25), pp. 5554-5568.e18. Available at: https://doi.org/10.1016/j.cell.2023.11.004.
Roesler, E. et al. (2020) E3SM Water Cycle Visualization Project (Final Report). SAND–2020-10233R, 1668925, 690902, p. SAND–2020-10233R, 1668925, 690902. Available at: https://doi.org/10.2172/1668925.
Saario, V. (2019) Visual Effects in SideFX Houdini. Available at: https://www.theseus.fi/handle/10024/170112.
Schroeder, W., Martin, K.M. and Lorensen, W.E. (1998) The visualization toolkit an object-oriented approach to 3D graphics. Prentice-Hall, Inc.
Sener, M. et al. (2021) ‘Multiscale modeling and cinematic visualization of photosynthetic energy conversion processes from electronic to cell scales’, Parallel Computing, 102, p. 102698. Available at: https://doi.org/10.1016/j.parco.2020.102698.
Sener, M. et al. (no date) ‘An accessible visual narrative for the primary energy source of life from the fulldome show Birth of Planet Earth’.
SideFX (no date) FLIP Solver 2.0 dynamics node, FLIP Solver 2.0 dynamics node. Available at: https://www.sidefx.com/docs/houdini/nodes/dop/flipsolver.html.
Wróżyński, R., Pyszny, K. and Wróżyńska, M. (2024) ‘Reaching beyond GIS for comprehensive 3D visibility analysis’, Landscape and Urban Planning, 247, p. 105074. Available at: https://doi.org/10.1016/j.landurbplan.2024.105074.
Xu, K. and Campeanuy, D. (2014) ‘Houdini engine: Evolution towards a procedural pipeline’, in Proceedings of the fourth symposium on digital production, pp. 13–18.
yt Project (no date) yt Project. Available at: https://yt-project.org/.