Cinematic Visualization of Multiresolution Data: Ytini for Adaptive Mesh Refinement in Houdini

TL;DR

This paper presents the general methodology for the import and visualization of nested multiresolution data sets into commercially available visual effects software, and describes a method for using Houdini to visualize uniform Cartesian data sets.

Abstract

We have entered the era of large multidimensional datasets represented by increasingly complex data structures. Current tools for scientific visualization are not optimized to efficiently and intuitively create cinematic production quality, time-evolving representations of numerical data for broad impact science communication via film, media, or journalism. To present such data in a cinematic environment, it is advantageous to develop methods that integrate these complex data structures into industry standard visual effects software packages, which provide a myriad of control features otherwise unavailable in traditional scientific visualization software. In this paper, we present the general methodology for the import and visualization of nested multiresolution datasets into commercially available visual effects software. We further provide a specific example of importing Adaptive Mesh Refinement data into the software Houdini. This paper builds on our previous work, which describes a method for using Houdini to visualize uniform Cartesian datasets. We summarize a tutorial available on the website www.ytini.com, which includes sample data downloads, Python code, and various other resources to simplify the process of importing and rendering multiresolution data.

Figures (14)

• Figure 1: Rendering created in Houdini combining multiple datasets - a volumetric solar interior simulation miesch, a solar surface simulation rempel, solar rim imagery sdo, and background star imagery allsky - along with a statistical core and streamlines traced through the solar interior volume, using the Houdini software. This image is not rendered with the multiresolution workflow described in this paper but is intended as an example of features in Houdini.
• Figure 2: Two different sets of opacity and color maps are applied to the metallicity field in a dataset from oshea2015. Left: A simple opacity map changes linearly from 0 to 1 across all data values (top left). Similarly, a simple color map changes from black (RGB 0,0,0) to white (1,1,1) (center left). These are the default opacity and color maps provided by Houdini. Right: A more complex opacity map (top right) and color map (center right) can change between many values and colors, and use various interpolation schemes. As with Figure \ref{['fig:sun']}, this imagery is an example of a feature in Houdini and is not rendered with the multiresolution workflow described described in this work.
• Figure 3: An emissive volumetric dataset from oshea2015, where intensity and wavelength of the emitted light is adjusted to highlight different aspects of the data. Both images show the same six variables from a similar camera position -- HI density, HII density, temperature, photo gamma, metallicity, and stars. The left image highlights regions of high temperature in orange while dimming the other attributes and removing their color to maintain context. The right image highlights areas of high metallicity and dims the rest. As with Figure \ref{['fig:sun']}, this imagery is an example of a feature in Houdini and is not rendered with the multiresolution workflow described in this work.
• Figure 4: Description of a raytracing process, shown with the dataset from enzo2013 that is used in the online tutorial. Each ray starts at a single camera position, passes through a 2D plane of pixels which will be the resulting image, and traverses through a translucent volume. Points are sampled through the volume along the ray. Each sample point has certain characteristics, such as color, opacity, and illumination, which are described by the shader. The result of the combination of these sample point characteristics is placed at the pixel on the 2D plane through which the ray passes.
• Figure 5: Two renderings of the density field of the dataset from wise. The image on the left shows edge effects resulting from placing volumes of different voxel sizes next to one another without further adjustment. The image on the right shows our method of data preprocessing and shading.