The IDIA Vislab team has published a paper on their immersive visualisation work. They describe a software package they are busy developing: The iDaVIE software suite currently under development reads from both “volumetric data cubes” and “sparse multi-dimensional catalogs”, rendering them in a room-scale immersive environment that allows the user to intuitively view, navigate around and interact with features in three dimensions.
The expressions in quotes above refer to the fact that this new software can handle both volume and catalog data. This is an essential challenge of big data visualisation. Indeed, the nature of the data is very different. Volumetric data is something that typically has a value everywhere, like temperature. And you have to render that value in every point of the space you are showing. Catalog data, for example a catalog of galaxies, will have sets of coordinates for each galaxy. Not only where they are located in the space you are showing, but also, say a relative size, brightness, colour, etc. And the computer has to render that data too,
As you can imagine, it is very different to paint a whole sky with nuances or a specific set of points with different values each – the approach to build the visualisation seamlessly of both types of data is very different from the computer’s point of view, just like a painter would use very different technique to paint a sky or the stars in it. In visualisation, however, there is an additional challenge: speed. This is very computationally intensive but for the visualisation to be navigable, the computer creating it needs to be fast – like in immersive computer games, In fact, there’s a reason why the technologies used for this are the same as for gaming: GPUs (graphics-specialised computer chips), software development frameworks for game development, and VR headsets for example.
Visualisation is even more challenging with the big data sets that we are working with. That’s why this software is so innovative. It can render both volumetric and catalog data, and work with the latest VR technology to provide an immersive big data interactive exploration and visualisation. To top it up even more, this ability to visualise big data is not limited to VR technology. The software is even able to build the visualisation for large digital domes, with several projectors working in unison to orchestrate a full dome experience.
So hats off to the IDIA vislab team!
References:
Alex Sivitilli, Angus Comrie, Lucia Marchetti, and Thomas H. Jarrett
Virtual Reality and Immersive Collaborative Environments: the New Frontier for Big Data Visualisation
ADASS 2019 Conference Proceedings
https://arxiv.org/abs/2103.14397
Lucia Marchetti, Thomas H. Jarrett, Angus Comrie, Alexander K. Sivitilli, Fabio Vitello, Ugo Becciani, A. R. Taylor
iDaVIE-v: immersive Data Visualisation Interactive Explorer for volumetric rendering
ADASS 2020 conference proceeding
https://arxiv.org/abs/2012.11553
T.H. Jarrett, A. Comrie, L. Marchetti, A. Sivitilli, S. Macfarlane, F. Vitello, U. Becciani, A. R. Taylor, J.M. van der Hulst, P. Serra, N. Katz, M. Cluver
Exploring and Interrogating Astrophysical Data in Virtual Reality
https://arxiv.org/abs/2012.10342
iDaVIE documentation: https://idavie.readthedocs.io/en/latest/
Image: T.H. Jarrett & al. Exploring and Interrogating Astrophysical Data in Virtual Reality