3

I hope this is not a vague question.

I would like to know if any researchers and students in this community have used or considered using 3D graphics software (such as Blender) or a game engine to visualise scientific data as opposed to specialist software. I understand there are many limitations to using popular 3D graphics for scientific purposes (such as accuracy of the simulation). Despite this, I have found studies by researchers using these resources to visualise their data (here's an example http://www.nature.com/articles/srep04228), and I am guessing that the existence of free and open-source software can be a benefit to researchers by being a cheaper alternative to hiring visual research groups or developing specialist programs, and may be quite easy to learn by comparison. Please correct me if I am wrong on that.

I would like to know if anybody here has ever made a decision to use 3D graphics software or even a game engine in their workflow, and that might be for the reasons above or entirely different ones. Furthermore, what were the limitations as well as benefits you found from using software not developed for scientists in mind?

Thanks

Re: I realise this question is dangerously close to asking for opinions. But if it is possible to answer factually by being based on experience, that would be great.

Toby Jake IB
  • 31
  • 3
  • Welcome to SciComp.SE. Can you be more specific about what do you mean with visualization? For example, I have used some. And regularly I use Inkscape (and have used Adobe Illustrator in the past) in some visualizations. – nicoguaro Jan 14 '16 at 23:02
  • Hello. What I meant was the sort of visualisations you may find in the molecular animation community: using data from simulations or experiments that has been imported into non-standard software like 3D, game engines, or other digital media (often with scripting if need be). Although I've seen examples of this outside the biochemistry field, such as geology too. – Toby Jake IB Jan 15 '16 at 17:24

2 Answers2

3

It seems you have answered your own question. As another example, computational astrophysicists have used Blender and Sketchfab, for instance, to render simulation output; see e.g. this page in the yt docs.

David Ketcheson
  • 16,522
  • 4
  • 54
  • 105
1

There are well designed, open source software packages for scientific visualization. In particular, there is the VTK toolkit, and the two major contenders for scientific visualization -- Visit and Paraview -- are both built on top of it. Given that both are free, there is really no reason for anyone to look for other tools that weren't specifically intended for scientific visualization.

That said, some software can at least interface with other packages. For example deal.II has a backend to generate Povray files that can then be rendered via raytracing. (Disclaimer 1: I'm one of the authors of deal.II. Disclaimer 2: I don't think anyone has tested the povray backend in the better part of a decade.)

Wolfgang Bangerth
  • 55,373
  • 59
  • 119
  • That's a very good point to make. While I have seen examples of scientific visualisation with non-standard software packages, I haven't seen many outside specialist interest. Obviously software like the VTK toolkit are probably the first things a researcher would choose if looking for open source tools. – Toby Jake IB Jan 20 '16 at 13:24