Wavelet Turbulence

from Wikipedia, the free encyclopedia
Wavelet Turbulence

Wavelet Turbulence Explosion.tif
Wavelet Turbulence Explosion
Basic data

developer ETH Zurich & Cornell University ( Theodore Kim , Nils Thürey , Doug James & Markus Gross )
Publishing year 2008
category 3D graphics software
License GPL ( Free Software )
cs.cornell.edu/~tedkim/WTURB

Wavelet Turbulence for Fluid Simulation is an open source algorithm with which explosion and liquid simulations can be made more heterogeneous and detailed after they have been calculated by the simulation tool. It is available as stand-alone software , contained in other programs (e.g. Blender ) and was published as a scientific publication .

Functionality

In addition to the resolution, which is increased by machining, the heterogeneity of the simulation can be influenced.

Postprocessing simulations with Wavelet Turbulence has the advantage over the directly detailed calculation of the course of a simulation that the approximate course of a simulation can be calculated at high speed, whereby when creating CGI effects it can be assessed whether the simulation with the scenery, In which it is to be inserted, the fact that the algorithm calculates details more quickly and that it makes it possible to calculate different points of the simulation with different levels of detail is harmonized.

functionality

The algorithm determines the UV coordinates of the raw version of the simulation without instructions and deforms them on the basis of two textures that it generates.

In the first step, details with low spatial resolution are added to the simulation. The Wavelet Noise algorithm is used for this. Spiral currents are added to the texture, which emphasize individual points. These have the shape of the cloud formation of a low pressure area .

Details with a high spatial resolution are calculated using a method that comes from the anisotropic noise algorithm . In contrast to this, Wavelet Turbulence calculates its position on the basis of the spatially poorly resolved currents.

With the help of its functional matrix , the algorithm can connect several streams.

Its pipeline is structured as follows:

Wavelet-Turbulence-Pipeline.png

history

The technological approach to improving computer simulations through wavelet transformations comes from Jos Stam's 1999 book Simulating the Effects of Turbulence on Flexible Structures . Wavelet Turbulence comes close to the Anisotropic Noise algorithm , which makes it possible to add coarse-grained details to simulations. This is integrated in the independent Wavelet Turbulence program as well as in the implementation of this in Blender.

The ETH Zurich decided to publish the article under the GNU Public License, because royalties have slowed the spread of the algorithm and because 25,000 euros would have cost a patent that is valid for 30 months. According to Markus Gross, a university like the ETH can only afford this if the prospect of success is extraordinarily high. After Wavelet Turbulence had developed into the industry standard, the developers criticized that the use of the algorithm was only rarely referred to by the studios. In many films there should have been no mention of the use in the credits. Although the GPL allows this, the practice is viewed as questionable by various quarters. The developer Nils Thürey commented on this in the mirror : "We underestimated how useful this tool would be."

Use in film projects

The algorithm was used, among other things, to implement the following projects:

2009:

2010:

2011:

2012:

2013:

Awards

After Wavelet Turbulence was used in many productions by Hollywood studios, the developers received the Technical Achievement Award of the Academy of Motion Picture Arts and Sciences for the algorithm in 2013 .

Web links

Commons : Wavelet Turbulence  - collection of images, videos and audio files

Individual evidence

  1. ↑ a b c Wavelet Turbulence Source. In: cs.cornell.edu. Retrieved May 29, 2016 .
  2. ↑ a b Frank Thadeusz: Film tricks: Steam, smoke - and rum! In: Der Spiegel . No. 45 , 2015 ( online ).
  3. ↑ a b c Smoke Domain. In: Blender Reference Manual. Blender Foundation, accessed June 3, 2016 .
  4. ↑ a b c d Markus Gross, Nils Thurey, Doug James, Theodore Kim: Wavelet Turbulence for Fluid Simulation . (PDF) Zurich 2008
  5. ↑ a b c Dennis Buehler: http://www.zeit.de/2013/07/Rauchsimulator-Markus-Gross-Oscar Special Effects: Oscar for the magician . In: The time . No. 7 , 2013.
  6. ↑ Fig. 1 by University of California: The Science Behind Hollywood Explosions. September 15, 2015, accessed June 11, 2016 .
  7. ^ Wavelet Turbulence for Fluid Simulation. In: cs.cornell.edu. Retrieved June 10, 2016 .
  8. ^ Peter Rüegg: Oscar-worthy smoke signals. ( Memento from November 6, 2016 in the Internet Archive ) In: ETH Life. January 8, 2013.
  9. ↑ The 85th Scientific & Technical Awards 2012 | 2013. In: Oscars.org | Academy of Motion Picture Arts and Sciences. Retrieved June 3, 2016 .