By the end of this week or the next I will release the second build based on my thesis code. It will NOT be the Matt Ebb’s because for thesis purposes I should only use my previous code but it will do include Dataset volumetrics in a primitive but working form.
Since I could not tackle on time the simulation obstacles and the simulator integration and due to the fact that Matt implementation is what will go into the main trunk I should start committing to it as soon as I finish the second build so this will be the last of my current implementation and latter everything will be with the “oficial “
Though it will not be the integrated one it do has few goodies and this release will serve as a gigantic backup (thesis shoul be very secured 🙂 ) and some day could serve to someone who likes to thinker code 🙂 , besides there’s out there few litle GNU 3D gases simulation implementation so this will be another one 🙂
Features it will include (added since first build):
1-“Correct” raytravel distance detection (had few issues regarding multiple intersecting volumetrics)
2- Adaptative sampling in the quality panel
3-Dataset volumetrics with full material integration (stack and nodes)
4-some bugfixes and little improvements
The simulator will be delivered in a separated file and I also will port it to a python script for more enthusiastic folks, that way you could not only load simulation dataset but other datasets (previously converted to the simple file format that blender will read) I also will public the file format.
Hope you will like it 🙂
ps: below some tests derived from the homemade violent fire algorithm.