Scientific visualization, sparse point set volume rendering

This is not modeling but I would appreciate your comments. It is a study on a fast and cheap pseudo volume-rendering technique for very sparse points sets. The image corresponds to a small sub-box (3 million light years side) cut from a large cosmological simulation. The bright blobs are dark matter halos where galaxies are expected to form. This is an extreme example of a poorly sampled density field but nevertheless produces beautiful images. Blender on python is the ideal tool for artistic scientific visualization.

Let me know what you think!

1 Like

Please quit Blender as soon as possible, before those black holes wipe out our whole existence!

Great work. Plan to share the code? How long did the file load take and also rendering take?

Oh there may be a few ultramassive black holes in there but they are tamed :slight_smile:

The dataset is ridiculously small, only ~500 particles and a few times that tetrahedra. It renders almost real time in cycles (GPU) using only one sampling per pixel. There is some aliasing in the bright/dense regions but it is mild.

yes of course, I will explain the technique and code in a tutorial I am preparing.



Sounds enticing, make it as long as possible please!

It looks really cool. As a fan of physics, space and 3d-modelling i find this very interesting :wink:
Looks pretty similar to matter distribution as well on large scales. I guess it makes sense since dark matter and matter interact trough gravity.

Also another little observation. I assume you started off with a cube shape here and it looks like it’s collapsing because there isn’t anything pulling it out from the outside. Maybe you could try making the effects warp around to the other side of the cube to keep the shape up better. I don’t know if that is considered more or less realistic. Just a little thought.