Animation of the comet we are landing on as I write this

http://www.youtube.com/watch?v=AwpimaBGiU8

(Produced in Codea, using a 3D model provided by the mission scientists. The rotation I am using is random and not realistic)

The data file is here, you can see for yourself, I counted 63,000 triangles.

https://gist.github.com/dermotbalson/e9aa33679ee69458a0a5

Yes, OpenGL is handling the whole thing, it is very efficient at culling (as an example, my 3D dungeon has 5 million pixels but it renders the room I’m in at over 50 FPS on my iPad3, which is very good). Apple’s advice is not to bother culling, you just get in the way of OpenGL.

Do you have any stats about the number of vertices / faces that your rendering along with the frame rate?

Also I’m assuming your just throwing the entire mesh at the GPU, you’re not doing any backface culling or anything like that - are you?

Cool demo all the same :slight_smile:

Wow that is impressive, what’s the file format - is it a standard one?
I know what you mean about culling, it’s like trying to write assembler that’s efficient on chip cache use, most people seem to agree that these day’s it’s better to let the compiler sort it out. :slight_smile:

Actually there was a fractal mountain demo a while back on the forum that just dumped the whole mesh every frame and it ran really well.

The file format is standard OBJ because that is readable text. OBJ has a separate MTL file, so I just combined it with the OBJ.

Writeup here

http://coolcodea.wordpress.com/2013/12/13/139-importing-full-3d-models-into-codea/

Thanks again. Even more stuff to read… :slight_smile: