How is this done?

[flashpipe1]

Any idea how this is created??

http://www.airtightinteractive.com/p...le1/index.html

Is it using papervision??

[lesli_felix]

papervision or sandy or away3d, most probably.

Although - it doesnt use solid shapes or 3d models, just flat images rendered in 3d space. If I were building it, I'd be tempted to use the distortimage class from sandy and build the rest from scratch. That would be more efficient and use less code.

[realMakc]

on the contrary, you would have to move all stuff around on per-vertex level.

[gerbick]

Read his blog (airtight.com/news)... I swear he was using PaperVision3D and DisplayObject3D sprites within PaperVision3D v2. He uses that a lot in his other motion based 3D stuff.

[lesli_felix]

on the contrary, you would have to move all stuff around on per-vertex level.

That's a very good point. The corners of the images would be accurately rendered in 3d, whilst the sub vertices of the skewed image wouldn't be.

I've got a feeling that's still an issue in those engines anyway. I remember reading that large surfaces in sandy suffer from distorting for this reason and they were looking at fixes that didn't involve killing the processor.

No idea how they work now, could be that they've ironed that out in the latest versions of the engines, however, if you keep your scaling and perspective at moderate levels, you wouldn't notice the difference anyway.

[realMakc]

actually, I was making another point, but whatever... as for sandy, and perspective on large planes, it was probably about having to use clipping, or otherwise resources are waisted for correcting perspective outside of area actually visible on screen.