Hey everybody, I have recently started using Papervision 3D for AS3 and had a couple questions regarding best practices and some performance issues i’m having.
Question 1:
First, in a lot of online examples i see some people are extending BasicView to write specialized 3D objects, making use of the internal onRenderTick function for their scene rendering. on the other hand, lots of other examples use standard MovieClip or Sprite extended classes which use onEnterFrame handling to process their scene rendering.
what have you guys experienced, is either of these methods better in terms of performance or integrating with a larger 3D site/application?
Question 2:
I built a pv3d class that simulates falling leaves, each leaf compising of two planes back-to-back in a DisplayObject3D wrapper, with each plane consisting of a MovieAssetMaterial from the library containing the graphics for the front and back of the leaf respectively. (transparency is true, smoothing, animated and interactive are false, and i’m using 2 triangles per plane). I then animate that leaf on the stage to simulate the falling aspect, but i am simultaneously rotating the leaf on all 3 axis’ so that it looks like it is tumbling down through the air. i update all the leaves via an array loop and only render the scene once for all the leaf position changes.
After more than 10 leaves, the performance gets kinda shoddy, and after 20 it has the potential of stalling out flash player. is this many objects rotating independantly too much to ask of papervision? are there better tactics to employ when having this many independant objects? for a while i thought about using a separate viewport for each leaf, but i assume that would be even more resource heavy. your thoughts?
thanks everyone in advance!
