My Shapes3D library implements shape picking by

1. creating an off-screen buffer (pickBuffer)
2. set the pickBuffer transofrmation matrix to match the on-screen matrix
3. then draw each shape onto the buffer in a different colour
4. find the colour at the mouse position
5. find the matching object (if any) and return it.

The actual code is shown below. In mousePresssed I call this method to identify the shape clicked on unfortunately it doesn't work now (it used to in 1.5.1) i.e.

1. void mouseClicked(){
2.   Shape3D picked = Shape3D.pickShape(this,mouseX, mouseY);
3.   if(picked != null){
4.     picked.fill(randomColor());
5.     println("Hit > " + picked.tag);
6.   }
7.   else {
8.     println("Missed " + millis());
9.   }
10. }

but it keeps missing the boxes unless I duplicate the camera transformation used in the draw() method in the mousePressed method e.g.

1. void mouseClicked(){
2.   camera(c * sin(a), 10, c * cos(a), 0, 0, 0, 0, 1, 0); // copied from draw()
   
3.   Shape3D picked = Shape3D.pickShape(this,mouseX, mouseY);
4.   ...
   

So my questions are

1. when the mousePressed (or any input event handler) method is executed does the transformation matrix still reflect changes made in draw?
2. would it be better to call this shape picker method from inside draw after any camera() commands?
3. is there a better / more 'proper' way of setting the transformation matrix in the off-screen buffer than the way I have done it (lines 7 & 8 in the code below)?
   

I suspect from my experience that the answers to the first 2 questions are going to be no/yes. The third question I don't know the answer so I would be interested in any response or detail related to how Processing works under the hood.

Thanks

Shape picking code

1. public static Shape3D pickShape(PApplet papplet, int x, int y){
2.   if(pickBuffer == null || pickBuffer.width != papplet.width || pickBuffer.height != papplet.height){
3.       pickBuffer = (PGraphicsOpenGL) papplet.createGraphics(papplet.width, papplet.height, OPENGL);
4.   }
5.   pickBuffer.beginDraw();
6.   // Set the camera same as the drawing surface
7.   pickBuffer.camera.set(((PGraphicsOpenGL)papplet.g).camera);
8.   pickBuffer.projection.set(((PGraphicsOpenGL)papplet.g).projection);
9.   pickBuffer.noLights();
10.   pickBuffer.noStroke();
11.   pickBuffer.background(WHITE);
12.     
13.   // Draw to the buffer
14.   pickModeOn = true;
15.   drawAll();
16.   int c = pickBuffer.get(x,y);
17.   pickModeOn = false;
18.   pickBuffer.endDraw();
19.   return pickMap.get(c);
20. }