hi forum

 

the following example is using a pgraphics to reproduce the screen drawing in a larger resolution (sphere in the middle of the screen).

 

since I've updated to processing 2.0.2 my code behaves completely different. so maybe i'm doing something fundamentally wrong or there's a glitch in 2.0.2. can someone look into the following example and let me know if it's me or processing that's doing it wrong?

 

to reproduce:

1. start the sketch

2. hit any key to dump the pgraphics into a tif-file

3. expected result: screen: sphere in the center / file: sphere in the center

4. result: screen: sphere is totally offset / file: sphere in the center

 

 

 

//toggle hires-pgraphics
boolean makehires = true;

int myWidth = 300;
int myHeight = 300;
int hires_scale = 3;
int myWidth_hires = myWidth * hires_scale;
int myHeight_hires = myHeight * hires_scale;

PGraphics hires;

////////////////////////////////////////////////////////////////////////////////
void setup() {
  // setup screen
  size(myWidth, myHeight, P3D);

  if (makehires){
    hires = createGraphics(myWidth_hires, myHeight_hires, P3D);
  }
}

void draw() {
  // drawing a sphere ono the screen
  background(0);
  stroke(0);
  fill(255);
  camera(width/2.0, height/2.0, (height/2.0) / tan(PI*30.0 / 180.0), width/2.0, height/2.0, 0, 0, 1, 0);
  translate(width/2, height/2, 0);
  sphere(100);

  // replicating the same sphere scaled in hires
  if (makehires){
    hires.beginDraw();
    hires.background(0);
    hires.stroke(0);
    hires.fill(255);
    hires.camera(hires.width/2.0, hires.height/2.0, (hires.height/2.0) / tan(PI*30.0 / 180.0), hires.width/2.0, hires.height/2.0, 0, 0, 1, 0);
    hires.translate(hires.width/2, hires.height/2, 0);
    hires.sphere(100);
    hires.endDraw();
  }
}

void keyReleased() {
  if (makehires){
    String timestamp = year() + nf(month(),2) + nf(day(),2) + "-"  + nf(hour(),2) + nf(minute(),2) + nf(second(),2); 
    hires.save(timestamp + ".tif"); // Saves a TIFF file named xxxx.tif
  }
}

 

 

 

 

 

 

hi forum.

 

i'm trying to figure out a way how to extract the global coordinates from a transformed point that i am about to draw. is there any way to do this that is non-destructive to the current stack of transformation?

 

 ...

 rotateX(objectRotationX);   // some object rotation

 for (int i = 0; i <= 10; i++){

   pushMatrix();

     translate(i,0,0); rotateX(i/10);   // or any other transforms that are changing through the loop

     //need global coordinates of 'current' position here to add some analysis

     point(0,0,0);                          // this would be the point in question

   popMatrix();

 }

// i need to be back in objectRotationX here

...

 

as far as i know, applyMatrix or resetMatrix would destroy the object's rotation. the reason for the question are multiple: it could for example be, that the z-depth or distance to a global camera have to be known...