Hi all,

I encountered a problem I can't grasp. When I change the camera perspective, load the pixels, get these into an image, and draw this image, the image's perspective is somehow different. But the thing I really don't get is that when I subsequently reset the camera's perspective, the image's pixels are changed as well; somehow these reset to the new perspective, as if some 3d information is still present.

This is my code; you can comment out 1. camera(), 2. image() and camera()...

PImage im;

void setup() {
 size(400, 200, P3D);
 im = createImage(width, height, RGB);
}

void draw() {
 background(127, 127, 127);
 translate(width/2, height/2, height/-2);
 box(width/4);
 camera(width/2.0 - 100, height/2.0, (height/2.0) / tan(PI*60.0 / 360.0), width/2.0 - 100, height/2.0, 0, 0, 1, 0);
 loadPixels();
 for(int i=0; i<pixels.length; i++) {
   im.pixels[i] = blendColor(pixels[i], color(0, 255, 255), SUBTRACT);
 }
 image(im, 0, 0);
 camera();
}  

Hope I explained things clear...

Thanks in advance, BEnm

Hi Manic,

The same problem occurs if I use get() instead of loadPixels():

im = get();
for(int i=0; i<im.pixels.length; i++) {
 // pixel manipulation...
}
...etc...

I write "im" to the screen, so I don't have to update the pixels.

If you comment out the last camera call (and also the last image call) you see the perspective I want: you see two sides of the box.
Two things go wrong:
1. If I draw the image (only comment out the last camera call) the camera seems to close in on the box
2. When I subsequently call camera the image seems to change - as if the future somehow changes the past, because now the image only shows one side of the box!?!

.ben (aka BEnm)

Hi Manic!

This is indeed the perspective I want.. But the way it's done seems a bit odd..

im.pixels = pixels !?!

It seems that a camera perspective is some sort of metadata in pixel arrays. That would explain why a subsequent camera call changes the perspective captured in previous image calls (and, maybe, your solution).
But, I need to change the camera perspective at least three times, and that would wreck previously drawn images. I want to make a stereoscopic (anaglyph) animation, that's why..

Thanks for thinking with me!

I think a potential problem is that im.pixels=pixels doesn't quite do what you think it does.. that will make both the display pixels, and the images pixels point to the same array of data. So if you edit im.pixels, you're editing pixels too, as it doesn't *copy* the array, it just sets both pointing to the same array.

You need to actually copy the pixels array instead of just assigning one to the other.

You want arraycopy(pixels,0,img.pixels,0,pixels.length); I think, assuming that img.pixels is big enough.