Use floats with get()

Uranhjorten

Hi!

I am trying to make a light field renderer based on this Nvidia research https://research.nvidia.com/sites/default/files/publications/Slim near-eye display using pinhole aperture arrays.pdf

Long story short - "int o = 1;" needs to be 0.75 but get() doesn't support floats.. any ideas?

This is my code so far:

int N = height/3;
int o = 1;
void setup() {
  size(400, 300);
}
void draw() {
  PImage myImage = loadImage("fox.jpg");
  image(myImage, 0, 0);
  for (int i = 0; i < myImage.width; i = i+1) {
    for (int j = 0; j < myImage.height; j = j+1) {
      PImage c = get(i*N*o, j*N*o, N, N);
      tint(0, 200, 255); //just for checking it works
      image(c, i*N*o, j*N*o);
    }
  }
}

GoToLoop

https://forum.Processing.org/two/discussion/15473/readme-how-to-format-code-and-text

We should only access width & height after size()! L-)

Uranhjorten

Whoops.. thanks :)

Uranhjorten

But still - anyone have any suggestions?

kfrajer

Ok... first point: Do not load your image in draw(). You are loading the same file 30 fps.

Second point: I would recommend calling your variable "o" something more meaningful based on what it does. If you write 100 or 200 lines of code, the letter o gets mixed with 0 and anything that looks round. If o is a scale factor, for example, call it scaleVal for sake of clarity.

A possible solution is to make your variable float instead of integer. If you are trying to interpolate between pixel colors, you can check lerpColor: https://processing.org/reference/lerpColor_.html
You need to be also aware of the colorMode. The default mode is RGB. Check https://processing.org/reference/colorMode_.html

You should also check the reference for tint() as you are calling it 30 x withd x height many times. If you need to use it, it will be used outside the for loop: https://processing.org/reference/tint_.html

Kf

Uranhjorten

All good suggestions, but it's not really what i'm looking for. Its just how i use get() with float?

TfGuy44

No. Stop. Fix the fact that you are loading your image in draw() first. i refuse to even look at the rest of your code until you fix that. Do not load things in draw().

I will shout this at you if I have to.

DO NOT LOAD THINGS IN DRAW.

Uranhjorten

Look I know I'm not supposed to do that and I know the code has flaws - but it's just a sketch and I just need to know if anyone knows if there's an alternative to using floats with get()

TfGuy44

DO NOT LOAD THINGS IN DRAW!!!

Cast them to int?

float x = 1.7567283832;

get( int(x), int(x) );

The problem, you see, is that it doesn't make any sense to use floats with get(). The get() function is trying to get you the color of a pixel. If you're get()ing the pixel at (x,y), you're getting the x'th pixel in the y'th row. You can't get() a color of a pixel between pixels, because, well, there's nothing between them. There's no, say first-and-three-fourths pixel. There's the first pixel, and the second pixel. And so on. They're counted by whole numbers, not floats.

Also, DO NOT LOAD THINGS IN DRAW().

Uranhjorten

Okay, I see - thank you very much :)

And I promise to never ever load any images outside of draw ;)