I'm trying to grasp how OOP work.  I've ran into a confusion.  Take the following code from a tutorial page:
Code:

void setup()
{
  size(200, 200);
  framerate(25);

  myBalls = new Ball[33]; // creating an (empty) array with a
				  // capacity of 33 elements of type Ball

  for (int i = 0; i < 33; i++) // iterating 33 times
  {
    myBalls[i] = new Ball();   // populating the array
					 // with new Ball objects
  
    myBalls[i].x = random(0, 200);
    myBalls[i].y = random(0, 200);
  }
} 

my confusion is with this line.  I don't see why it's necessary:
Code:

    myBalls[i] = new Ball(); 

this seems redundant to me since the ball objects are already initialized earlier in the setup with this line:
Code:

 myBalls = new Ball[33]; 

Am I missing something?

A concept that seems to trip lot of beginners... I can understand that, coming from the C world, where defining an array of objects allocates the number of bytes of the object multiplied by the number of slots.

But Java, in new Ball[33], only allocate an array of references to objects, these references being null until we make them point to real objects, created classically with new Ball().

The first new allocates the array, the second one creates the object (maybe with constructor parameters, etc.).

What is confusing too is that for primitive types, new int[42] or new double[99] actually allocate all is needed...