No, it's the line

movement = (height/100)*sin(theta);  

followed by...

theta += 0.07;

the sin of theta will vary from 0 to +1/-1 so your movement variable will change accordingly - think (height/100) * 0 compared to (height/100)*1 :D

yes, the sin of theta will vary, that's my intention, i'm moving the image up and down with a amplitude of height/100

here's a sample of what i'm trying to do, to speed is constant

It is hard to tell, since the movement isn't big anyway, and quite slow from the start.
I tested your code with the following:

1. float theta;
2. PVector position, initialPosition;
3. PImage a;
4.  
5. void setup() {
6.   size(100, 500);
7.   initialPosition = new PVector(width / 2, height / 2);
8.   position = new PVector(width / 2, height / 2);
9.   a = loadImage("H:/Temp/cursor.png");
10. }
11.  
12. void draw() {
13.   background(255);
14.   image(a, position.x, position.y);
15.   float movement = (height/100)*sin(theta);  
16.   position.x = initialPosition.x + movement;
17.   theta += 0.07;
18. }

but that's the kind of thing you should provide yourself, because it is hard to help from a simple code fragment, and when we have to do this, we can miss something crucial present in you code. For example, some people make the error to do the loadImage() in draw().
Beside, providing the small code snippet yourself, beside improving your chances of getting an accurate and fast answer, allows you to reduce the problem to its essence, to think about it differently, and sometime to see it doesn't lie where you thought it was...

I ran the above sketch while I was typing this message, and saw no noticeable slow down.

shouldn't we after line 17 check if theta > 2 PI and if so

1. theta = theta-(2*PI);

or

1. theta = 0;

or so?
maybe sin() gets slower when getting values of 22000 instead of 1.7.....................

Greetings, Chrisir  

Hi,