I'm not sure if I follow what you mean by vector.

A vector image is an image made of mathematical curves and lines which can be re-sized to any size you like. Hope you didn't get that mixed up with PVector(). These are two different things.

But maybe vector images aren't necessary.

My idea is to make a memory efficient converter, so in case the image is extremely huge, I wouldn't have to waste memory on an array to store the pixel color, and the output image.

This would work by getting the first pixel, then putting a resized version of the same pixel (A rect() in this case), and store it immediately to the image ( without erasing the previous pixel), and this would keep repeating until the whole image is re-sized.

Hope that makes a little more sense.

- MenteCode.

1) I'm planning on using the built-in ones.

2) I'm not sure I know what you're saying, as there are several types of curves (bezier(), arc(), curve()).
But if I'm understanding correctly, then that would mean to use curveVertex() in beginShape, to translate each point from the previous relative center. Some code would be very useful, to better understand what you have in mind.

Thanks for the help.

- MenteCode.

To be honest, I don't have the smallest idea on what you mean by writing my own transformation matrices, nor did I know that you could do that, so I said in the previous response that I'll be using the built-in functions.

Processing's built in transformation matrices manipulate the coordinate system which changes where things are drawn (translation() / rotation()), how big they are (scale()), etc. They do not however, actually affect the saved coordinates that make up your drawings. They only change how it is drawn.

Could you describe concisely what the post in the link is about?