Joshua Knapp

I'm new to the Jitter world and I'd like to...

1) ...generate a random jit.matrix with pixels ranging from black to white (grayscale). (I can do this!)

2) ...then process this initial matrix so that...

2.1) ...every black pixel rgb(0, 0, 0) "grows" into a square figure which is 509 pixels wide. The pixels get whiter the 

further they are from the center pixel: Center px has rgb (0, 0, 0) -> neighbouring pixels have rgb (1, 1, 1) 

Konzept.png

When a center pixel has rgb (27, 27, 27), it grows 455 pixels wide, so it's wide enough, that the rgb values of the 

pixels growing outwards increase +1 until they reach almost-white (254, 254, 254), then their goal is reached. 

This means that a center pixel with rgb (254, 254, 254) does not grow and a center pixel with rgb (253, 253, 253) 

2.2) ...when 2 figures overlap, I want to calculate the average of the rgb values of the overlapping pixels.

AverageOut.JPG

I'm happy about any solutions, thank you!

1) ...generate a random jit.matrix with pixels ranging from black to white (grayscale). (I can do this!)

2) ...then process this initial matrix so that...

   2.1) ...every black pixel rgb(0, 0, 0) "grows" into a square figure which is 509 pixels wide. The pixels get whiter the 
   further they are from the center pixel: Center px has rgb (0, 0, 0) -> neighbouring pixels have rgb (1, 1, 1)                         
   -> then (2, 2, 2) and so on. 

   When a center pixel has rgb (27, 27, 27), it grows 455 pixels wide, so it's wide enough, that the rgb values of the     
   pixels growing outwards increase +1 until they reach almost-white (254, 254, 254), then their goal is reached.     
   This means that a center pixel with rgb (254, 254, 254) does not grow and a center pixel with rgb (253, 253, 253)     
   grows only 3 pixels wide. 

   2.2) ...when 2  figures overlap, I want to calculate the average of the rgb values of the overlapping pixels.

Hello Jitter people,
I'm new to the Jitter world and I'd like to...
1) ...generate a random jit.matrix with pixels ranging from black to whi

looking-for-a-matrix-processing-algorithm

I don’t understand the meaning of your answer

The longer I try to build this patch idea, the harder it seems. 

I thought maybe jit.expr would do the job, because it can generate gradient masks, but only from the center of the matrix it seems. Help I'm lost and I don't know if the idea is event worth my time :D :/

Nobody? 
The longer I try to build this patch idea, the harder it seems. 
I thought maybe jit.expr would do the job, because it can generate gradient masks, but only from the center of the matrix it seems. Help I'm lost and I don't know if the idea is event worth my time :D :/

I wouldn't be able to implement that "visually" (although I am not saying that it is impossible), I'd rather look for a programming method (writing code in a classic text form).

In my opinion, the specific feature of your idea is that you have to create "buffers" that remember the shapes that build up on each pixel, and then combine these buffers together (eg creating a matrix - by mixing shapes/pixels brightness accordingly - this method is generating a lot of data dynamically) or the implementation will be doing a lot of iterative processess. However it should be relatively easy to implement in any programming environment allowing access to pixel data ("matrices") and typical data structures.

If I was in a rush, I would do it in Processing or OpenFrameworks, for example, and connect to Max via Syphon or Spout. I'm not sure it's easy to write something like this with jit.pix or shaders (or generally Max internal mechanics). You could also experiment with JavaScript (the mechanisms built into Max like js / jsgui, Node, jweb) - but for high resolution images that might be inefficient (but I might be wrong, I didn't check it out). Of course you may also create an external for Max. Java "inside" Max will be a great mechanics for this kind of tasks, however it is a little bit deprecated component of Max (but maybe my knowledge about this is deprecated too).

In my opinion, the specific feature of your idea is that you have to create "buffers" that remember the shapes that build up on each pixel, and then combine these buffers together (eg creating a matrix - by mixing shapes/pixels brightness accordingly - this method is generating a lot of data dynamically) or the implementation will be doing a lot of iterative processess. However it should be relatively easy to implement in any programming environment allowing access to pixel data ("matrices") and typical data structures. 

sounds like you are talking about a cellular automata? check out the jit.gl.pix example patch called 

sounds like you are talking about a cellular automata? check out the jit.gl.pix example patch called pix.game.of.life (search the file browser)

Looking for a matrix processing algorithm