I'm currently trying to build a patch which loads an image file, then performs an analysis of each pixel in the image, putting the RGB data into a jit.matrix.

To do this I'm overlaying a suckah on my image, and feeding it a stream of co-ordinates which i create with a counter. However I mis-understood this function of suckah and it seems that the co-ordinate feature is for the whole screen, not the area of the individual suckah object.

Can anyone recommend another object or external which I could use for something similair?

suckah-image-analysis-question

> I'm currently trying to build a patch which loads an image file, then performs

> an analysis of each pixel in the image, putting the RGB data into a

Feb 4th update: David Fiuczynski at Harvard

> I'm currently trying to build a patch which loads an image file, then performs
> an analysis of each pixel in the image, putting the RGB data into a
> jit.matrix.
>

getcell message to jit.matrix.
J-F.
-- 
Feb 4th update: David Fiuczynski at Harvard
http://www.jeanfrancoischarles.com


Hi, i'm aware of how to retrieve data from the matrix, currently i'm sending it to jit.window to view the analysed data (which should theoretically look like an exact replica of the original image, but stored as rgb data in a jit.matrix. And it does).

However what i was originally asking was if there's a way to send suckah a continuos stream of co-ordinates relative to it's own position on top of an image (fpic) rather than co-ordinates for the whole desktop. If not is there another object or external for this?

The reason i need to get rgb data for each pixel in an image is for producing sound from an image's colour data, i plan on moving away from RGB once i have the prototype working, just sticking with the defaults for now however.

Turns out that i can use jit.qt.movie to load the image, send to a 4 plane matrix and the getcell on that matrix. No no need to use suckah.

Now onto playing with rgb2hsl to analyse what the results are.

Suckah Image Analysis Question