On 07/17/2016 02:41 AM, Bald Eagle a dit:
> I sampled an image_map pigment of a photo of my son with eval_pigment, and
> plotted the resulting vectors in RGB-space (step 10 in x and y).
> The scanning takes 33 sec (2448 x 3264).

    I've done something like that, but with some C and shell coding.

    Original video is here :
http://la.buvette.org/photos/myrys/20ans/avance-rapide.avi

    (gory) C code is here :
http://la.buvette.org/vrac/colorcube.c

    And the result is here :
http://la.buvette.org/photos/myrys/20ans/fx/colorcube.avi

> I want to find a way to "cluster" similar values - I'm experimenting with
> rounding the RGB channels, etc.

    I'va also done (a long time ago) some search in this field

> I also made a sort of "parametric" RGB map - translate along x by gray, rotate y
> by green, then x by red, then z by blue

    Mmmm, sound nice :)



-- 
http://weblog.mixart-myrys.org/?post/2016/03/De-la-temporalisation

Post a reply to this message