Monday, May 14, 2007

Intertwined 3



Last one for today.

Intertwined 2


Not really better. Try darker one.

Intertwined



I find quite strange the "plastic" effect that emerges from this image. Maybe blue part is too or not enough dark.

Lights and shadows


My daily one.

Sunday, May 13, 2007

Variations



Varying gradient center. Next : change the gradient shape, by interpolating control points from initial corner dispersion to planar and regular center

New gradient method



Compared to this new gradient computation method, my previous one was silly (but allows nevertheless the "tiger flower", which is not the case of this new method). This new method simply compute decreasing centered subpiece of bezier surface.

Tiger flower



Same as before, except that only one frame is not drawn for two computed.

Reiteration



Same as previous, with changed iteration pattern

Color gradient


Same picture as before, but with some more iterations, and color rendering. Nice indeed.

Some results

Here is a first result with automatically created gradient, but
  • some errors appear in bezier surface cut
  • the gradient pattern occurs to reflect the way cut is done : may be nice, but it is not what I was expecting

Oups

Something's wrong in coord computation. Interesting though

Manual color gradient


Soon more automatic and gradual ... gradient

Daily new

Saturday, May 12, 2007

More cutting edges

Playing again.Post them before trying some improved features.

Metart


Just for fun : my class hierarchy. Obviously cleaning is needed though ...

Friday, May 11, 2007

Cutting edge 2

Just varying the motif. I was just wondering if arbitrary drawing on bezier surface were not analogous to curve spaces in physics ... Need to experiment this.

Cutting edge

Still generalizing bezier shape patching. New algo, by considering :
  • a bezier surface defined by a close four-parts bezier curve
  • a generalized 2D coord system can be defined on this surface, coord being between 0.0 and 1.0
  • algo can then compute a sub surface defined by two "points" of the previous bezier surface
Thus can be defined and computed the image iterated form :
  • the bezier surface is defined by a pic bezier, completed to be closed
  • four sub-surfaces are computed, from top to bottom

Thursday, May 3, 2007

Slaughter



Just mapping the previous result on a spiral pattern, with some color playing, and it occurs. Badly in time ...

Rusted metalflower


Keep on improving, still rebuilding a lot of things, and extending. This image uses a new splitter operator, which takes a surface (made of four bezier curves forming a close area), and splits it according to one orientation (two are available given the two possible opposite curve pa) and one ratio. Computation of the boundary is then done by computing the two new points on the 'base" curves, and aggregating control vectors of the "untouched" two other curves.

The previous curve corresponds to a splitter recursion on a simple pic bezier, alternating "horizontal" and "vertival" splitting. Chase the red ...