Just uploaded a 3D rendered (side-by-side 1080p half-res)...
// May 14, 2013 11:00:26 PMJust uploaded a 3D rendered (side-by-side 1080p half-res) capture of ?Kohlenstoffeinheit? (my entry for the PC 4k intro competition @Revision 2013) to Youtube.
To get a more ?stable? 3D effect I removed the tv-scanlines and noise post-processing effects from the original release. The rendering is far from optimal (just ?Toe-In? stereo). To make things worse I haven?t even cared about depth budget management at all :) Anyway, enjoy!
Btw: Setup code for the stereo conversion can be found here. Fragment shaders here and here.
Version 3.1.4
admin // May 14, 2013 2:06:06 AMThis release of dyn4j is a small maintenance release to fix a bug introduced in 3.1.3. With this some flip methods have been added to the Geometry class to allow flipping of polygons about arbitrary axes and points.
This release should not contain any breaking changes unless you were using or storing shape, fixture, body, joint, etc. ids. These ids were of type String, but have been changed to UUID for performance (they were actually UUID.toString() anyway). You can still get the old String value by getId().toString().
No new release of the Sandbox was published for this release.
Version 3.1.3
admin // May 13, 2013 2:01:45 AMThis release fixes a bug in the Polygon inertia and center of mass calculations. This bug would also throw off multi-fixture body’s inertia. This bug would manifest if a polygon was created away from the origin. This version has no API changes.
No new release of the Sandbox was published for this release.
?Toe-In? experiment #5, this time parallax occlusion...
// May 11, 2013 2:43:45 PM?Toe-In? experiment #5, this time parallax occlusion mapping (kinda raymarched too) in stereo3D. Get the code here (shader code stays the same).
?Toe-In? experiment #4, a stereoscopic version of a...
// May 11, 2013 2:37:05 PM?Toe-In? experiment #4, a stereoscopic version of a 3D Mandelbrot fractal (not the real ?holy-grail? one). JOGL code can be found here and here.
Next try: Rubber Mandelbulb in stereo 3D (using my old...
// May 11, 2013 2:26:43 PMNext try: Rubber Mandelbulb in stereo 3D (using my old bruteforce raymarcher). Code as always here and here (java and shader).
And another ?Toe-In? stereoscopic port of my old...
// May 11, 2013 2:18:02 PMAnd another ?Toe-In? stereoscopic port of my old metaballs bruteforce raymarcher. Code can be found here (java) and here (shader).
Currently experimenting with stereoscopic rendering and...
// May 11, 2013 2:07:21 PMCurrently experimenting with stereoscopic rendering and different methods for calculating the stereo pairs. The most simple method I guess is ?Toe-in? stereoscopy, where the camera has a fixed and symmetric aperture, and each camera is pointed at a single focal point. The downside of this method is the introduction of a vertical parallax, which generally causes discomfort while viewing. Anyway good enough for a first shot ?
As a first routine I ported my quaternionic Julia brute-force raymarcher to stereo 3D. Code can be found here (java) and here (fragment shader).
Ofcourse in addition to the original ?minified?...
// Apr 7, 2013 3:28:47 PM
Ofcourse in addition to the original ?minified? version of the Kohlenstoffeinheit fragment shaders, I also comitted an unminified development version, that should be way more understandable (at least I hope so): main.fs post.fs
Here?s the promised direct 1:1 JOGL2 port of my PC 4k intro...
// Apr 7, 2013 3:07:00 PM
Here?s the promised direct 1:1 JOGL2 port of my PC 4k intro competition entry for Revision 2013. Sure it got a little bigger while porting but the shader and control code remained more or less untouched. The intro renders a fullscreen billboard using a single fragment shader.? The shader basically encapsulates a sphere-tracing based raymarcher for a single fractal formula with camera handling. The rendering technique is a little bit different than usual as the renderer accumulates the spheretraced distances of the estimation function over the ray-length, and converts them into a pixel brightness, resulting in some sort of ?X-Ray? look for the marched volume.
Additionally a second post-processing shader is applied to the render output from the raymarching shader. Post effects are radial blur, vignette, color abbrevation, tv-lines and noise to make the overall look more interesting and less ?sterile?. The different intro parts are all parameter and camera position variations of the same fractal. Anyway, code as always on Github: GL3_Kohlenstoffeinheit_Port.java and the corresponding main+postprocessing shaders: main.fs post.fs