Form Over Function

Finished my post processing anti-aliasing evaluation and settled with a combination of FXAA and 2xFSAA. As before I used the Mandelbrot and Displaylist routines as basis for the experiments. Source on Github: GL3_FXAA_Mandelbrot.java and GL3_FXAA_DisplayLists.java. JOGL “friendly” fragment shader (eliminated the need for a dedicated vertex shader): .fs

Continuing with my experiments regarding anti aliasing, I took a look at the idea of combining post-processing anti-aliasing (FXAA) with super sampling as explained by Timothy Lottes. As the above picture clearly shows the results of 2xFSAA+FXAA line up somewhat between 2xFSAA and 4xFSAA at the cost of 2xFSAA (as FXAA is neglectable performance wise). JOGL2 port/implementation coming up shortly …

With the current rate post-processing antialiasing techniques are invented my projection is that we will run out of four letter acronyms by the end of the year: MLAA, FXAA, SRAA, GPAA, NFAA (to name only a few). So I felt it was time to revisit the whole anti-aliasing problem and started with porting/evaluating FXAA. Looks promising, more to come …

My eighth (first, second, third, fourth, fifth, sixth and seventh) “Monkey Mathica” plane deformation shader series try. Setup code is the same as in the other tries: GL3_MonkeyMathica.java but ofcourse the fragment shader differs: .fs

My seventh (first, second, third, fourth,fifth and sixth) “Monkey Mathica” plane deformation shader series try. Setup code is the same as in the other tries: GL3_MonkeyMathica.java but ofcourse the fragment shader differs: .fs

My sixth (first, second, third, fourth and fifth) “Monkey Mathica” plane deformation shader series try. Setup code is the same as in the other tries: GL3_MonkeyMathica.java but ofcourse the fragment shader differs: .fs

My fifth (first, second, third and fourth) “Monkey Mathica” plane deformation shader series try. Setup code is the same as in the other tries: GL3_MonkeyMathica.java but ofcourse the fragment shader differs: .fs

My fourth (the first, second and third) “Monkey Mathica” plane deformation shader series try. Setup code is the same as in the other three tries: GL3_MonkeyMathica.java but ofcourse the fragment shader differs: .fs

My third (the first, and the second) “Monkey Mathica” plane deformation shader series try. As before no textures involved, everything generated procedurally from scratch. Setup code is the same as in the first and second try: GL3_MonkeyMathica.java but ofcourse the fragment shader differs: .fs

My second (here’s the first) “Monkey Mathica” plane deformation shader series try. Also no textures involved, the fragment shader generates everything procedurally from scratch. Setup code is the same as in the first try: GL3_MonkeyMathica.java but the fragment shader differs: .fs

First try in my “Monkey Mathica” plane deformation shader series. No textures involved, the fragment shader generates everything procedurally from scratch. Was originally planned for a PC 1K intro entry at the “Revision Easterparty 2011” but I ended up making a 4k intro :) … Anyway, code as always on Github: GL3_MonkeyMathica.java and the corresponding shader: .fs

Awesome! @twbombo ported “Elektronenmultiplizierer” to WebGL after I published the JOGL2 code here yesterday. Thanks alot man! Really cool to see how quick and seamless u can port code from JOGL2 to WebGL. Maybe someone in the mood for a JOCL port ? :) Anyway, enjoy here …

As the fragment shader of my Revision 2001 PC 4k entry port was processed or better “obfusicated” with the shader minifier tool to meet the size restriction of <=4096 bytes, it became nearly impossible to understand or make any sense of the GLSL code. So here is the “unminified” version with additional comments I used during development (as good as it gets): .fs

As promised before, here’s the direct 1:1 JOGL2 port of my PC 4k intro competition entry for Revision 2011. 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 encapsulates basically two different routines: A sphere-tracing based raymarcher for a single fractal formula and a bitmap orbit trap julia+mandelbrot fractal renderer. Additionally an inline-processing analog-distortion filter is applied to all rendered fragments to make the overall look more interesting. Anyway, code as always on Github: GL3_Elektronenmultiplizierer_Port.java and the corresponding shader: .fs

I have been quite lazy in the last couple of months about posting something new here, but at least I had a good reason: I was working fultime on my entry for the Revision 2011 PC 4k intro competition (executable size has to be <=4096 bytes) called “Elektronenmultiplizierer”. As this was my first 4k intro since a long time (my last one was under MS-DOS) it took nearly three month to finish. Anyway, I made it just in time for the competition and successfully released it at Revision. In the end it placed 4th and me and my buddy (who released a new graphics mode for the VIC20: “Yes VIC can”) were awarded with the “Newcomer Award”. After Revision I took the liberty of doing some chillout for a couple of weeks, wich are over as of now. Ofcourse the next logical step was to do a JOGL2 port of my Revision PC 4k entry …