View Demo

Some of you might remember my Chaotic Particles demo from last year. That demo was featuring 10'000 particles on a plain old 2d canvas. I decided to optimize that demo a bit in order to support 100'000 particles. I also fixed a little issue where numeric inaccuracy allowed particles to escape and made the influence map more fine grained.

Want to see the source? Just use view source and feel free to ask questions.

Next up: 4'000'000 Particles using WebGL.

View Demo

I think this is my favorite canvas demo I have created so far. It is an interactive drawing tool based on particle effects. It is the result of me trying to create some generative art using canvas. The techniques used are actually pretty similar to the ones shown in my frontendconf talk on particle systems. In short:

p.vx = p.vx*0.8 + getNoise(p.x, p.y, 0)*4+fuzzy(0.1);
p.vy = p.vy*0.8 + getNoise(p.x, p.y, 1)*4+fuzzy(0.1);

p.x += p.vx;
p.y += p.vy;

data[index] = tonemap(hdrdata[index] += r);
data[index+1] = tonemap(hdrdata[index+1] += g);
data[index+2] = tonemap(hdrdata[index+2] += b);

I'm planning to play with a few improvements especially in tone mapping and controls in the future but feel free to take a look at the source on github. I hope you enjoy it.

Video

Video from my talk at frontendconf on particles and html5 canvas, 100% live coding.

Video on ustream

Demo

View Demo

Sourcecode

on github

Note

This is an experiment. It's hacky and you should not use it as an example of good style - it's not.

I finally found a way to optimize 2d canvas drawing on the iPhone 4. Because of the retina display the canvas seems to be rescaled in a slow way (in software?). So even though the rendering itself is relatively fast, the end result is slow.

Setting the Viewport

The first step is to set the viewport scale to 0.5 which will result in having one pixel per css pixel.

<meta name="viewport" content="width=device-width, initial-scale=0.5, user-scalable=no"/>

So now the rendering is fast, but the picture is tiny.

3D transforms to the rescue

To scale the picture up we can use css 3d transforms which are fast.

canvas { -webkit-transform: scale3d(2, 2, 0) translate3d(200px, 110px, 0); }

Also note the translation to move the image back into place as it is being scaled from the origin.

That's it you got yourself a massive performance boost in about two lines of code. :)

Play it now

Space Break

Space Break is a html5 acrade 'ball and paddle' game written in coffee-script. It features levels full of explosives, extra balls and even nukes. I hope you will enjoy it.

History

Space Break started out as coffee-break - a little project I did to get into coffee script. Because it turned out to be a lot of fun I decided to turn it into a complete game.

Technology

Space Break uses the canvas tag for rendering. The sound effects were created using csound and played using the audio tag. The graphics were made using the gimp and blender. Rake is used for controlling the asset pipeline. All the assets of the game (music, sounds, graphics) are self made - programmer art.

Browsers

Firefox 4 (3.6 kind of works), Chrome 9 and Safari (no ogg - no audio) work. Opera is crashy. Mobile safari on the iPad also works but the frame rate is low. Could be fixed with a bit of optimization.

Source and License

The code and assets are of course on github.

The Sourcecode (break.coffee, Rakefile) is licensed under the GPL V3.

The assets are licensed as Creative Commons BY NC SA

PLEASE LET ME KNOW IF YOU DO SOMETHING COOL WITH IT.

If the licensing doesn't work for you, please let me know.

Feedback

Feedback, both positive and negative is of course appreciated. Just leave me a comment, tweet or send me a email.

I created this demo more or less by accident when prototyping some particle mechanics for a game. It is based on a particle system, and an acceleration map but more on that later. It reminds me quite a bit of simulations of galaxy formation or a lava lamp, it's really nice to watch.

Demo

Click the image to open the demo. In the demo, set the particles in motion by dragging with your mouse.

Mobile devices

This demo also works on Android including fullscreen (tested on a nexus one with 2.2), iPhone and iPad! Give it a try ;)

How it works

Each particle has a position and a velocity stored in a typed Float32Array (yes, they work on iOS). In addition to that there is an acceleration 'map' in the background. When you drag over the particle system, you initialize the acceleration map which sets the whole system in motion.

accelerationMap[particle.position] += particle.velocity*someFactor
particle.velocity += accelerationMap[particle.position];
particle.position += particle.velocity;
...

Finally after a long time I can present you a new demo. It is inspired by the wave propagation formula mr.doob demoed a while ago. I was quite amazed by how simple it was and so I decided to build myself a little pond.

Demo

Click the screenshot to start the demo. Move your mouse over the pool to disturb the water.

How it works

I think the really interesting part here is the wave propagation formula:

waveHeight = (buffer0[i-1] + buffer0[i+1] + buffer0[i+width] + buffer0[i-width])/2-buffer1[i];
buffer1[i] = waveHeight;
aux = buffer0;
buffer0 = buffer1;
buffer1 = aux;

The rest is pretty simple, and made up on the spot. The refraction is done by offsetting the texture coordinates by the height difference between the current point and it's neighbors. The lightning and caustics are faked similarly based on the height differences and the height of the current point.

In short it's all fake.

Performance

Chrome is fast, firefox is slow, at least thats the way it usually is. To my pleasant surprise the jaeger monkey builds of firefox seem to finally catch up. Another interesting thing is that safari seems to be the fastest (for this demo).

Credits

The photo was shot by Travholt licensed CC-BY-SA. As mentioned in the intro, I got the wave propagation function from mr.doob.

Other Experiments

Be sure to check out my other experiments. If you like what you saw you should subscribe to my blog or follow me on twitter.

This is an older demo I did a while ago but haven't published yet. It's pretty slow unoptimized and hackish and the projection is wrong and I don't feel like fixing it. But it still looks cool! So thought I should publish it anyway.

What does it do? It simulates a spiral galaxy with 5000 stars.

Demo

Click the screenshot to start the demo. It's pretty slow in firefox, works well in chrome.

How it works

Doing calculating the gravitational forces for 5000 particles in realtime using javascript is hard, and it would probably be quite hard to get stable spirals. So I cheated a little (a lot actually). To simulate the slowdown of the motion of stars in more dense areas I'm simply using a texture:

That's pretty much it: Generate a bunch of stars spinning around a center and modulate their speed according to the density map. Feel free to take a look at the source, but remember, I told you that it's a mess and all wrong.

Other Experiments

Be sure to check out my other experiments too.

When I saw the JS1K demo contest I knew I had to enter. I first tried to do some GFX demo, but I was lacking an impressive idea. So when I tried to sleep last night at 2 in the morning I had an idea which I needed to implement: Demonstrating evolution by natural selection in 999b/666b gzipped:

x = ('var_a=Math,b=a.random,c=document.g'+
'etElementById("c"),e=c.getContext("2d")'+
',f=[],g=b()*		 16777216<<0,h,i'+
',j,k,l=(g		  &16711680)>>16'+
',m=(g&65		    280)>>8,n=g&'+
'255,o,p		    ;c.width=320'+
';c.height=320;		    e.shadowBlur'+
'=10;for /* */	      (p=   0;p<64;p++)f'+
'.push(b(		    )*16777216<<'+
'0);func		    tion_q(d){fo'+
'r(d=d.t oStr		  ing(16);d.leng'+
'th<6;)			d = " 0"+d;retur'+
'n"#"+		      d } d ocument.body'+
'.sty		     l e .b a c kground='+
'q(g);		    setInterval(function'+
'(){e.		 c l e a r R e c t ( 0 ,'+
'0,320,		3 2 0 ) ; f o r ( p = 0 '+
';p<64;	    p + + ) { h = f [ p ] ; i = '+
'l-((h&	  1 6 7 1 1 6 8 0 ) > > 1 6 ) ;j'+
"=m-((h&65 2 8 0 ) > > 8 ) ; k = n - ( h"+
'&255);o=a.sqrt(i*i+j*j+k*k)/443;if(b()<'+
"o)h=f[p]=(f[b()*64<<0]&16773120|f[b()*6"+
'4<<0]&4095)^1<<(b()*24<<0);e.fillStyle='+
"q(h);e.fillRect(1+p%8<<5,1+p/8<<5,32,32"+
')}},100);e.shadowColor="#000";').replace(
/\s/g,"").replace(/_/g,' ');eval(x);
/*evolution in 999b by 29a.ch**/

+1 if you recognize that guy.

Demo

It's basically a few cells colored cells. The selection criteria is the color difference between the cell and the background. So cells that stand out have a higher chance of getting eaten.

Click me

How it works

You can see the full unminimized sourcecode here evolution.js I'll only explain the meaty stuff.

env = random()*(1<<24)<<0,
...
population.push(random()*(1<<24)<<0)

A random environment and population is initialized. The shift by 0 is a short way to round a number. Both the environment and the genes of the population are random 24 bit integers, that can be viewed as rgb colors.

subject=population[i];
r = r_-((subject&0xff0000)>>16);
g = g_-((subject&0xff00)>>8);
b = b_-(subject&0xff);
distance=M.sqrt(r*r+g*g+b*b)/443;
if(random()<distance){

The fitness for survival is determined by measuring the distance between the color of the individual and the environment. The lower the distance, the higher the chance for survival.

subject = population[i] = (
                // breed
                ((population[random()*popsize<<0]&0xfff000)|
                (population[random()*popsize<<0]&0xfff))
                // Mutate
                ^(1<<(random()*24<<0))
            );

When an individual dies, a new one is bred by sexual reproduction (parthenogenesis and necrophilia are possible as well). The new individual then gets one random mutation. That's pretty much all there is to it.

Last Saturday I have released my little javascript pathtracer that rendered a very simple scene. This time I want to show off some of the more advanced effects that can be simulated nicely using path tracing. I decided to set up a scene similar to the well known Cornell Box to demonstrate effects such as soft shadows, color bleeding, reflection, refraction and caustics. I did not include Depth of Field in this experiment because it didn't fit the scene well in my opinion.

The experiment

Click the image to start the rendering. Warning, this experiment is slow. Run it using a fast web browser (google chrome) and be a little bit patient. Opera won't work because it doesn't support web workers.

You might also want to check out my previous path tracing experiment.

Path tracing

Path tracing is a way of solving the rendering equation using monte carlo integration. It is a form of ray tracing. According to wikipedia 'Path tracing is the simplest, most physically-accurate and slowest rendering method'. Sounds like the perfect target for a javascript experiment!

Implementation

The implementation is basically an extended version of the path tracer in my last post. It is using html5 web workers to render the image on up to 4 cores without locking up the page. One of the cool things I came up with is jittering the points on the view plane for each pixel to get rid of aliasing. I also changed the color of the light to that of warm sunlight (~5400k). I probably got the fresnel equations wrong and generally things are quite rough. I'm not an expert in the field of raytracing and I probably did a lot of things in a suboptimal way. But feel free to have a look at the source code (~300 LOC) and ask questions. I think the code should be quite easy to understand.

Please write a comment if you've got anything to say. If you think my experiments are interesting, follow me on twitter or subscribe to my atom feed.