===================================
SOME LINKS 
===================================

Start here:
http://www.adobe.com/devnet/flash/articles/pixel_bender_basics_03.html

Adobe Labs page
http://labs.adobe.com/technologies/pixelbender/

Twitter feed
http://twitter.com/pixelbender

Pixel Bender Exchange
http://www.adobe.com/cfusion/exchange/index.cfm?event=productHome&exc=26

Forum
http://www.adobe.com/cfusion/webforums/forum/categories.cfm?forumid=72&catid=661&entercat=y

MAX presentation 11/2008
http://tv.adobe.com/#vi+f15383v1072

Adobe TV video
http://tv.adobe.com/#vi+f1552v1003

Embedding into SWF
http://www.mikechambers.com/blog/2008/09/08/embedding-pixel-bender-filters-within-a-swf/

Apply to video..
http://www.brooksandrus.com/blog/2009/01/19/pixel-bender-effects-video-killer-runtime-effects/

Fuckin color keying
http://www.video-flash.de/index/flash-player-10-pixel-bender-realtime-color-keying/

Simple example, math only
http://www.boostworthy.com/blog/?p=243

Audio mixing
http://www.kaourantin.net/2008/10/audio-mixing-with-pixel-bender.html

Endian-ness is opposite:
http://www.adobe.com/cfusion/webforums/forum/messageview.cfm?forumid=72&catid=675&threadid=1373666&enterthread=y

===================================
SOME EXPOSITION...
===================================

What about the 4th? Well, as you can see the ones in the list are designed 
for graphics only. The last one is more powerful than that. Instead of 
targeting a specific graphics primitive in the Flash Player, you can target 
BitmapData objects, ByteArrays or Vectors. Not only that but if you use 
ByteArray or Vector the data you handle are 32-bit floating point numbers 
for each channel, unlike BitmapData which is limited to 8-bit unsigned 
integers per channel. In the end this means you can use Pixel Bender 
kernels to not only do graphics stuff, but generic number crunching. 
If you can accept the 32-bit floating point limitation.

# The internal RGB color space of the Flash Player is alpha 
pre-multiplied and that is what the Pixel Bender kernel gets.

# Output color values are always clamped against the alpha. This is not 
the case when the output is a ByteArray or Vector.

# The maximum native JIT code buffer size for a kernel is 32KB, if you 
hit this limit which can happen with complex filters the Flash Player 
falls back to interpreted mode mode like it does in all cases on 
PowerPC based machines.

# Math functions apart from simple multiplication, division, addition 
and subtracting work slightly differently on different platforms, 
depending on the C-library implementation or CPU.

In the Pixel Bender model, images do not have a size. Instead, each 
image is thought of as being defined over an infinite plane of discrete
 pixel coordinates. 

This Pixel Bender coordinate model has implications for filter design. 
For example, it is not possible to write a Pixel Bender kernel that 
reflects an image around its “center,” because the center is undefined. 
Instead, the coordinates of the circle must be denoted explicitly by 
passing them in as kernel parameters.

Pixel Bender images have 32 bits per channel, but graphics in Flash 
Player 10 have only 8 bits per channel. When a kernel is run in Flash 
Player, the input image data is converted to 32 bits per channel
and then converted back to 8 bits per channel when kernel execution is 
complete.

No arrays for Flash.
No dependent values.

===================================
PBK / TOOLKIT NOTES
===================================

Toolkit

	Make sure to check Build > Enable Actionscript warnings

File types

	pbk - source code
	pbj - abc
	pbg - graph code

Types

	float
	int
	bool
	pixel (represented internally as a float)

	image1/2/3/4 - represents an entire image with 1,2,3 or 4 channels

	No arrays.

Vectors

	Above Types can be declared as 2,3 or 4 element vectors (eg, "float2", "pixel4")

	Declared like this: float3(0.0, 0.5, 1.0);

	Accessed like this: myFloat[0], or myPixel.r

	Also,	r,g,b,a == 0,1,2,3
		x,y,z,w == 0,1,2,3
		s,t,p,q == 0,1,2,3


	"Swizzing":

		float3 noAlpha = vec4.rgb;
		float4 reversed = vec4.abgr;
		vec2 = vec4.xz;

Casting

	Works between bool, int, and float:
	Eg, "bool(myFloat)"


- Declaring the input and output variables

	input image4 src;
	output pixel4 dst;

- Declaring parameters (AS-controlled variables)

	parameter float exposure;

- Declaring parameters for 'UI'

    parameter float red
    < minValue:float(0);
      maxValue:float(1);
      defaultValue:float(0.0);
    >;

- Declaring variables

	float4 inputColor = sampleNearest(src, outCoord());

- Declaring constants

	const pixel3 color1 = pixel3(1,0,0);

- Main function

	void evaluatePixel()
	{
	}

- Flow control

	if else, that's it

- Functions

	outCoord():float2 - the current coordinates

	sampleNearest(imageN, float2 f):pixelN - nearest pixel to f
	sampleLinear(imageN, float2 f):pixelN  - nearest pixel to f, bilinear

- Math functions

	pow, exp, exp2, log, log2, sqrt
	sin, cos, tan, asin, acos, atan
	abs, sine, ceil, floor, fract, mod, 
	step, clamp, mix, smoothStep
	radians, degrees
	length, distance, dot, cross, normalize
	matrix stuff...

- Region functions

	...


================================
USE IN ACTIONSCRIPT
================================

	Example:

		[Embed(source="alpha.pbj", mimeType="application/octet-stream")]
		private var PbjAlpha:Class;

		_shader = new Shader(new PbjAlpha());
		_shaderFilter = new ShaderFilter(_shader);
		_sprite.filters = [ _shaderFilter ];

	Feeding extra inputs to the shader:

		They live in "_shader.data.*". Are of type "ShaderInput". Eg:

			_shader.data.myInputVariable.input = a BitmapData or ByteArray
			_shader.data.myInputVariable.width/height - need to be specified only for ByteArray inputs
			_shader.data.myInputVariable.channels = 3 or 4; // read only?
			_shader.data.myInputVariable.name = "myInputVariable"; // read only?
			_shader.data.myInputVariable.index = 0, 1, 2, ...; // read only?

		The first input in a shader is (always?)(by default?) the object 
		that the shaderfilter is applied to.

	Using parameters declared in the PBK

		Found in "_shader.data.*" also. Of type "ShaderParameter". Eg:

			_shader.data.myParam.defaultValue/minValue/maxValue - 
				Array of 1 or more elements; 
				defined by PBK metadata
			_shader.data.myParam.value - Array
			_shader.data.myParam.type = "float"
			_shader.data.myParam.index

		!important: 
			After changing any parameter, you have to reapply the shader:
			"s.filters = [ _shaderFilter ];"

	Using ShaderJob:

		_shaderJob = new ShaderJob(_shader, _outputBitmap, _width, _height);
		_shaderJob.addEventListener(Event.COMPLETE, onShaderJobComplete);

		protected function onShaderJobComplete(event:Event):void 
		{
			// do stuff with "_outputBitmap" or event.bitmapData or event.byteArray, etc.
		}