uniform float4x4 WorldViewProj;
uniform float4 prmAtlasTilesHalfPixel;
uniform sampler2D filteredAtlasSampler : register(s0);
uniform sampler2D clusterWeightIndexSampler : register(s1);
uniform sampler2D clusterWeightSampler : register(s2);
uniform sampler2D colorSampler : register(s3);
uniform float allClusterCount; 
uniform float clusterCount;

struct vsInputWalk
{
    float4 pos			: POSITION;
    float2 tex			: TEXCOORD0;
    float2 texAtlas		: TEXCOORD1;
};

struct vsOutputWalk
{
    float4 pos			: POSITION;
    float2 tex			: TEXCOORD0;
    float2 texAtlas		: TEXCOORD1;
};


vsOutputWalk vsWalk(vsInputWalk input)
{
	vsOutputWalk output = (vsOutputWalk)0;
	output.pos = mul(WorldViewProj, input.pos);
	output.tex = input.tex;
	output.texAtlas = input.texAtlas;	
	return output;
}

float4	psWalk(vsOutputWalk input) : COLOR
{
	float2 texAtlas = input.texAtlas;
	
	int2 prmAtlasTiles = prmAtlasTilesHalfPixel.xy;
	float2 atlasHalfPixel = prmAtlasTilesHalfPixel.zw;
	
	float3 col = 0;
	for(int iCluster = 0; iCluster < 32; iCluster++)
	{
		float2 prmTexPos = float2(
			(texAtlas.x + (iCluster % prmAtlasTiles.x)) / prmAtlasTiles.x,
			1.0 - (texAtlas.y + (iCluster / prmAtlasTiles.x)) / prmAtlasTiles.y) + atlasHalfPixel;

		float weightIndex = tex2D(clusterWeightIndexSampler, float2(((float)iCluster + 0.5) / clusterCount, 0.5)).r;
		float3 weight = tex2D(clusterWeightSampler, float2((weightIndex + 0.5)/allClusterCount, 0.5)).rgb;
		float3 val = tex2D(filteredAtlasSampler, prmTexPos).xyz;
		//if(length(val - float3(1,1,1))== 0)
		//val = 0;
		col += val.xyz * weight;	
		//col = 0.00000000001 * col + float3(prmTexPos,1);
	}
	
	float3 color = tex2D(colorSampler, input.tex).rgb;

//return float4(col,1);
//return float4(color + 0.0000000001 * col, 1)	;
return float4(col * color, 1);


//return float4(col,1) * 0.0000000000001 + tex2D(filteredAtlasSampler, float2(input.texAtlas.x / 32.0, input.texAtlas.y));
}