ok here is my shader code so far:

Code:

---

 material mat_diffuseperpixel
{
	skin2 = zombiebump;
	flags = tangent;
	effect =
	"
		texture texSkin1;
		texture texSkin2;
		matrix matWorldViewProj;
		matrix matWorld;
		matrix matWorldView;
		vector vecSkill41;
		vector vecFog;
		
		technique diffuseperpixel
		{
			pass p0
			{
				Texture[0]=<texSkin1>;
				Texture[1]=<texSkin2>;

				VertexShaderConstant[0]=<matWorldViewProj>;
				VertexShaderConstant[4]=<matWorld>;
				VertexShaderConstant[8]=<matWorldView>;
				VertexShaderConstant[16]=<vecSkill41>;	
				VertexShaderConstant[17]=<vecFog>;
				VertexShaderConstant[95]=(0.5f,0.0f,0.0f,0.0f);	


				VertexShader=
				decl
				{
					stream 0;
					float v0[3]; //position
					float v3[3]; //normal
					float v7[3]; //uv
					float v8[3]; //tangent
				}
				asm
				{
					vs.1.1
					
					m4x4 oPos,v0,c0
					mov oT0,v7
					mov oT1,v7

					m3x3 r5,v8,c4				// transform tangent to world space
					m3x3 r7,v3,c4				// transform normal to world space

					mul r0,r5.zxyw,-r7.yzxw			// tangent x normal -> binormal
					mad r6,r5.yzxw,-r7.zxyw,-r0
					
			//		dp3 r8.x,r5,-c16			// transform light vector to texture space
			//		dp3 r8.y,r6,-c16
			//		dp3 r8.z,r7,-c16
			//		mad oD0.xyz,r8.xyz,c95.x,c95.x
				
				//new code for point lights
				
				m4x4 r10,v0,c4 // transform position to world space 
				sub r10,v0,c16 // lightdir=vertexpos-lightpos 
				dp3 r10.w,r10,r10 // renormalize it 
				rsq r10.w,r10.w 
				mul r10,r10,r10.w 

				dp3 r8.x,r5,r10 // transform light vector to texture space 
				dp3 r8.y,r6,r10 
				dp3 r8.z,r7,r10 
				mad oD0.xyz,r8.xyz,c95.x,c95.x 


					mov r1.w,c17.w				// r1.w=1
					dp4 r0,v0,c10 				// distance to camera position
					add r0,r0,-c17.x 			// distance-fog_start
					mad r0.x,-r0.x,c17.z,r1.w 		// 1-(distance-fog_start)*(1/(fog_end-fog_start))
					max oFog.x,r0.x,c95.w			// clamp with custom max value		
			
				};
			
				PixelShader= 
				asm
				{
					ps.1.3
					tex t0				// sample colormap
					tex t1				// sample normalmap
					dp3_sat r1,t1_bx2,v0_bx2		// normal . light
					
					
					mul r0,t0,r1			// modulate against base color
					
				};
			}
		}
	";
} 

---

here is the function i use to get the ight positions:
Code:

---

 function find_nearest_light()

{
	temp.pan =360;
	temp.tilt=360;
	temp.z=3000;
	you=scan_entity(my.x,temp);
	
	if your.light==ON
	{
	
		temp.x = your.x;
		temp.y = your.y;
		temp.z = your.z;
		my.skill41=float(temp.x); 
		my.skill42=float(temp.y); 
		my.skill43=float(temp.z); 
		my.skill44=float(1); 
		
		return ;	
	}
} 

---

its does appear to work.. buts it difficult to tell exactly.. i cant tell if its getting the light positions always.. or if it just samples it once.. or if my nearest light code just alwasy returns 0 for everything.. which of course would make the light source always 0,0,0 ..sigh
im not a programmer.. just artist.. im trying to muddle through tho

seems to be working...maybe you can help me with a problem... my texture is like this:


here is the code i'm using...just this applied as an action.
EDIT!: I replaced this code with the fixed version.

Code:

---

 //Matt_Aufderheide's modified ventilator shader

//Important! generate a normal map from your greyscale hightmap. You can do this via ATI's "TGAtoDOT3" tool

// it uses the models regular map as a clor map and the normal map as a bump

//here is the function i use to get the ight positions:



bmap zombiebump=<bumpmapDOT3.tga>;//normal map



function find_nearest_light()



{

	temp.pan =360;

	temp.tilt=360;

	temp.z=3000;

	you=scan_entity(my.x,temp);

	

	if your.light==ON

	{

	

		temp.x = your.x;

		temp.y = your.y;

		temp.z = your.z;

		my.skill41=float(temp.x); 

		my.skill42=float(temp.y); 

		my.skill43=float(temp.z); 

		my.skill44=float(1); 

		

		return ;	

	}

} 





 material mat_bumpmapping

{

	skin2 = zombiebump;

	flags = tangent;

	effect =

	"

		texture texSkin1;

		texture texSkin2;

		matrix matWorldViewProj;

		matrix matWorld;

		matrix matWorldView;

		vector vecSkill41;

		vector vecFog;

		

		technique bumpmapping

		{

			pass p0

			{

				Texture[0]=<texSkin1>;

				Texture[1]=<texSkin2>;



				VertexShaderConstant[0]=<matWorldViewProj>;

				VertexShaderConstant[4]=<matWorld>;

				VertexShaderConstant[8]=<matWorldView>;

				VertexShaderConstant[16]=<vecSkill41>;	

				VertexShaderConstant[17]=<vecFog>;

				VertexShaderConstant[95]=(0.5f,0.0f,0.0f,0.0f);	





				VertexShader=

				decl

				{

					stream 0;

					float v0[3]; //position

					float v3[3]; //normal

					float v7[3]; //uv

					float v8[3]; //tangent

				}

				asm

				{

					vs.1.1

					

					m4x4 oPos,v0,c0

					mov oT0,v7

					mov oT1,v7



					m3x3 r5,v8,c4				// transform tangent to world space

					m3x3 r7,v3,c4				// transform normal to world space



					mul r0,r5.zxyw,-r7.yzxw			// tangent x normal -> binormal

					mad r6,r5.yzxw,-r7.zxyw,-r0

					

			//		dp3 r8.x,r5,-c16			// transform light vector to texture space

			//		dp3 r8.y,r6,-c16

			//		dp3 r8.z,r7,-c16

			//		mad oD0.xyz,r8.xyz,c95.x,c95.x

				

				//new code for point lights

				

				m4x4 r10,v0,c4 // transform position to world space 

				sub r10,v0,c16 // lightdir=vertexpos-lightpos 

				dp3 r10.w,r10,r10 // renormalize it 

				rsq r10.w,r10.w 

				mul r10,r10,r10.w 



				dp3 r8.x,r5,r10 // transform light vector to texture space 

				dp3 r8.y,r6,r10 

				dp3 r8.z,r7,r10 

				mad oD0.xyz,r8.xyz,c95.x,c95.x 





					mov r1.w,c17.w				// r1.w=1

					dp4 r0,v0,c10 				// distance to camera position

					add r0,r0,-c17.x 			// distance-fog_start

					mad r0.x,-r0.x,c17.z,r1.w 		// 1-(distance-fog_start)*(1/(fog_end-fog_start))

					max oFog.x,r0.x,c95.w			// clamp with custom max value		

			

				};

			

				PixelShader= 

				asm

				{

					ps.1.3

					tex t0				// sample colormap

					tex t1				// sample normalmap

					dp3_sat r1,t1_bx2,v0_bx2		// normal . light

					

					

					mul r0,t0,r1			// modulate against base color

					

				};

			}

		}

	";

}



action Shaders_Bump

{ 

//bmap_to_normals(bmap_for_entity(my,10),16); //1,16 convert the bumpmap to a normal map with scale factor 16 

my.material=mat_bumpmapping; 

//... // causes error but leave it here

my.pan=random(360); //rotate for testing

my.tilt=random(360); //rotate for testing

my.roll=random(360); //rotate for testing

while(1) { //rotate for testing

my.pan+=sin(my.tilt)/2*time; //rotate for testing

my.tilt+=cos(my.roll)/2*time; //rotate for testing

my.roll+=sin(my.pan)/2*time; //rotate for testing

wait(1); //rotate for testing

} 

} 

 

---

ok i had the same problem using the bmap_to_normals command. Instead.. generate a normal map from your greyscale hightmap. You can do this via a plugin or a standalone app.. check ATI website, they have a good tool. once you have the normal map then you can just use it as is.. no need to invoke bmap_to_normals..
then it should work.

EXCELLENT!!! thanks... I use ATI's "TGAtoDOT3",
commented out the bmap_to_normals line,and replace the name of the targa being called... I edited the above code in case anyone wanted to copy&paste it without the need to edit..

works great!! seems to be a real bump map!
A good way to test is to use the object as a player, switch to orbit mode (F7), and rotate around the player. You really get to see the Skin change as the light rolls across the surface.

whats the best way to control the amount? tweaking the contrast of the texture or in code?

the amount of the hightmapping?.. yeah eitehr contrast on the hightmap/normal map or i think there is a way to bias the dotproduct.. but i dont know the assembly

thanks...one last question (hopefully
Is there a way to have it also react to realtime light code attached to entites? light code like this:
my.lightrange = 100; // RANGE
my.lightred = 100;// 85
my.lightgreen = 100;//22
my.lightblue = 100;//22

yes sure but damned if i know :P
im waiting for ventilator to comeback so i can ask a few more questions...

like first: for rgb color of the lights.. i have no clue what to do in the assembly..

all i want is to use different normal maps for different entities.. not differnt ones per the same models...see. maybe your script already can do that..but i dont see how. I dont want to define differetn materials for each model.. just use the same mat_diffuse. How does it know which model/skin gets which normal map?
instead of defining skin2=<wahtever.tga> as the normal map.. how do you make the shader know to use differnt normal maps .. maybe im being dense.. thank you for your patience.
my first example does exactly that. it uses the second texture (which you assigned in med) of the current model as normalmap. don't assign skin2 in the material.

Code:

---

action ball // the action of your entity

{

  //bmap_to_normals(bmap_for_entity(my,1),16); // you don't need this line if you already generated the normalmap with an external tool

  my.material=mat_diffuseperpixel;

  ...

}



material mat_diffuseperpixel

{

  flags=tangent;

  effect=

  "

    texture texSkin1;

    texture texSkin2;

    matrix matWorldViewProj;

    matrix matWorld;

    matrix matWorldView;

    vector vecSkill41;

    vector vecFog;



    technique diffuseperpixel

    {

      pass p0

      {

        Texture[0]=<texSkin1>; // colormap (texture 1 of 2 in med's skin editor)

        Texture[1]=<texSkin2>; // normalmap (texture 2 of 2 in med's skin editor) make sure you use a 24bit tga bitmap

---

.........
sub r10,v0,c16 // lightdir=vertexpos-lightpos

this line should be:

sub r10,r10,c16 // lightdir=vertexpos-lightpos

i corrected this 5 minutes after i posted the code but you were faster!
if the model space vertex position gets used, the light direction won't be correct. but in the current beta this doesn't matter much. the light direction will behave strange anyway because of the wrong normals...

the amount of the hightmapping?.. yeah eitehr contrast on the hightmap/normal map or i think there is a way to bias the dotproduct.. but i dont know the assembly
i don't think you should manipulate the normalmap in any way (for example giving it more contrast). the normalmap's normalvectors won't be well behaved (normalized) anymore. the photoshop and gimp plugins provide amplification parameters which get used when the bumpmap gets converted to a normalmap. you should do it this way. if you convert the bumpmap to normalmap with bmap_to_normals you can scale the bumps with the second parameter.

im not a programmer.. just artist..
i am more artist than programmer too!

here is the complete shader again:

Code:

---

action ball // your entity's action

{	

	//bmap_to_normals(bmap_for_entity(my,1),16); // don't use this if you already have a normalmap

	my.material=mat_diffuseperpixel;

	

	while(1)

	{	

		my.skill41=float(sun_pos.x); // use your own lightpos here

		my.skill42=float(sun_pos.y);

		my.skill43=float(sun_pos.z);

		my.skill44=float(1);

		wait(1);

	}

	...



material mat_diffuseperpixel

{

	flags=tangent;

	effect=

	"

		texture texSkin1;

		texture texSkin2;

		matrix matWorldViewProj;

		matrix matWorld;

		matrix matWorldView;

		vector vecSkill41;

		vector vecFog;

		

		technique diffuseperpixel

		{

			pass p0

			{

				Texture[0]=<texSkin1>; // colormap (texture 1 of 2 in med's skin editor)

				Texture[1]=<texSkin2>; // normalmap (texture 2 of 2 in med's skin editor) make sure you use a 24bit tga bitmap



				VertexShaderConstant[0]=<matWorldViewProj>;

				VertexShaderConstant[4]=<matWorld>;

				VertexShaderConstant[8]=<matWorldView>;

				VertexShaderConstant[16]=<vecSkill41>;

				VertexShaderConstant[17]=<vecFog>;

				VertexShaderConstant[95]=(0.5f,0.0f,0.0f,0.0f);	



				VertexShader=

				decl

				{

					stream 0;

					float v0[3]; //position

					float v3[3]; //normal

					float v7[3]; //uv

					float v8[3]; //tangent

				}

				asm

				{

					vs.1.1

					

					m4x4 oPos,v0,c0				// transform vertexpos to clip space

					mov oT0,v7				// output uvs

					mov oT1,v7				// output uvs



					m3x3 r5,v8,c4				// transform tangent to world space

					m3x3 r7,v3,c4				// transform normal to world space

					

					mul r0,r5.zxyw,-r7.yzxw			// tangent x normal -> binormal

					mad r6,r5.yzxw,-r7.zxyw,-r0

					

					m4x4 r10,v0,c4				// transform vertexpos to world space

					sub r10,r10,c16				// lightdir=vertexpos-lightpos

					dp3 r10.w,r10,r10			// renormalize it

					rsq r10.w,r10.w

					mul r10,r10,r10.w

					

					dp3 r8.x,r5,r10				// transform lightdir to texture space

					dp3 r8.y,r6,r10

					dp3 r8.z,r7,r10

					mad oD0.xyz,r8.xyz,c95.x,c95.x



					mov r1.w,c17.w				// r1.w=1

					dp4 r0,v0,c10 				// distance to camera position

					add r0,r0,-c17.x 			// distance-fog_start

					mad r0.x,-r0.x,c17.z,r1.w 		// 1-(distance-fog_start)*(1/(fog_end-fog_start))

					max oFog.x,r0.x,c95.w			// clamp with custom max value

				};

			

				PixelShader= 

				asm

				{

					ps.1.3

					tex t0				// sample colormap

					tex t1				// sample normalmap

					dp3 r1,t1_bx2,v0_bx2		// normal . light

					mul r0,t0,r1			// modulate against color

				};

			}

		}

	";

}

---

ahh.. ok i now i get it... i didnt realize you could add mulitple skins in med.. lol.

ok your point light code seems to work correctly. Great work vent!
Maybe could you point me in the right direction about where in the assembly you add information about light rgb values?