• Advertisement
Sign in to follow this  

DX11 Normal Interpolation issues on my generated terrain

This topic is 1821 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi all,
I am having a weird problem with normals on my generated terrain. I am not sure whether is it is shader or mesh issue, but here is how it looks:
 
 
 
 
Dx11FW%202013-01-22%2022-51-45-78.png
 
As you can see, I get this pattern along the edges of the triangles. This reminds me of per vertex shading  however I am aiming for per pixel shading.
 
Here are my vertex and pixel shaders:

 

VS:

cbuffer cbToProjection
{
	float4x4 matToProj;
}

struct VS_IN
{
	float4 Position : POSITION;
	float3 Normal: NORMAL;
};

struct VS_OUT
{
	float4 Position : SV_POSITION;
	float3 NormalWS: TEXCOORD1;
};

VS_OUT main(VS_IN IN)
{
	VS_OUT OUT;
	OUT.Position = mul(IN.Position,matToProj);
	OUT.NormalWS = normalize(IN.Normal);
	return OUT;
}

 

PS:

struct PS_IN
{
	float4 Position : SV_POSITION;
	float3 NormalWS: TEXCOORD1;
};

float4 main(PS_IN IN): SV_TARGET0 {
	float3 normal = normalize(IN.NormalWS);
	float3 toLight = normalize(float3(1,3,-2));
	float NDotL = saturate(dot(toLight,normal));
	float4 color = float4(1.0f,1.0f,1.0f,1.0f);	
	color.rgb *= NDotL;
	return color;
}

 

 

 

so what am I doing wrong? 

Share this post


Link to post
Share on other sites
Advertisement

From your shader code, it looks like you are doing this correctly (although you could remove the normalize call on the normal vector in the VS since you renormalize in the PS after rasterization).  My guess is that your terrain is defining three vertices for each triangle face, rather than one vertex at each grid point.  You can verify this by checking the number of vertices  you are passing in with your draw call, or you can also check this with PIX/Graphics Debugger to see how many primitives are generated from how many input vertices.

Share this post


Link to post
Share on other sites
From your shader code, it looks like you are doing this correctly (although you could remove the normalize call on the normal vector in the VS since you renormalize in the PS after rasterization).  My guess is that your terrain is defining three vertices for each triangle face, rather than one vertex at each grid point.  You can verify this by checking the number of vertices  you are passing in with your draw call, or you can also check this with PIX/Graphics Debugger to see how many primitives are generated from how many input vertices.

 

So, I moved the rendering into indexed rendering, so there is only one normal per vertex and still get same result. 

Share this post


Link to post
Share on other sites

Have you tried with different light directions? The dark areas simply look like they are dark because they are facing away from the parallel light you have hard-coded in your shader (ie. the dark sections are always on the -z axis in your image).

 

It may also be worth outputing the normal to the frame buffer so you can visually see any issues with the interpolation.

 

eg: change your color computation (in the pixel shader to) to:

 

color = normal * 0.5f + 0.5f;

return color;

 

You can also change your light vector so that it's always directly above the terrain (float3 toLight = float3( 0.0f, 1.0f, 0.0f );) which should give you a more even lighting across the terrain, again helping to see any issue with the normals. With the off-center light angle, it makes it difficult to say what is wrong really sorry :) But, certainly, your shader code looks fine. If it isn't the light direction confusing you, then it maybe the normals themselves.

 

My first guess is just that the light is at an angle really ;)

 

n!

Share this post


Link to post
Share on other sites

It could also be that I'm misunderstanding what you're complaining about.

 

Sometimes people build their terrain such that the vertices look like:

 

pW3fUDy.jpg

 

Whereas you can avoid some artifacts on terrain lighting if you structure your vertices like:

 

ZO7pxKa.jpg

 

n!

Share this post


Link to post
Share on other sites
From your shader code, it looks like you are doing this correctly (although you could remove the normalize call on the normal vector in the VS since you renormalize in the PS after rasterization).  My guess is that your terrain is defining three vertices for each triangle face, rather than one vertex at each grid point.  You can verify this by checking the number of vertices  you are passing in with your draw call, or you can also check this with PIX/Graphics Debugger to see how many primitives are generated from how many input vertices.

 

So, I moved the rendering into indexed rendering, so there is only one normal per vertex and still get same result. 

 

That may or may not mean that there is exactly one vertex normal being used at each grid point.  How many vertices are in your vertex buffer, how many indices in your index buffer, and how many primitives are you drawing?  Compare that with your grid size and make sure that you only have N+1 x N+1 vertices for a grid of size N x N.

Share this post


Link to post
Share on other sites

Have you tried with different light directions? The dark areas simply look like they are dark because they are facing away from the parallel light you have hard-coded in your shader (ie. the dark sections are always on the -z axis in your image).

 

It may also be worth outputing the normal to the frame buffer so you can visually see any issues with the interpolation.

 

eg: change your color computation (in the pixel shader to) to:

 

color = normal * 0.5f + 0.5f;

return color;

 

You can also change your light vector so that it's always directly above the terrain (float3 toLight = float3( 0.0f, 1.0f, 0.0f );) which should give you a more even lighting across the terrain, again helping to see any issue with the normals. With the off-center light angle, it makes it difficult to say what is wrong really sorry smile.png But, certainly, your shader code looks fine. If it isn't the light direction confusing you, then it maybe the normals themselves.

 

My first guess is just that the light is at an angle really ;)

 

n!

 

Setting the light to 0,1,0 doesn't help, the problem persists. Rendering the normals into the frame buffer shows the same problem.

 

 

 

It could also be that I'm misunderstanding what you're complaining about.

 

Sometimes people build their terrain such that the vertices look like:

 

pW3fUDy.jpg

 

Whereas you can avoid some artifacts on terrain lighting if you structure your vertices like:

 

ZO7pxKa.jpg

 

n!

 

I am building the mesh the way shown in the first picture, will try the other way that thanks.

 

 

From your shader code, it looks like you are doing this correctly (although you could remove the normalize call on the normal vector in the VS since you renormalize in the PS after rasterization).  My guess is that your terrain is defining three vertices for each triangle face, rather than one vertex at each grid point.  You can verify this by checking the number of vertices  you are passing in with your draw call, or you can also check this with PIX/Graphics Debugger to see how many primitives are generated from how many input vertices.

 

So, I moved the rendering into indexed rendering, so there is only one normal per vertex and still get same result. 

 

That may or may not mean that there is exactly one vertex normal being used at each grid point.  How many vertices are in your vertex buffer, how many indices in your index buffer, and how many primitives are you drawing?  Compare that with your grid size and make sure that you only have N+1 x N+1 vertices for a grid of size N x N.

 

For a 16x16 grid, there are 256 vertices, index buffer holds 1350 indices

Share this post


Link to post
Share on other sites

It could also be that I'm misunderstanding what you're complaining about.

 

Sometimes people build their terrain such that the vertices look like:

 

pW3fUDy.jpg

 

Whereas you can avoid some artifacts on terrain lighting if you structure your vertices like:

 

ZO7pxKa.jpg

 

n!

 

So, I tried building the mesh as in the second image, unfortunately the problems did not go away. 

 

16x16 mesh:

 

Dx11FW%202013-01-24%2022-50-58-53.png

 

normals:

 

Dx11FW%202013-01-24%2023-01-47-90.png

Share this post


Link to post
Share on other sites
Hey, thanks for the images. The lighting in the first one shows your problem much clearer for me.It looks like your normals are incorrect. For terrain normals I take the height sample x,y and compute the normal for that sample with:

float h0 = GetSample( x + 0, y - 1 );
float h1 = GetSample( x - 1, y + 0 );
float h2 = GetSample( x + 1, y + 0 );
float h3 = GetSample( x + 0, y + 1 );

Vector3 normal;

normal.x = h1 - h2;
normal.y = separation; // separation = distance between samples (I use 1.0f).
normal.z = h0 - h3;

normal.Normalize();

return normal;

If that doesn't help, perhaps posting your normal calculation code?

n! Edited by nfactorial

Share this post


Link to post
Share on other sites

Hey, thanks for the images. The lighting in the first one shows your problem much clearer for me.It looks like your normals are incorrect. For terrain normals I take the height sample x,y and compute the normal for that sample with:

float h0 = GetSample( x + 0, y - 1 );
float h1 = GetSample( x - 1, y + 0 );
float h2 = GetSample( x + 1, y + 0 );
float h3 = GetSample( x + 0, y + 1 );

Vector3 normal;

normal.x = h1 - h2;
normal.y = separation; // separation = distance between samples (I use 1.0f).
normal.z = h0 - h3;

normal.Normalize();

return normal;

If that doesn't help, perhaps posting your normal calculation code?

n!

 

Are you using y-up for your normals, but z-up for your sample locations in this response?

 

EDIT:  Also, I think your y value for the normal should be 2 * the separation.  According to this post, at least http://www.gamedev.net/topic/163625-fast-way-to-calculate-heightmap-normals/

Edited by Vexal

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
  • Advertisement
  • Popular Tags

  • Advertisement
  • Popular Now

  • Similar Content

    • By AxeGuywithanAxe
      I wanted to see how others are currently handling descriptor heap updates and management.
      I've read a few articles and there tends to be three major strategies :
      1 ) You split up descriptor heaps per shader stage ( i.e one for vertex shader , pixel , hull, etc)
      2) You have one descriptor heap for an entire pipeline
      3) You split up descriptor heaps for update each update frequency (i.e EResourceSet_PerInstance , EResourceSet_PerPass , EResourceSet_PerMaterial, etc)
      The benefits of the first two approaches is that it makes it easier to port current code, and descriptor / resource descriptor management and updating tends to be easier to manage, but it seems to be not as efficient.
      The benefits of the third approach seems to be that it's the most efficient because you only manage and update objects when they change.
    • By evelyn4you
      hi,
      until now i use typical vertexshader approach for skinning with a Constantbuffer containing the transform matrix for the bones and an the vertexbuffer containing bone index and bone weight.
      Now i have implemented realtime environment  probe cubemaping so i have to render my scene from many point of views and the time for skinning takes too long because it is recalculated for every side of the cubemap.
      For Info i am working on Win7 an therefore use one Shadermodel 5.0 not 5.x that have more options, or is there a way to use 5.x in Win 7
      My Graphic Card is Directx 12 compatible NVidia GTX 960
      the member turanszkij has posted a good for me understandable compute shader. ( for Info: in his engine he uses an optimized version of it )
      https://turanszkij.wordpress.com/2017/09/09/skinning-in-compute-shader/
      Now my questions
       is it possible to feed the compute shader with my orignial vertexbuffer or do i have to copy it in several ByteAdressBuffers as implemented in the following code ?
        the same question is about the constant buffer of the matrixes
       my more urgent question is how do i feed my normal pipeline with the result of the compute Shader which are 2 RWByteAddressBuffers that contain position an normal
      for example i could use 2 vertexbuffer bindings
      1 containing only the uv coordinates
      2.containing position and normal
      How do i copy from the RWByteAddressBuffers to the vertexbuffer ?
       
      (Code from turanszkij )
      Here is my shader implementation for skinning a mesh in a compute shader:
      1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 struct Bone { float4x4 pose; }; StructuredBuffer<Bone> boneBuffer;   ByteAddressBuffer vertexBuffer_POS; // T-Pose pos ByteAddressBuffer vertexBuffer_NOR; // T-Pose normal ByteAddressBuffer vertexBuffer_WEI; // bone weights ByteAddressBuffer vertexBuffer_BON; // bone indices   RWByteAddressBuffer streamoutBuffer_POS; // skinned pos RWByteAddressBuffer streamoutBuffer_NOR; // skinned normal RWByteAddressBuffer streamoutBuffer_PRE; // previous frame skinned pos   inline void Skinning(inout float4 pos, inout float4 nor, in float4 inBon, in float4 inWei) {  float4 p = 0, pp = 0;  float3 n = 0;  float4x4 m;  float3x3 m3;  float weisum = 0;   // force loop to reduce register pressure  // though this way we can not interleave TEX - ALU operations  [loop]  for (uint i = 0; ((i &lt; 4) &amp;&amp; (weisum&lt;1.0f)); ++i)  {  m = boneBuffer[(uint)inBon].pose;  m3 = (float3x3)m;   p += mul(float4(pos.xyz, 1), m)*inWei;  n += mul(nor.xyz, m3)*inWei;   weisum += inWei;  }   bool w = any(inWei);  pos.xyz = w ? p.xyz : pos.xyz;  nor.xyz = w ? n : nor.xyz; }   [numthreads(1024, 1, 1)] void main( uint3 DTid : SV_DispatchThreadID ) {  const uint fetchAddress = DTid.x * 16; // stride is 16 bytes for each vertex buffer now...   uint4 pos_u = vertexBuffer_POS.Load4(fetchAddress);  uint4 nor_u = vertexBuffer_NOR.Load4(fetchAddress);  uint4 wei_u = vertexBuffer_WEI.Load4(fetchAddress);  uint4 bon_u = vertexBuffer_BON.Load4(fetchAddress);   float4 pos = asfloat(pos_u);  float4 nor = asfloat(nor_u);  float4 wei = asfloat(wei_u);  float4 bon = asfloat(bon_u);   Skinning(pos, nor, bon, wei);   pos_u = asuint(pos);  nor_u = asuint(nor);   // copy prev frame current pos to current frame prev pos streamoutBuffer_PRE.Store4(fetchAddress, streamoutBuffer_POS.Load4(fetchAddress)); // write out skinned props:  streamoutBuffer_POS.Store4(fetchAddress, pos_u);  streamoutBuffer_NOR.Store4(fetchAddress, nor_u); }  
    • By mister345
      Hi, can someone please explain why this is giving an assertion EyePosition!=0 exception?
       
      _lightBufferVS->viewMatrix = DirectX::XMMatrixLookAtLH(XMLoadFloat3(&_lightBufferVS->position), XMLoadFloat3(&_lookAt), XMLoadFloat3(&up));
      It looks like DirectX doesnt want the 2nd parameter to be a zero vector in the assertion, but I passed in a zero vector with this exact same code in another program and it ran just fine. (Here is the version of the code that worked - note XMLoadFloat3(&m_lookAt) parameter value is (0,0,0) at runtime - I debugged it - but it throws no exceptions.
          m_viewMatrix = DirectX::XMMatrixLookAtLH(XMLoadFloat3(&m_position), XMLoadFloat3(&m_lookAt), XMLoadFloat3(&up)); Here is the repo for the broken code (See LightClass) https://github.com/mister51213/DirectX11Engine/blob/master/DirectX11Engine/LightClass.cpp
      and here is the repo with the alternative version of the code that is working with a value of (0,0,0) for the second parameter.
      https://github.com/mister51213/DX11Port_SoftShadows/blob/master/Engine/lightclass.cpp
    • By mister345
      Hi, can somebody please tell me in clear simple steps how to debug and step through an hlsl shader file?
      I already did Debug > Start Graphics Debugging > then captured some frames from Visual Studio and
      double clicked on the frame to open it, but no idea where to go from there.
       
      I've been searching for hours and there's no information on this, not even on the Microsoft Website!
      They say "open the  Graphics Pixel History window" but there is no such window!
      Then they say, in the "Pipeline Stages choose Start Debugging"  but the Start Debugging option is nowhere to be found in the whole interface.
      Also, how do I even open the hlsl file that I want to set a break point in from inside the Graphics Debugger?
       
      All I want to do is set a break point in a specific hlsl file, step thru it, and see the data, but this is so unbelievably complicated
      and Microsoft's instructions are horrible! Somebody please, please help.
       
       
       

    • By mister345
      I finally ported Rastertek's tutorial # 42 on soft shadows and blur shading. This tutorial has a ton of really useful effects and there's no working version anywhere online.
      Unfortunately it just draws a black screen. Not sure what's causing it. I'm guessing the camera or ortho matrix transforms are wrong, light directions, or maybe texture resources not being properly initialized.  I didnt change any of the variables though, only upgraded all types and functions DirectX3DVector3 to XMFLOAT3, and used DirectXTK for texture loading. If anyone is willing to take a look at what might be causing the black screen, maybe something pops out to you, let me know, thanks.
      https://github.com/mister51213/DX11Port_SoftShadows
       
      Also, for reference, here's tutorial #40 which has normal shadows but no blur, which I also ported, and it works perfectly.
      https://github.com/mister51213/DX11Port_ShadowMapping
       
  • Advertisement