• Advertisement
Sign in to follow this  

DX11 Instancing

This topic is 2154 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

Hey.

I have been following [url="http://rastertek.com/dx11tut37.html"]this tutorial[/url] , but while experimenting I reached a dead end. The addition I was trying to implement is a dynamic updating rotation per instance.

So first, i define

[CODE]
// TriangleTest.h
struct InstanceType
{
D3DXVECTOR3 position;
D3DXVECTOR3 rotation;
};
[/CODE]
for my Instance buffer Type, which leads me to define

[CODE]
// vertexshader.vs
struct VertexInputType
{
float3 position : POSITION;
float4 color : COLOR;
float3 instancePosition : POSITION1;
float3 instanceRotation : POSITION2;
};
[/CODE]
in my vertex Shader file. Note that I have changed the usage of semantic type TEXCOORD from the tutorial, as I feel more comfortable having positions use POSITION, as well as giving the position a [b]float3 type[/b].

I set up my InputElementDescription array like this

[CODE]
//shaderclass.cpp
polygonLayout[0].SemanticName = "POSITION";
polygonLayout[0].SemanticIndex = 0;
polygonLayout[0].Format = DXGI_FORMAT_R32G32B32_FLOAT;
polygonLayout[0].InputSlot = 0;
polygonLayout[0].AlignedByteOffset = 0;
polygonLayout[0].InputSlotClass = D3D11_INPUT_PER_VERTEX_DATA;
polygonLayout[0].InstanceDataStepRate = 0;
polygonLayout[1].SemanticName = "COLOR";
polygonLayout[1].SemanticIndex = 0;
polygonLayout[1].Format = DXGI_FORMAT_R32G32B32A32_FLOAT;
polygonLayout[1].InputSlot = 0;
polygonLayout[1].AlignedByteOffset = D3D11_APPEND_ALIGNED_ELEMENT;
polygonLayout[1].InputSlotClass = D3D11_INPUT_PER_VERTEX_DATA;
polygonLayout[1].InstanceDataStepRate = 0;
polygonLayout[2].SemanticName = "POSITION";
polygonLayout[2].SemanticIndex = 1;
polygonLayout[2].Format = DXGI_FORMAT_R32G32B32_FLOAT;
polygonLayout[2].InputSlot = 1;
polygonLayout[2].AlignedByteOffset = 0;
polygonLayout[2].InputSlotClass = D3D11_INPUT_PER_INSTANCE_DATA;
polygonLayout[2].InstanceDataStepRate = 1;
polygonLayout[3].SemanticName = "POSITION";
polygonLayout[3].SemanticIndex = 2;
polygonLayout[3].Format = DXGI_FORMAT_R32G32B32_FLOAT;
polygonLayout[3].InputSlot = 1;
polygonLayout[3].AlignedByteOffset = D3D11_APPEND_ALIGNED_ELEMENT;
polygonLayout[3].InputSlotClass = D3D11_INPUT_PER_INSTANCE_DATA;
polygonLayout[3].InstanceDataStepRate = 1;
[/CODE]

and create the input layouts.
In my model class (TriangeTest), i use this to update the dynamic instance buffer ( completely created as D3D11_USAGE_DYNAMIC, with D3D11_CP_ACCESS_WRITE)

[CODE]
void TriangleTest::Frame(ID3D11DeviceContext* deviceContext) {
for (int i = 0; i < this->i_instanceCount; i++) {
D3DXVec3Add(&(m_instanceList[i].rotation), &(m_instanceList[i].rotation), &D3DXVECTOR3(1,0,0.0f));
}
UpdateBuffers(deviceContext);
}
void TriangleTest::UpdateBuffers(ID3D11DeviceContext* deviceContext) {
HRESULT result;
D3D11_MAPPED_SUBRESOURCE mappedResource;
InstanceType* instancePtr;

result =deviceContext->Map(m_instanceBuffer, 0, D3D11_MAP_WRITE_DISCARD, 0, &mappedResource);
if(FAILED(result)) throw CException("Could not remap instance buffer");
instancePtr = (InstanceType*)mappedResource.pData;
memcpy(instancePtr, (void*)m_instanceList, (sizeof(InstanceType) * i_instanceCount));
deviceContext->Unmap(m_instanceBuffer, 0);
}
[/CODE]
, where m_instanceList is an array of InstanceType holding all information on the position and rotation per triangle. Of course, theFrame(...) function gets called once per frame.

Then I use the shader as provided and replace instancePosition with [b]instanceRotation[/b],which should make my triangles slide out to the right.

[CODE]
PixelInputType TextureVertexShader(VertexInputType input)
{
PixelInputType output;

// Change the position vector to be 4 units for proper matrix calculations.
input.position.w = 1.0f;
Here is where we use the instanced position information to modify the position of each triangle we are drawing.
// Update the position of the vertices based on the data for this particular instance.
input.position.x += input.[b]instanceRotation[/b].x;
input.position.y += input.[b]instanceRotation[/b].y;
input.position.z += input.[b]instanceRotation[/b].z;
// Calculate the position of the vertex against the world, view, and projection matrices.
output.position = mul(input.position, worldMatrix);
output.position = mul(output.position, viewMatrix);
output.position = mul(output.position, projectionMatrix);

// Store the texture coordinates for the pixel shader.
output.tex = input.tex;

return output;
}
[/CODE]

But when I run it (complete without errors), the triangles stay still on the screen. Even when using [b]instancePosition[/b] again. They move however, if I change the semantic type of inputPosition to TEXCOORD and update the position in the Frame procedure, but even then I get no motion from the instanceRotation.

Which leads me to think, that my InputElementLayoutDescription is wrong. But how? I have tried many a combination already, but don't understand.

Can anyone enlighten me with their ideas?

Share this post


Link to post
Share on other sites
Advertisement
I'm taking a bit of a guess here but it might be because your first semantic usage of POSITION is not POSITION0 (zero) which is causing the binding to fail for some reason; try giving it the semantic POSITION0 and see if that helps.

Share this post


Link to post
Share on other sites
That actually did the trick.
Thank you so very much, i thought I had tried this before, but my mind was so messed up with this, I screwed it completely and tested it with some update code disabled. I was already going to tell you how it doesn't work, but kept trying once more to consider your answer, and Voila.

It actually works.

Share this post


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

  • Advertisement
  • Advertisement
  • Popular Now

  • Advertisement
  • Similar Content

    • 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
       
    • By xhcao
      Is Direct3D 11 an api function like glMemoryBarrier in OpenGL? For example, if binds a texture to compute shader, compute shader writes some values to texture, then dispatchCompute, after that, read texture content to CPU side. I know, In OpenGL, we could call glMemoryBarrier before reading to assure that texture all content has been updated by compute shader.
      How to handle incoherent memory access in Direct3D 11? Thank you.
  • Advertisement