• Advertisement
Sign in to follow this  

DX11 (Order Independent) Transparency

This topic is 1852 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

Hello guys,

I started implementing transparency in my engine, and I got stuck on the usualy problem, I think, even thought I can't find very much helpful stuff on this.

I use a classic blendmode for transparent objects, and order them by the distance from the cam before drawing. This works fine if they are just a bunch of objects on behind another, but as soon as I rotate them or a model contains multiple polygons which have to been blended over oneanother, it becomes problematic.

For the rotationpart, I suppose I need a way to check the point of the object closest to the cam, and order them by this point rather than by just their position.

But I can't see someone preordering the triangles inside a mesh to get which transparent one needs to get drawn first.

As a solution, I have read about Order Independent Transparency, which is only for dx11, but that would be fine with me. I can't seem to pick up a good tutorial on this, so I may aswell ask you for one.

Or do you have another way of fixing those two problems?

 

Thanks so much!

-gnomgrol

Share this post


Link to post
Share on other sites
Advertisement

I had some help earlier on a similar issue, if you search for the topic "Rendering Intersecting Premultiplied Transparent Planes" and look at the Billboard example in the Microsoft XNA samples site they show how two passes can be used to prevent intersecting planes from creating weird depth effects.

Share this post


Link to post
Share on other sites

There's the 'OIT11' sample in the DirectX SDK. It's an expensive technique, though. You can often get away by using simple additive blending.

Share this post


Link to post
Share on other sites

Additive blending is the only blending that can be drawn without drawing back to front, yes, for everything else I always implemented a simple binary tree that receives all the triangles being sent to the renderer, inserting them into the tree based on their distance, and then when its time to draw them to the screen I just walk the tree drawing from back to front, deconstructing the tree as the triangles are rendered. This entails building a new tree each frame, but it could be set up to only build a new tree once things have moved around enough to require a new tree being built. Also, for large triangles that could be intersecting eachother this won't work properly, but it works well enough for most cases. If you have a lot of triangles in your scene, its better to precompute something like a BSP tree for the static geometry, as it wouldn't be very efficient to rebuild a distance binary tree every frame (or even every couple of frames) for tens-of-thousands of triangles.

Share this post


Link to post
Share on other sites

Hi!

 

What you ended up using is called „alpha test“ and is not truly a solution to the blending of many transparent objects, though it is a common hack to avoid the problem.

Whenever you really need to approach true order-independent transparency, there are a number of methods:

  • Depth peeling [Everitt 01, Bavoil & Myers 08] (Requires to know the number of layers, but was used quite often in research papers. Also works on Dx9.)
  • Concurrent linked list construction [Yang et al. 10, Yakiimo 10]-> do not look into the DxSDK sample (OIT11). Last time I looked, it was poorly implemented (didn’t use shared memory at all).  Check the version of Yakiimo. He implemented it faster, even with multi-sampling. This technique grew more useful by the recent advances of the graphics hardware. Requires Dx11.
  • Stochastic transparency [Enderton et al. 10] (also rather a research thingy. It can handle an arbitrary number of layers, but consumes much performance to get it frame-to-frame coherent). Needs Dx 10.1 if I recall correctly, as it works on the coverage mask.
  • The list goes on... stencil routing etc... (Look into the related work of the papers I linked, if you want to learn more.)

AFAIK, true order-independent transparency was yet too expensive for games. (If someone knows a game that used OIT, please let me know!) It is easier to work around it or just sort the transparent objects by depth. (Of course, this doesn’t work in all cases, but you can tell your artist to circumvent the ugly cases.)
I’m creating technical demos and prototypes, so I can afford to use the concurrent linked list construction.

 

Best regards!
 

PS: Links don't seem to work at the moment, so I post the URLs.

Everitt 01: http://gamedevs.org/uploads/interactive-order-independent-transparency.pdf

Baviol & Myers 08: http://developer.download.nvidia.com/SDK/10/opengl/src/dual_depth_peeling/doc/DualDepthPeeling.pdf

Yang et al. 10: http://onlinelibrary.wiley.com/doi/10.1111/j.1467-8659.2010.01725.x/abstract

Yakiimo 10: http://www.yakiimo3d.com/2010/07/19/dx11-order-independent-transparency/

Enderton et al. 10: http://www.enderton.org/eric/pub/StochasticTransparency_I3D2010.pdf

Share this post


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

  • Advertisement
  • Advertisement
  • Popular Now

  • Advertisement
  • 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