Tagged with directx - NVIDIA Developer Forums

LiveKernelEvent

Michi10 — Wed, 16 May 2012 07:11:50 -0400

Hello support NVIDIA

Since 3 weeks, I have a big problem in the video or games comes with me on the screen such a picture.

[url=http://www.abload.de/image.php?img=foto0089s2dri.jpg][img]http://www.abload.de/thumb/foto0089s2dri.jpg[/img][/url]

After a few minutes, this image comes back.

[url=http://www.abload.de/image.php?img=foto00903gckl.jpg][img]http://www.abload.de/thumb/foto00903gckl.jpg[/img][/url]

Then I looked at the Fehlerrebort me and understand only station so I can make only so enter it you:

Produkt
Windows

Problem
Grafikkartenfehler

Datum
16.05.2012 11:06

Status
Nicht berichtet

Beschreibung
Aufgrund eines Videohardwareproblems ist Windows nicht mehr voll funktionsfähig.

Problemsignatur
Problemereignisame: LiveKernelEvent
Betriebsystemversion: 6.0.6002.2.2.0.768.3
Gebietsschema-ID: 1031

Dateien zur Beschreibung des Problems
WD-20120516-1106.dmp
sysdata.xml
Version.txt

Weitere Informationen über das Problem
BCCode: 117
BCP1: 87FA1510
BCP2: 92358ACE
BCP3: 00000000
BCP4: 00000000
OS Version: 6_0_6002
Service Pack: 2_0
Product: 768_1

I was with my computer at stores where I bought the parts and the graphics card and now even the technician is verwiert and white but not what he do sol my graphics card is only 1 month old. The burden has not revealed the whole PC and the driver version is 296.10 this software I de 5 times and installed but always the same result.

What can that this be it suddenly is pink and one freezes the whole computer?

From the UIG wonder what can be even?

The technician said to me the computer is fine, but what is the error?

I defend glad if you please me an only for the problem very soon can give a response?

With friendly regards
Michael Hirschegger

texture sampling in a CG vertex shader

nicoteen — Thu, 10 May 2012 10:39:13 -0400

Hi guys,

I am trying to access a texture in the vertex shader. So I need the tex2D-function. But the compiler says, that this function is not supported in my profile, which is vs_3_0(Dirext3d). In the internet I get two contradicting results on this topic:
CG documentation says that tex2D is supported in fragment programs. But for vertex programs it names only OpenGL-profiles. Nothing about Direct3D.
In the General FAQ it says "tex2D will be supported in the vs_3_0 vertex profile and beyond in HLSL and in Cg, as it is a requirement of Vertex Shaders 3.0 and beyond. Similar functionality will be available in Cg's OpenGL profiles."
I am using vs_3_0 but the compiler says, that tex2D is not supported. Am I doing something wrong?

Using tex2D in a CG vertex shader compiled for OpenGL is working with the vp40 profile. So my hardware is definitely capable of doing a vertex texture fetch. But there are some reasons for using DirectX in my project.

I also found the similar function tex2Dlod which is supported in vs_3_0. But it always returns zero and I am sure that the colors in the tetxure are not zero, because I use them in the pixel shader afterwards.

So, what is the right way to acces a texture in a vertex shader on d3d?

Greetings,
nicoteen

nvstereo.h

colneilens — Wed, 09 May 2012 17:41:39 -0400

Is there an updated version of nvstereo.h for dx11? The one in SDK 11 only deals with DX9 and DX10?
Thanks,
Colin Neilens
Digital Extremes.

pfnIaSetIndexBuffer in NVIDIA display driver

gzMarshall — Thu, 03 May 2012 05:46:50 -0400

Hi,

I'm trying to render overlay in game via hooking the NVIDIA display driver. I hooked the pfnPresent and render the overlay in the function. When I called the NVIDIA's native function pfnIaSetIndexBuffer, it returns device removed error. The MSDN about the WDDM is too simple to debug. Does anyone in this forum know the issue?

D3D10DDIARG_CREATERESOURCE mtxResource;
mtxResource.pInitialDataUP = NULL;
mtxResource.ResourceDimension = D3D10DDIRESOURCE_BUFFER;
mtxResource.Usage = D3D10_DDI_USAGE_DEFAULT;
mtxResource.MapFlags = 0;
mtxResource.MiscFlags = 0;
mtxResource.Format = DXGI_FORMAT_UNKNOWN;
mtxResource.SampleDesc.Count = 1;
mtxResource.SampleDesc.Quality = 0;
mtxResource.MipLevels = 1;
mtxResource.ArraySize = 1;
mtxResource.pPrimaryDesc = NULL;

D3D10DDI_MIPINFO indiciesSize;
indiciesSize.PhysicalWidth = indiciesSize.TexelWidth = sizeof(ovlIndicies);
indiciesSize.PhysicalDepth = indiciesSize.PhysicalHeight = indiciesSize.TexelDepth = indiciesSize.TexelHeight = 1;
mtxResource.pMipInfoList = &indiciesSize;

D3D10_DDIARG_SUBRESOURCE_UP indiciesInitialData;
indiciesInitialData.pSysMem = ovlIndicies;
indiciesInitialData.SysMemPitch = 0;
indiciesInitialData.SysMemSlicePitch = 0;
mtxResource.pInitialDataUP = &indiciesInitialData;

mtxResource.BindFlags = D3D10_DDI_BIND_INDEX_BUFFER;
SIZE_T szIndicies = m_pDevice->m_deviceFuncs.pfnCalcPrivateResourceSize(m_pDevice->GetDrv(), &mtxResource);
m_hIndicies.pDrvPrivate = (void *)(new char[szIndicies]);

D3D10DDI_HRTRESOURCE hrtNull;
hrtNull.handle = new char[szIndicies];

m_pDevice->m_deviceFuncs.pfnCreateResource(m_pDevice->GetDrv(), &mtxResource, m_hIndicies, hrtNull);

m_pDevice->m_deviceFuncs.pfnIaSetIndexBuffer(m_pDevice->GetDrv(), m_hIndicies, DXGI_FORMAT_R32_UINT, 0);

Problem with DirectX and PhysX 3

rwirz — Wed, 18 Apr 2012 14:04:23 -0400

Hi,

I can run in my computer PhysX 2.X (Samples in OpenGL) and PhysXViewer (DirectX).
However, when I try to run PhysX 3.X Samples, the samples crash.
I invetigate and my computer dont have Hardware Vertex, so I change to Sofware Vertex (In the code). But a new error appear in CreateVertexShader, so I dont run the samples.

How can i run the samples in OpenGL?

Thank you

Problem with second monitor artifacts (DirectX 9, multihead mode, Windows XP)

fotis — Wed, 25 Apr 2012 04:22:52 -0400

Hello,

General Problem Description

We have artifacts on second monitor when working with two monitors in multihead mode, locked with VSYNC. The artifact looks like an horizontal bar that moves up from the bottom of the second monitor, sometimes goes all the way to the top of monitor, sometimes dissapears in the middle.
Looks like some kind of sychronization problem. Doesn't exist on the back buffer of the second monitor (checked with various kinds of captures, inside and outside of our application). I attach a video of the problem.

Test environment

We use DirectX 9.0c. Tested with GeForce GT 220, GeForce 8400GS, GeForce GT 520 and the latest driver (296.10).
OS: Three different Windows XP Professional machines both with SP2 and SP3.
We create one DIRECT3DDEVICE9 with D3DCREATE_ADAPTERGROUP_DEVICE to guide simultanously two monitors full screen with the same card (Dual View mode). We have used exact microsoft help guidelines to implemement it.

Exact problem behavior

All work OK when presentation interval is D3DPRESENT_INTERVAL_IMMEDIATE.
The problem exists when we choose D3DPRESENT_INTERVAL_DEFAULT to lock FPS with vertical refresh (We have certain needs that force us to do so).
The problem exists even in case we use the same resolution on the two monitors (always in Dual View) but becomes worse when the resolutions are different.
Activation/deactivation of hw vertex processing doesn't affect it.
Set/unset of D3DCREATE_MULTITHREADED also doesn't affect it.
Increasing refresh rate of monitors to 75Hz (from 60Hz) seems to make the problem more frequent.
The same code/executable works OK in a test machine with Windows 7 Ultimate (machine had DX 11 runtime installed by default).

Other information

We use the same OS/cards for years in production environment without problems using horizontal span. We used dual view mode in the past with Intel onboard graphics cards without problems.
This time though we need to use different resolutions in the two monitors, so there is no other way than dual view.

Looking forward for some help because we have an ongoing project that depends on this and is very critical.

Thank you very much in advance,

Fotis Kapetanias

DirectX integration: What happened to getGlobalPose().getColumnMajor44()?

Daniel Sefton — Wed, 25 Apr 2012 21:42:51 -0400

Hi,

I've been trying to implement PhysX 3.2 (Beta 3) with DirectX, and one of the problems I came across is how to sync DirectX meshes with PhysX shapes. It's often recommended to use the getColumnMajor44() function of getGlobalPose() like so:

D3DXMATRIX d3dMat;
actor->getGlobalPose().getColumnMajor44(d3dMat);
g_pd3dDevice->SetTransform(D3DTS_WORLD, &matWorld);

But apparently this function is no longer existant. Instead I saw that using .q and .p to obtain vector and quaternion positions was the new way, so I tried:

PxTransform pose = actor->getGlobalPose();
D3DXMatrixRotationQuaternion(&mMatRotate, &MappingsD3D9::makeD3DXQuat(pose.q));
D3DXMatrixScaling(&mMatScale, scale.x, scale.y, scale.z);
D3DXMatrixTranslation(&mMatTranslate, pose.p.x, pose.p.y, pose.p.y);
mMatWorld = mMatRotate * mMatScale * mMatTranslate;
mRenderSystem->getD3D9Device()->SetTransform(D3DTS_WORLD, &mMatWorld);

It sort of works... But (the rotations) are totally messed up. When a shape starts rotating the mesh goes in all sorts of crazy offset positions. It may be to to with the fact that PhysX uses a right-hand coordinate system, whereas DirectX's is left (the getColumnMajor44() function solved that before).

If someone could point me in the right direction, I would appreciate it.

Cheers.

PhysX 3.2: LPD3DXMESH to PhysX Scene

CyberKevin — Thu, 08 Mar 2012 23:23:34 -0500

I have looked on the internet and through the forum and I cannot find any references on how convert .x mesh loaded with DirectX as LPD3DXMESH to mesh usable in a PhysX scene through PxTriangleMeshDesc or PxDeformableMeshDesc. The only references that I have found online to do this were for PhysX 2. Even after I tried to change those codes, it still did not work. Is there a way to do this and if so, what is it?

DirectX 11 Tesselation shaders in Cg

Knuckledust — Mon, 23 Apr 2012 05:08:52 -0400

I am working on a DirectX11 engine with Cg, but there are no details online on how the hull and domain shader syntax works in and no examples, at least none I could find.
Can anyone give me an example or a link to the shader syntax?

I need a shader

Kamlah — Tue, 03 Apr 2012 07:52:38 -0400

Hi All,
I would like to get some help. I work in a TV as a graphic artist. Here we use a virtual studio system, where I need to create or somehow solve the problem showing frozen glass like material. The point is that the objects behind need to be blured. I found a good example here: http://www.polycount.com/forum/showthread.php?t=87743
The first picture shows some teapots behind a glasslike surface. Thats what I need.
My biggest problem is that iam a total lama for the Cg topic. I started to learn and go into deep into the topic, but now i see that it is much more complex as i thought.
So i would be really thankful ...
The studio render engine use Nvidia cards but I dont know the exact parameters now. If necessary i look for it.
thx
KG

Scenix DirectX Rendering

cdechenne — Mon, 16 Apr 2012 13:42:27 -0400

I'm not sure if anyone is interested, but I've created a DirectX renderer for Scenix. It's actually doing the rendering using OpenGL, but then uses a OpenGL DirectX interop to transfer the scene to a DirectX context. So you shouldn't have to worry about your scene looking any different than the OpenGL version (but it is in fact a DirectX scene).

More information can be found here:
http://sites.google.com/site/snippetsanddriblits/scenix-directx-rendering

CreateDeviceAndSwapChain fails with unkown error code

aerlfredith — Mon, 16 Apr 2012 07:54:23 -0400

I am trying to create a D3D10Device with

hr = D3D10CreateDeviceAndSwapChain( 0, D3D10_DRIVER_TYPE_HARDWARE, 0, createDeviceFlags, D3D10_SDK_VERSION, &sd, &m_pSwapChain, &m_pD3DDevice);

which works on my older laptop but NOT om my new desktop with a GTX 680 it does not return a valid HRESULT the return is -2005270527. I think this could be some issue with the GPU compatibility but am not sure and am unable to find an answer on msdn. Help would be sorely appreciated.

for those wondering this is de descriptor i give along

DXGI_SWAP_CHAIN_DESC sd;
ZeroMemory( &sd, sizeof( sd ) );
/* buffer desc first */
sd.BufferDesc.Width = m_Options->m_WindowWidth;
sd.BufferDesc.Height = m_Options->m_WindowHeight;
sd.BufferDesc.RefreshRate.Numerator = 60;
sd.BufferDesc.RefreshRate.Denominator = 1;
sd.BufferDesc.Format = DXGI_FORMAT_R8G8B8A8_UNORM;
sd.BufferDesc.ScanlineOrdering = DXGI_MODE_SCANLINE_ORDER_UNSPECIFIED;
sd.BufferDesc.Scaling = DXGI_MODE_SCALING_UNSPECIFIED;

/* sample once */
sd.SampleDesc.Count = 1;
sd.SampleDesc.Quality = 0;

sd.BufferUsage = DXGI_USAGE_RENDER_TARGET_OUTPUT;
sd.BufferCount = 1;
sd.OutputWindow = m_Options->m_GameWindow;
sd.Windowed = m_Options->m_bWindowed;
sd.SwapEffect = DXGI_SWAP_EFFECT_DISCARD;
sd.Flags = 0;

Best way of going about this? (hlsl, sm4+)

hazarada — Fri, 13 Apr 2012 18:39:09 -0400

I'm looking to divert more calculations from cpu to gpu side. One of the more major resource hogs in my project was skin matrix creation which was originally done by the cpu. So i figured i'll just give the gpu the source data and have it create the matrices itself and i did do so in the vertex shader but the next morning i realized (was a long night) that this way the matrices are being built and combined thousands of times over while they only need to be calculated once. This epic inefficiency aside the result was still faster than doing it on the cpu side. Now to do it efficiently i would need some way to feed the source data (the source material unit length varies from 4 to 16 bytes depending on the type) to the shaders and have them chew through it and store it in vram. I think it would be possible to switch out the vertex shader and feed it bogus vertices to get them to work on the source data (currently located in a stream in a cbuffer) but it seems messy. Is there any "good" way of doing this? thanks.

Breakpoints not being hit in NSight 2.1(and 2.0)

dblack — Sun, 20 Nov 2011 14:35:35 -0500

Hi,

I am currently trying to debug some d3d 11shaders, however when I set breakpoints in my project they are not triggered. (The variance shadow mapping sample works fine).

However there are a couple of things which differ with my main project and I am not sure if they should work?

* Shaders are pre-built, however they have the correct flags set(eg debug, skip opt, prefer flow control). One thing which might cause an issue is that they content isnt built on the host, so the path name supplied when building is not available. Still nsight says it has symbols and when clicking the shader it opens a file in the temp dir.

* My solution is loaded over a share from the client, then it is copied back using NSights built in synchronization. (the reason for this funny setup is that I only have one machine with d3d11 capable hardware).

FWIW exactly the same thing happens with 2.0. Frame profiling and timing work fine.

Thanks,

David

Why copy time so long to NVidia video card?

ygwangking — Fri, 13 Apr 2012 16:14:18 -0400

Hello,

I have two NVidia video card, 9800 GTX and GTS 450.
My application drive 3 mintors.

Main application window display at 9800 GTX DVI port 1.

Anther two HD monitor connect GTX DVI port 2 and GTS 450 DVI port.
and take these two monitor full screen display image unsing directX 10 Windowed mode.

Image capturerate is 60 and I have to display every frame So I set mintor refreshrate as 60.

If I only drive only one HD monitor(any one), works fine directX refresh rate 60 fps.

But If I drive two HD monitors refresh rate dropdown to 30 fps and copy time from host to GPU take very long. I checked it's directX map() function spend much time in some frame.

I don't understand why I seperate copy two image to two video card take so long and why refresh rate dropdown to 30.

Could you help me?

Compact 2D vertex format problem, FVF, D3D9

ArcticHammer — Fri, 13 Apr 2012 01:35:26 -0400

We are successfully using compact vertex structure for Ati(AMD) Radeon GPUs, they work in all models about 5 years old or newer. We want to minimize the data amount to be written to the GPU for maximizing the performance, we are rendering 2D graphics with D3D9. The same vertex structure type doesn't work with any nVidia or Intel GPU in our app.

We are currenlty using SlimDX .NET library, but the same problem occurs in Microsoft Managed DirectX too. And most likely also native DirectX (not tested).

We can successfully use the following struct for Ati(AMD):

internal

struct CompactVertex
{

public float X;

public float Y;

public int Color;

public static VertexFormat Format = VertexFormat.None;

public const int StrideSize = 12;

public static VertexElement[] VertexElements =
{

new VertexElement(0, 0,

DeclarationType.Float2,

DeclarationMethod.Default,

DeclarationUsage.PositionTransformed,0),

new VertexElement(0, 8,

DeclarationType.Color,

DeclarationMethod.Default,

DeclarationUsage.Color, 0),

VertexElement.VertexDeclarationEnd
};
}

And for nVidia and Intel, we have to use larger one:

internal

struct LargeVertex
{

public float X;

public float Y;

public float Z;

public float RHW;

public int Color;

public static VertexFormat Format = VertexFormat.PositionRhw | VertexFormat.Diffuse;

public const int StrideSize = 20;

public static VertexElement[] VertexElements =
{

new VertexElement(0, 0,

DeclarationType.Float4,

DeclarationMethod.Default,

DeclarationUsage.PositionTransformed,0),

new VertexElement(0, 16,

DeclarationType.Color,

DeclarationMethod.Default,

DeclarationUsage.Color, 0),

VertexElement.VertexDeclarationEnd
};
}

We have tried to detect the compatible type by investigating
bool compactSupported = (caps.FVFCaps & VertexFormatCaps.DoNotStripElements) == 0;

Ati/AMD has flag value set to 0, Intel and nVidia 1.

and also tried to force the compact format into use in nVidia and Intel, but they fail to render it.

It's very difficult for me to believe that nVidia wouldn't support this kind of compact vertex type.

So, can you say what are we doing wrong, and how to get a compact vertex type (x,y,color) working with nVidia?

Thanks for any help...

Parallel Nsight 2.2 Release Candidate 1 is available!

Rafael Campana — Wed, 11 Apr 2012 18:33:03 -0400

The NVIDIA Parallel Nsight development team is proud to announce Release Candidate 1 of Parallel Nsight™ 2.2. This new release brings support for single GPU Debugging for CUDA developers on systems equipped with any GPU that supports hardware GPU debugging. For graphics developers, Nsight now supports DirectX 9 in the Frame Debugger, Frame Profiler, Analysis and Nsight HUD.

For more information about the release, and how to download it, please visit:
http://developer.nvidia.com/content/nvidia-parallel-nsight-22-rc1-now-available

DX11.1 drivers for new GTX 680 cards on Windows 8 CP..

oscarbg — Wed, 11 Apr 2012 06:47:03 -0400

Are coming soon?

Are some Windows 8 WDDM 1.2 drivers for new GTX 680 cards?

oscarbg — Wed, 11 Apr 2012 06:46:18 -0400

As a developer wanting to exploit some WDDM 1.2/DXGI 1.1 features want a driver for new GTX 680 card..

Idea for demo: ptex

jogshy — Tue, 10 Apr 2012 13:55:51 -0400

I would like to see a demo from NVIDIA featuring realtime PTEX.
thx

[DirectX11] : Resource creation, initial data and synchronism

DZ4 — Tue, 10 Apr 2012 06:03:55 -0400

Hi,
Since DirectX11, we can now create resources from any thread, including on older hardware where the feature is emulated.
But when creating a resource on a device, is it guaranteed to be totally synchronous regardless the calling thread ?

I've experienced really strange behavior in really particular conditions where successive calls to an immutable buffer creation using initial data seems to be asynchronous: If the data behind the pointer given to the D3D11_SUBRESOURCE_DATA.pSysMem is changed just after the CreateBuffer() call, the resource content after creation is corrupted.
(using a default resource usage with updatesubresource instead of initial data was a working workaround).

Better explanation on concurrent resource creations and synchronism NVIDIA device implementation would be greatly appreciated.

Thanks.

SDK10 sample not working correctly

caianbene — Fri, 27 Jan 2012 07:17:19 -0500

yesterday I've downloaded Parallax Mapping code sample and a few others from NVidia's DirectX10 SDK page, I remember downloading the full SDK in the past and it worked fine, in fact most of samples are working except for Parallax Mapping. The silhouette effect is all messed up:

http://i.imgur.com/Xr7xB.jpg

Recompiling the sample and DXUT not solved the problem, does Microsoft introduced changes in the mesh processing or shader compilation? Or could this be a driver issue? (I'm currently using 285.67 because of Parallel NSight)

Thanks in advance

Crash in CgFx DX10 when assigning texture

JoakimD — Sat, 19 Nov 2011 04:08:06 -0500

If you try to assign a texture to a sampler that isn't used in the shader CgFx will crash with 'trying to access memory at 0x00000040'.

Start the DX10 CgFx sample called 'cgfx_texture'.
In the shader, comment out the line "OUT.color = tex2D( decal, input.tex0 );" and run it.
It will crash.

This causes some problem for me since knowing exactly if a texture is used or not is difficult because I have built a realtime shader editor using CgFx but want to move to DX10 instead of OpenGL. I'm using the Feb 2011 Cg version

Any time scale when this will be fixed?

PhysX import/export actors

SoulSharer — Fri, 30 Mar 2012 04:37:48 -0400

Hello there I wonder if I do export as intended and how do I need to import?

What I do is switching PhysX plugin in 3ds max to 3.x (I have 3.2 [Beta 3] SDK), then I chose a model with physx modifier and exporting it as a PhysX scene with XML format. Do I need to export all models at once, or I can export one model at a time like I'm doing now?

When I finish exporting I get two files .RepX (which is needed to load up an actor via deserialize I guess) and .PxProj (don't know what's it for).
Am I doing export right?
And how do I import that stuff?

~~I also confused how I should attach a mesh model to its PhysX actor or whatever. I think I just need to multiply mesh by a matrix PhysX provide, but it's only a guess.~~ (answered)

I already searched google, documentation, forum and other stuff. Yet I don't have any clear answer just yet.
Thanks for any help you can offer.

---Some info
PhysX SDK: 3.2 Beta 3
API: Direct3D10
Modeling tool: 3ds Max 2012

Trying to integrate PhysX into game.

Why does Stencil-Routed K-Buffer DX10 Sample not work correctly (strange effects) on DX11 hardware?

SoulWiz — Fri, 30 Mar 2012 16:09:52 -0400

sample code: http://news.developer.nvidia.com/2007/12/sdk-10-update-n.html
I have the problem that I get strange artifacts when using Stencil-Routed K-Buffer on my Geforce GTX 560. Depth peeling however works fine.
I tried it with both Geforce 9600 GT (works fine) and Geforce GTX 560 (strange artifacts). Is this a DirectX11/DirectX10_1-Bug or some driver problem?

Implementing HW Alpha Test in HLSL

inbilla — Tue, 27 Mar 2012 22:10:20 -0400

I'm currently trying to implement the HW Alpha Test from SM2 in HLSL.

I know there are fairly straight forward methods of achieving this, i.e.
clip(alphaVal <= alphaRef ? -1 : 1);

However, in trying these techniques, I've noticed that they deviate from results that HW Alpha test returned (as evidenced by extra, or less pixels being affected).

The technique I'm currently using involves doing the following:

float alphaVal = readFromTexture(uv).a;
float alphaRef = 0.5;

float enableLessTest = 0.0; // 1.0 (off/on)
float enableGreaterTest = 1.0; 
float enableEqualTest = 0.0;

// Greater test
clip((alphaRef - alphaVal) * enableGreaterTest);
// Less test
clip((alphaVal - alphaRef) * enableLessTest);

// Equal test
float equality = (alphaVal == alphaRef) ? -1 : 1
clip(equiality * enableEqualTest);

Now the above exhibits the issues described where pixels on the edge of the alpha tested area are discarded by the HW, but included by the SW version.

I'm pretty sure it is a problem related to the precision that this calculation is being performed in, I assume the alpha test in HW uses Integer comparisions instead of float, however, I am trying to keep this shader ps_2 compliant.

I tried performing the following operation on alphaVal:
alphaVal = trunc(alphaVal * 255)/255;
with no luck.

I also tried flooring/truncing the various calculations inside the clip functions, however I can't seem to get exactly the right pixels returned, there are always either more or less pixels returned.

Does anyone have any ideas? I'm sure someone must have done this before?

Thanks

Josh

Windowed Stereo DirectX Rendering

vskryabin — Tue, 07 Feb 2012 10:22:31 -0500

Hello! I'm an app developer and want to support 3D Vision/3D Vision Pro. I use DirectX9 for rendering of the 3D scene in my projects. Recently, I learned that windowed DirectX applications can work in 3D stereo now. But I couldn't find any example with it. Can you help me with my problem? I need a simple example of stereo 3D rendering by means of DirectX9 in window (not fullscreen), or the detailed guide. Can I use nvapi for this purpose? In advance thanks!

NVS 295, deferred context and D3D_DRIVER_TYPE_HARDWARE

haust — Tue, 27 Mar 2012 10:38:32 -0400

Hello,

latelly i've been playing with deferred context.

Nothing seemed to work as it should with the hadrware driver. Then I decided to give a try to the reference driver and all worked as intended....

My test was to render an instanced geometry in a deferred context with different textures. Nothing fancy (no geomtry shader, no compute shader, no shadow, etc.) just a simple textured quad rendered many times through instancing.

As indicated I'm using a Quadro NVS 295 with driver 295.73 and don't have access to another video card.
Does anyone know how to fix this issue or can tell what I'm doing wrong :)

haust.

3D Vision only renders one eye on laptop

cdechenne — Thu, 22 Mar 2012 14:30:33 -0400

I'm trying to use DirectX 9 in conjunction with 3D vision in order to do stereo. My application is written to do windowed mode stereo, but the problem is the same in full screen as well. The problem that I'm having is that I only see one eye when the app is rendered in stereo. Here's what I debugged so far:

My app has three render targets - one for left, one for right, and one to combine the two. The left render target is just rendered with a clear color of purple. The right render target is rendered with a clear color of yellow.
When I run my app without stereo, I see the purple and yellow side by side in the window.
When I run my app with stereo, I see just yellow (both in windowed and full screen).
The emitter comes on with a bright green and the screen flashes black a few times, so I'm pretty sure it is recognizing my StereoImageHeader that is required for the stereo to work.
If I change the dwFlags of the StereoImageHeader to use a SWAP_EYES, I see purple instead. That says to me that it is interpreting the stereo surface just fine.
Somehow I've thought that it must be some sort of problem with the size of the RECTs used for the StretchRect (to copy the left and right onto the combined surface), but I've tried just about every combination I can think of with no luck.
I'm using driver version 295.73
I'm trying to get this to work on a laptop with a GeForce GTX 260M. My test app seems to work fine on a desktop with a GeForce GTX 470.

Anybody got any ideas as to what's going on? I can post my code if needed...

Getting into CgFX and fx composer

terjeloe — Sun, 18 Mar 2012 09:57:07 -0400

Hi,

I'm getting into shaders and using CG with CgFX and fx composer 2.5. I've written a basic shader that compiles with no warnings, but when I apply it to a teapot object, rendering with directx9/10 I only see a wireframe mesh? With the default effect its fine though.

If I render my effect with OpenGL the object becomes black and it show up black in the material preview. The default effect is also black in the material preview, but if I apply it to the teapot it renders it fine.

My program have code fore transforming the object aswell.. how do I get fx composer to send the viewPerspective matrix from the camera and the model matrix so I can test that my transformations are right?


void mainVS(float4 position : POSITION,
			float3 normal 	: NORMAL,
			float2 texCoord	: TEXCOORD0,

			out float4 oPosition : POSITION,
			out float2 oTexCoord : TEXCOORD0,
			uniform float4x4 viewProjection,
			uniform float4x4 world)
{
	oPosition = mul(world, position); 
	oPosition = mul(viewProjection, position);
}

void mainFS(float2 texCoord	: TEXCOORD0,
			out float4 color 	: COLOR,
			uniform sampler2D texSampler)
{
	color = tex2D(texSampler, texCoord);
}


float4x4 world : World;
float4x4 viewProjection : ViewProjection;

sampler2D texSampler = sampler_state
{
	generateMipMap = true;
	minFilter = LinearMipMapLinear;
	magFilter = Linear;
};

technique basic_textured
{
	pass
	{
		FragmentProgram = compile fp40 mainFS(texSampler);
		VertexProgram = compile vp40 mainVS(viewProjection, world);
	}
}

RWByteAddressBuffers

SeanT — Fri, 16 Mar 2012 11:35:41 -0400

I'm working on a technique that uses a RWByteAddressBuffer from a pixel shader. The UAV is bound to u0. Clearing the UAV works, but the pixel shader writes never seem to the update values (tried InterlockedExchange and Store).

This is using a GTX 580 card, and the same shader runs fine on a similar ATI card.

Is there a known issue, or something I might be missing?

Problem with DirectX Application Trace when rendering using multiple deferred contexts

CharlieB — Thu, 15 Mar 2012 12:45:06 -0400

Hi

I'm having a problem when attempting a DirectX application trace when I use multi-threaded rendering. The application consists of a simple test level with 6 way split screen, where each view is rendered simultaneously using a deferred context, then each command list is executed on the main device context. The app runs for several seconds and then freezes. The amount of time it lasts for seems dependent on the amount of stuff being drawn - if I render less instances for example, it lasts longer, but still freezes eventually. It's still possible to trace it in this time, though it will freeze whilst tracing if you don't stop soon enough, and in this case "Load Status" will be "Failed" for the .nvevents file.

This problem only occurs if I launch through the NSight Analysis Activity with DirectX trace enabled and when using more than 1 deferred context at the same time. Any other circumstances and it is fine, I've obviously tried debugging with full D3D11 checks and so on and nothing is coming up. The app exits with a clean D3D11 shutdown with no warnings or errors.

The MultithreadedRendering11 sample from the D3D SDK seems OK, so there is something else going on.

Any suggestions as to how I might go about tracking this down?

What kind of level editor can I use along with NVIDIA Graphics SDK 10.6?

BigLouis — Thu, 15 Mar 2012 03:46:12 -0400

I'm an aspiring indie game developer looking for a suitable way to build a high quality DirectX 10 game and would like to know if there's any DirectX 10 game engine that you know of. Since I'm a newbie a complete game building suite is highly desirable. Thanks in advance for any information.

Nsight: how to set texture name?

steel3d — Thu, 05 Jan 2012 21:39:43 -0500

I'd like to know how to name my resources such that I can see them in nsight captures.

Thanks

Crash in nvgf2um.dll

esrever2357 — Mon, 12 Mar 2012 13:38:55 -0400

Hi,

My application is sometimes (apparently randombly) crashing in nvgf2um.dll.

Higher level description:
Two shader resource views are bound in an effect, and used in one of the compute shader techniques it contains.
The application occasionally crashes somewhere in nvgf2um.dll, at some point in IDXGISwapChain::Present(), as I've found out in PIX. It only happens when the aforementioned compute shader is run.

More details:
The two views use the same resource, just binding different parts of it - it's actually animated position data which are lerped in the shader. I generate 100 views for my test model.
Resource structure byte stride: sizeof(float3)
The view descriptions set the 'FirstElement' and 'NumElements' fields in terms of elements (float3's), not bytes.

Specs:
GTX 580 using 64-bit drivers, tried 285.16 as well as the latest ones.

Any ideas?
I could provide a PIX log if needed, but I'm not sure if it will be of any help.

Thanks,
Babis

Bug in cl_nv_d3d11_sharing: Cannot map DirectX buffer for interop

mchajdas — Mon, 12 Mar 2012 03:25:21 -0400

There seems to be a bug in cl_nv_d3d11_sharing which prevents more than 526 buffers being mapped for interop, even if there is plenty of memory free. In the attached application, we create, map and release 1 MiB sized buffers and on both a 560 Ti with 1 GiB of memory and a 480 with 1.5 GiB it fails at around buffer 526 (depending on the buffer size.) We have reproduced the issue also on a 3 GiB Tesla C2050. What's going on here? As we release each and every buffer, we should be able to map them until the end of days, shouldn't we?

HullShader / HullShaderConstantFunction Bug

krauthaufen — Fri, 09 Mar 2012 11:38:40 -0500

Hi all,

I have two problems using HullShader and HullShaderConstantFunction:

1) The HullShader does not support passing "non-trivial" attributes:
Here is some Code that does not work on current nVidia Drivers:
HS_Output HS(uint id: SV_OutputControlPointID, InputPatch input)
{
HS_Output output;
output.Positions = input[(id + id) / 2].Positions;
return output;
}
In fact the HullShader does not allow any index-calculation at all. And outputs can not depend on more than one input-index.
I tested several things like above with ATI-Cards and the DirectX11 Reference Device.

2) The HullShaderConstantFunction cannot use an input which is not passed (or used) by the HullShader:

HSC_Output HSC(InputPatch input)
{
HSC_Output output;
//...
//Set the Tessellation-Factors
//...
output.SomeAttribute = input[0].TexCoord + input[1].TexCoord + input[2].TexCoord + input[3].TexCoord;
return output;
}

//...
[patchconstantfunc("HSC")]
HS_Output HS(uint InputControlPointIndex : SV_OutputControlPointID, InputPatch input)
{
HS_Output output;
output.OutputPositions = input[InputControlPointIndex].InputPositions;
return output;
}

This Combination does (as above) work on the Reference-Device and on ATI-Cards.
If i add "output.TexCoord = input[id].TexCoord;" to the HullShader it works also on nVidia-Cards.

Does anyone know why that happens or is it just a Driver-Bug?

GeForce GTX 560 Ti Clock-Texture Corruption Issue?

keiichi25 — Mon, 05 Sep 2011 13:54:29 -0400

I do not know where to ask this, so I will put it in general and hope it will be moved to an appropriate forum. I was wondering if there is any known issues with having the 500 series Geforce cards, namely the GeForce GTX 560 Ti OEM series, having a slight issue with having the card set at Adaptive and running for long periods of time?

The reason I am asking, is that I happen to notice with the Alienware provided OEM GeForce GTX 560 Ti card, under Windows 7, if the card is running longer than 20+ hours, the card will invariably have two events happening:

1) The driver for the card will experience a Timeout, Driver not responding, but has recovered.
2) Any games using the 3D portion of the GPU start to suffer graphically Texture corruptions, such as textures gaining weird colors, textures being stretched/torn from models in weird directions.

This happens with all drivers, from 270.66 to 280.26. Short of putting the card into Performance, which locks the card to its performance clock times (Mind you, I did not tweak any of the card's clock speeds from the way the card was given to me) I was not able to run the card past 20 hours for the OpenGL/Cuda enabled games without some sort of noticeable corruption in the graphics. The desktop, on the other hand, was functioning without problems.

Speaking with nVidia tech support, who had me try the various drivers and also looking at some of the programs running, nothing short of putting the card in performance mode seem to have rectified the problem.

Cg equivalents to HLSL functionality

keithoconor — Wed, 07 Mar 2012 18:39:59 -0500

We have used Cg to great success as a common shading language for our previous cross-platform titles. We're in the middle of upgrading our PC renderer to D3D11, and I'm starting to come across SM5 HLSL functionality that doesn't seem to have any equivalents in Cg.

For example, in DX11 HLSL there is a Texture2D function GatherRed() - http://msdn.microsoft.com/en-us/library/windows/desktop/ff471558(v=vs.85).aspx. However I can't find any equivalent in Cg. Similarly, HLSL allows switch statements but Cg doesn't according to the documentation.

These are just a couple of examples, but it's making me concerned that in sticking with Cg we might be backing ourselves into a corner and that we will eventually come across a case where we want to do something that would work in HLSL, but just isn't exposed in Cg. Am I just missing something, or are HLSL and Cg starting to diverge?

Problems trying to compile/link Direct3D 10 SDK Smoke sample

HIDEBU — Tue, 15 Nov 2011 01:39:11 -0500

I've been trying to compile this sample (http://developer.download.nvidia.com/SDK/10/Samples/Smoke.zip) that's included with the set of samples from the Direct3D 10 SDK so that I can experiment with it, but I've been having a lot of trouble trying to make it work. I have installed the Direct3D 10 SDK, Cg 3.0, Microsoft DirectX SDK (February 2010), and am attempting to compile the sample on Visual Studio 2008. I've encountered a number of issues so far which I've documented below, but I'm stuck on how to fix a linker error with the .lib file nvutd.lib. Any help would be very much appreciated!

Problem 1: FXC custom build step fails as "'fxc.exe' is not recognized as an internal or external command.
- FXC.rules file modified to include the path where the x86 FXC executable is, in particular:
CommandLine=""$(DXSDK_DIR)Utilities\bin\x86\fxc.exe" [output] [target] [strict] [compat] [perf] [noopt] [nopre] [child] [Inputs]"

Problem 2: Voxelizer.fx(281,19): error X4580: emitting a system-interpreted value which is not written in every execution path of the shader
- Fixed by forcing default values onto the GsGenVelOutput object before Appending it to the TriangeStream object.

Problem 3:smoke\direct3d\include\dxut\core\dxut.h(93) : fatal error C1083: Cannot open include file: 'd3dx9.h': No such file or directory
- Fixed by adding "$(DXSDK_DIR)\Include;" to the Additional Include Directories of the project.

Problem 4: 1>LINK : fatal error LNK1104: cannot open file 'dxerr9.lib'
- dxerr9.lib doesn't exist anymore; it's been deprecated and replaced with dxerr.lib (see: http://www.microsoft.com/download/en/details.aspx?displaylang=en&id=23549). Change the reference to it in the Linker accordingly.

Problem 5: DirectX library files still not found
- Fixed by adding "$(DXSDK_DIR)\Lib\x86;" to the additional library directories of the Linker.

Problem 6: 1>nvutd.lib(nvutmedia.obj) : error LNK2001: unresolved external symbol "unsigned int (__stdcall* ATL::g_pfnGetThreadACP)(void)" (?g_pfnGetThreadACP@ATL@@3P6GIXZA)
- I don't know what the correct resolution to this problem is.
Adding the line "#include " to the precompiled header file Smoke.h results in a big stream of errors from the project's fcollada\futils\fuerror.h. If I add the file atls.lib and the path "C:\Program Files (x86)\Microsoft Visual Studio 8\VC\atlmfc\lib" (which is where atls.lib is located) to the Linker, I get link errors with various DXUT obj files, all claiming that "LNK4099: PDB 'vc80.pdb' was not found", followed by the linker crashing out.

nsight - get certain statistics

Jofo — Tue, 28 Feb 2012 07:02:16 -0500

Hello,

I'm working on a shadow mapping demo application in DX11, and my project supervisor asked me to get certain performance data, but I really don't know if nsight is capable of acquiring them

- avg. time spent in pixel/vertex/geometry shader (so far I've seen only general shader utilization)
- avg. texture access count per frame, and time GPU spents with texturing

Thank You very much for help

Technical problem with latest Nvidia graphic drivers and DirectX9

ccw — Thu, 01 Mar 2012 16:03:12 -0500

It would appear that the latest Nvidia graphic drivers are hooking the D3D9.DLL function 'Direct3DCreate9' in such a way that any previous hooks are overwritten.

The previous hook in my test case was applied using Microsoft Corporation Detours (Version 1.5 Build 46) during a DllMain() DLL_PROCESS_ATTACH of a dll injected into another process. If the test hook is applied later it does not get overwritten.

direct3d9 program, render dds textures crash!

fly821008 — Thu, 01 Mar 2012 09:29:41 -0500

test procedures are as follows: use DXUT, test direct3d dds textures support under windows xp .
The test environment:
windows xp, video memory 512MB, 8600GT, 2010 june the directx sdk, vc 2008. latest video card driver .

Is a d3d 3D GIS program, dynamic scheduling scene model data, then compile and render. The program will crash.

My question is to ask what I created less than 300MB of dxt1 and dxt3 format of the texture, my program will load and compile of 270 dxt1 texutres of 256*256 and over 3200 dxt3 texutres of 256*256, more than 20 dxt3 texutres of 512*512. when program Loaded over 2800 dxt3, Direct3d will crash.

Statistics: vb: MAMAGED 35MB, default 3MB; the texture all MANAGED: the number of dds textures is 3000, render patch is 3500, more than 160,000 of the triangles, the vertex is less than 1.7 million.

windows 7 is no problem.
windows xp will always crash in
pd3dDevice-> the the DrawPrimitive (D3DPT_TRIANGLESTRIP, *, *).

Prompt the following error message:
Direct3D9: (ERROR): Unexpected error in the Clone 8007000e
Direct3D9: (ERROR): The resource manager failed to promote or update a dirty texture. The texture will be disabled.

Direct3D9: (ERROR): river returned error: DDERR_CURRENTLYNOTAVAIL
Direct3D9: (ERROR): driver failed command batch. Attempting to reset device state. The device may now be in an unstable state and the application may experience an access violation.

MANAGED texture are the dds, whole mipmap. All the D3D api return S_OK!
If all dds textures only create a level 0 surface, is operating normally. And if all dds are converted to argb format and enable automipmap, direct3d can well manage the graphics resources.

Also to tried D3DCREATE_DISABLE_DRIVER_MANAGEMENT, also crash.

What are the limits of dds textures when realtime render many ddx textures of whole mipmaps, the total number? the total size? how to solve these problems.
Really very, very Thanks.

NVIDIA Parallel Nsight 2.1 Release Candidate 2 now available!

Sebastien Domine — Mon, 05 Dec 2011 19:16:15 -0500

NVIDIA Parallel Nsight 2.1 Release Candidate 2 now available!

Dear Parallel Nsight User,

Building on the NVIDIA Parallel Nsight™ 2.1 Release Candidate 1 release with multiple bug fixes and stability improvements, we are proud to announce the release of NVIDIA Parallel Nsight™ 2.1 Release Candidate 2. This release
brings support for the new CUDA Toolkit 4.1 Release Candidate 2, which can be downloaded under the CUDA Registered Developer Program (www.developer.nvidia.com/join). Parallel Nsight 2.1 adds a number of new features to enhance debugging and profiling capabilities.

This release requires NVIDIA Display Driver Release 285.86, available on the same download site.

Traced workloads can now navigate the dependencies and call stack to allow the developer to follow through GPU workloads, corresponding API calls and host code that was the cause of the activity.
CUDA warp watch visualizes variables and expressions across an entire CUDA warp.

DirectX Sample browser shows blank screen on the right

isvicbhasme — Fri, 17 Feb 2012 09:54:50 -0500

Hi, I installed DirectX SDK (June2010) and when I open SampleBrowser.exe I get a blank window. The reason why I am posting it on this forum is because I was told that the problem might lie with the NVIDIA driver.

Kindly view the original discussion at directXForum

I run NVIDIA GeForce 8400M GS graphic card with driver version 285.62
If you need any other information like DirectX Diagnostic report please let me know.

Thank you in advance!

CG Toolkit and DirectX 9 AlphaRef and AlphaFunc (Bug?)

pfeifers — Thu, 23 Feb 2012 13:09:56 -0500

Background:
I am using the latest CG Toolkit (3.1, issue occurs with 3.0, haven't tested previous versions) with DirectX 9. DX9 is a hard requirement for this project, so switching versions is not a possible solution. I am loading and running HLSL and CgFX shaders using CG Toolkit, with shaders that may or may not be supplied by users of my engine.

Issue:
AlphaRef and AlphaFunc CgFX pass states do not seem to set the proper alpha reference values in DX9. When specifying these values for a pass, strange alpha blending artifacts occur (white rim around the edge of alpha-blended textures).
D3DRS_ALPHAREF calls for "an 8-bit value placed in the low 8 bits of the DWORD render-state value. Values can range from 0x00000000 through 0x000000FF" (D3DRS_ALPHAREF reference).
The issue is that if I call cgSetPassState() with any value set for AlphaRef or AlphaFunc, the CG Toolkit will pass a float to the D3DRS_ALPHAREF state assignment, rather than the requested DWORD.

Example:
AlphaRef = 127;
If I debug in PIX, I see the call to set D3DRS_ALPHAREF:
IDirect3DDevice9::SetRenderState( D3DRS_ALPHAREF, 0x42FE0000 )
Which is float 127.0 on a function whose valid range is 0x00000000 through 0x000000FF. Same issue occurs with 0.5f (0x3F000000).

I can get the proper 127 (0x0000007F) int value using an AlphaRef value of 1.780e-043f, however this solution is not feasible, as other users are expected to be able to write shaders for my engine (and pull in shaders from FX Composer and the like).

Unfortunately, it seems as though the CG Toolkit is directly passing the float representation of my values to the IDirect3DDevice9::SetRenderState call. Is there some way around this issue that I am missing?

Multiple pass shaders failing Z-test when antialiasing is forced on with drivers

megaderek — Sat, 18 Feb 2012 00:43:35 -0500

I have some shaders that use multiple passes and when antialiasing is forced on in the drivers then the subsequent shader passes are failing the ztest if zenable is true. If I set zenable to false or set zfunc to always then it succeeds rendering.

I noticed that the DirectX SDK (March 2009) / Direct3D 9 sample program called PostProcess has exactly the same issue.

This happens on my GTX 480 (driver 285.62) and on my GT 420M.
It doesn't happen on my ATI card.

Developing Windows applications for the Tesla GPU

paysonwelch — Fri, 10 Feb 2012 06:51:21 -0500

Greetings I have a very quick question. I am a Windows application developer who needs to crunch lots of data. Can anyone tell me if it is relatively easy to write code that can be offloaded to a Tesla GPU using Visual Studio?

Specifically my application has a GUI front-end, at the very least is it possible to compile a C++ library that can be included without much hassle?

Also any tips or suggestions for research would be greatly appreciated. Thanks in advance.

geforce 8400 gs black_bar artifacts in old games (Dangerous Waters, Falcon Bms4.32)

mstram — Mon, 13 Feb 2012 21:35:39 -0500

Hello,

In all of the u.i. pages of Dangerous Waters 1.04 and Falcon Bms 4.32 I get thin horizontal bars (see attached image).

They are not in the "main sim window" ("3d" fullscreen )

I'm running Win xp sp3, and nvidia drivers ver 6.14.12.7061 (270.61). DirectX 9.0c (4.09.0000.0904), on an HP DC5000 / 1gig / and the geforce 8400 gs. I also tried installing older drivers (169.21), but the problem remains

I'm downloading the Perfkit to see if it will show anything, but having not used it before, even if it does show "something", I'm wondering how that will translate to a particular driver version / setting.

Does anybody here know what the particular DirectX program call / technique that is being used to cause this ?

The card works fine with many other graphic intensive applications (xplane, iRacing, Orbiter, Blender, ClearviewRC...etc)

Mike