Cortex 9, Arnold 5, exploratory PR #1

danieldresser-ie · 2017-09-26T02:23:40Z

No description provided.

This is needed for some of the Arnold tests to pass.

…passed due to substantial watermarks baked into the bright values of the comparison image. We now assume tests will run with a license, and I've regenerated the comparison image

…lues of rendered pixels, to account for floating point error in filter normalization. ( I couldn't figure out any easy way to swtich to a box filter )

…amples to reliably converge ( in Arnold 4, the tests were passing due to good luck )

…till exist in Arnold 5

…requires undocumented functions

…red in the header, and instead use accessor functions

…for Arnold 5

…terface

…tructor for the writable UniverseBlock will always call it anyway

…ng to float

…to no output In Gaffer::IECoreArnoldPreview or Cortex10, there are options to set verbosity Currently, without this, we would dump a huge pile of output every time you made a UniverseBlock just to query metadata or something.

… can check it ( There is no way to do this through Cortex 9 at the moment, so we use Arnold's python API )

johnhaddon

All looks good to me - just minor comments that aren't particularly important. Thanks for getting everything up to date, even for the stuff that'll get deleted pretty soon. My plan is to port over those existing tests to work with the IECoreArnoldPreview::Renderer when that moves to Cortex, so having them updated is great.

johnhaddon · 2017-09-27T15:20:34Z

SConstruct

 		arnoldTest = arnoldTestEnv.Command( "contrib/IECoreArnold/test/IECoreArnold/results.txt", arnoldPythonModule, pythonExecutable + " $TEST_ARNOLD_SCRIPT" )
 		NoCache( arnoldTest )
-		arnoldTestEnv.Depends( arnoldTest, [ arnoldPythonModule + arnoldProceduralForTest + arnoldDriverForTest ] )
+		arnoldTestEnv.Depends( arnoldTest, [ arnoldPythonModule + arnoldProceduralForTest + arnoldDriverForTest + arnoldLibrary ] )


I'm surprised this is necessary, since arnoldPythonModule already depends on arnoldLibrary.

The issue is the cache. If you make a change to arnoldLibrary that does not produce any changes in the python/IECoreArnold/_IECoreArnold.so binary, then the cache considers the test to have already been run with this exact set of inputs.

johnhaddon · 2017-09-27T15:32:32Z

contrib/IECoreArnold/src/IECoreArnold/ParameterAlgo.cpp

 				{
 					const Imath::M44f v( data->readable() );
-					AiNodeSetMatrix( node, name, const_cast<float (*)[4]>( v.x ) );
+					AiNodeSetMatrix( node, name, *reinterpret_cast<AtMatrix*>( const_cast<float(*)[4][4]>( &v.x ) ) );


Do we really need all the nested casting? I don't suppose reinterpret_cast<const AtMatrix &>( v ) work?

I think at the time, I felt better about being more explicit about the reason this cast works is that the classes can be cast to float[4][4] - but you're right that this is pretty ugly. I'll go ahead and change this

johnhaddon · 2017-09-27T15:40:21Z

contrib/IECoreArnold/include/IECoreArnold/NodeAlgo.h

+/// Arnold does not support non-uniform sampling.  It just takes a start and end time, and assume
+/// the samples are distributed evenly between them.  We need to throw an exception if given data
+/// we can't render.
+void ensureUniformTimeSamples( const std::vector<float> &times );


Can we just make this internal to the Renderer class? I'm trying to minimise the API surface area...

I currently use this from both IECoreArnold::RendererImplementation and GafferArnold::IECoreArnoldPreview::Renderer, so I need it to be public. I can add a /todo to make it private in Cortex 10 when we don't have this split.

johnhaddon · 2017-09-27T15:42:30Z

contrib/IECoreArnold/src/IECoreArnold/RendererImplementation.cpp

 	}

-	m_motionTimes.insert( m_motionTimes.end(), times.begin(), times.end() );
+	// TODO - is it worth the time to do this, in order to avoid silently giving incorrect results?


Seems unlikely to have significant overhead compared to the rest of the work we're doing.

johnhaddon · 2017-09-27T15:43:23Z

contrib/IECoreArnold/src/IECoreArnold/RendererImplementation.cpp

+	NodeAlgo::ensureUniformTimeSamples( timesVector );
+
+	m_motionStart = *times.begin();
+	m_motionEnd = *(--times.end());


timesVector.back() might be a bit more idiomatic?

johnhaddon · 2017-09-27T15:44:54Z

contrib/IECoreArnold/src/IECoreArnold/RendererImplementation.cpp


 	const FloatData *pixelAspectRatio = cortexCamera->parametersData()->member<FloatData>( "pixelAspectRatio" );
-	AiNodeSetFlt( options, "aspect_ratio", 1.0f / pixelAspectRatio->readable() ); // arnold is y/x, we're x/y
+	AiNodeSetFlt( options, "pixel_aspect_ratio", 1.0f / pixelAspectRatio->readable() ); // arnold is y/x, we're x/y


I could have sworn I saw something saying that they'd also changed the interpretation to be x/y - am I wrong about that?

Ah - good catch, thanks, will fix.

johnhaddon · 2017-09-27T15:48:30Z

contrib/IECoreArnold/src/IECoreArnold/RendererImplementation.cpp

+		// just like built in nodes - we don't reference the filename of the dso that defines the node type.
+		// So here we just interpret "filename" as the node type to create.
+		// Where should this be documented.  Should we change the name of the parameter to ExternalProcedural
+		// to be something other than "filename" in Cortex 10?


Seems reasonable, maybe just calling it "name" is generic enough? Let's not mix that up with the Arnold update though - what you have here seems fine for now.

johnhaddon · 2017-09-27T15:53:09Z

contrib/IECoreArnold/src/IECoreArnold/procedural/Procedural.cpp

+	AiParameterArray( "parameterValues", AiArray(0, 1, AI_TYPE_STRING) );
+}
+
+procedural_init


For what it's worth, I hate the way Arnold uses macros to obfuscate the function parameters, and I explicitly wrote out the signatures instead so they were obvious.

OK, though I liked having less typing ...

johnhaddon · 2017-09-27T15:55:58Z

contrib/IECoreArnold/src/IECoreArnold/UniverseBlock.cpp

 void begin()
 {
 	AiBegin();
+	// Default to logging errors / warnings only - we may not even be using this universe block to perform a render,


So Arnold 5 defaults to all logging being on?

Not sure if it is all logging, maybe it was reporting the paths it was loading shaders from into "info"? It was a lot of annoying output every time you instantiated a universe block anyway

`DisplayTileCallbackFactory::create()` returned the same DisplayTileCallback instance every time, and up until now took responsibility for deleting it. This has always been sketchy because Appleseed holds the returned DisplayTileCallback's with `auto_release_ptr`, so `DisplayTileCallback::release()` must avoid double deletes by just doing nothing. This worked if `DisplayTileCallback::release()` was called before `DisplayTileCallbackFactory::release()`, but Appleseed 2.1 has reversed the call order so that the factory is destroyed before the callback is. This means `DisplayTileCallback::release()` is called on an already-deleted instance, and crashes, with a stack trace like so : ``` #0 0x0000000000000000 in ?? () #1 0x00007fffe18153a6 in foundation::auto_release_ptr<renderer::ITileCallback>::reset (this=0x7fff8c012148, ptr=0x0) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/foundation/utility/autoreleaseptr.h:187 ImageEngine#2 0x00007fffe181137f in renderer::(anonymous namespace)::ProgressiveFrameRenderer::~ProgressiveFrameRenderer (this=0x7fff8c0120a0, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/kernel/rendering/progressive/progressiveframerenderer.cpp:178 ImageEngine#3 0x00007fffe1811592 in renderer::(anonymous namespace)::ProgressiveFrameRenderer::~ProgressiveFrameRenderer (this=0x7fff8c0120a0, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/kernel/rendering/progressive/progressiveframerenderer.cpp:187 ImageEngine#4 0x00007fffe18115ca in renderer::(anonymous namespace)::ProgressiveFrameRenderer::release (this=0x7fff8c0120a0) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/kernel/rendering/progressive/progressiveframerenderer.cpp:191 ImageEngine#5 0x00007fffe1712fb7 in foundation::auto_release_ptr<renderer::IFrameRenderer>::~auto_release_ptr (this=0x7fff8c01f928, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/foundation/utility/autoreleaseptr.h:124 ImageEngine#6 0x00007fffe17121be in renderer::RendererComponents::~RendererComponents (this=0x7fff8c01f890, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/kernel/rendering/renderercomponents.h:73 ImageEngine#7 0x00007fffe1712288 in std::default_delete<renderer::RendererComponents>::operator() (this=0x7fff8c000ae0, __ptr=0x7fff8c01f890) at /opt/rh/devtoolset-6/root/usr/include/c++/6.3.1/bits/unique_ptr.h:76 ImageEngine#8 0x00007fffe1711e0d in std::unique_ptr<renderer::RendererComponents, std::default_delete<renderer::RendererComponents> >::reset (this=0x7fff8c000ae0, __p=0x7fff8c01f890) at /opt/rh/devtoolset-6/root/usr/include/c++/6.3.1/bits/unique_ptr.h:347 ImageEngine#9 0x00007fffe1710632 in renderer::CPURenderDevice::~CPURenderDevice (this=0x7fff8c0009b0, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/device/cpu/cpurenderdevice.cpp:112 ImageEngine#10 0x00007fffe1710988 in renderer::CPURenderDevice::~CPURenderDevice (this=0x7fff8c0009b0, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/device/cpu/cpurenderdevice.cpp:129 ImageEngine#11 0x00007fffe181f21c in std::default_delete<renderer::IRenderDevice>::operator() (this=0x3093868, __ptr=0x7fff8c0009b0) at /opt/rh/devtoolset-6/root/usr/include/c++/6.3.1/bits/unique_ptr.h:76 ImageEngine#12 0x00007fffe181eafd in std::unique_ptr<renderer::IRenderDevice, std::default_delete<renderer::IRenderDevice> >::~unique_ptr (this=0x3093868, __in_chrg=<optimized out>) at /opt/rh/devtoolset-6/root/usr/include/c++/6.3.1/bits/unique_ptr.h:239 ImageEngine#13 0x00007fffe181d545 in renderer::MasterRenderer::Impl::~Impl (this=0x3093820, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/kernel/rendering/masterrenderer.cpp:168 ImageEngine#14 0x00007fffe181cfdb in renderer::MasterRenderer::~MasterRenderer (this=0x3088d70, __in_chrg=<optimized out>) at /disk1/john/dev/gafferDependencies/Appleseed/working/appleseed-2.1.0-beta/src/appleseed/renderer/kernel/rendering/masterrenderer.cpp:584 ImageEngine#15 0x00007fffc67d0ae0 in (anonymous namespace)::AppleseedRenderer::~AppleseedRenderer() () from /disk1/john/dev/build/gaffer/lib/libGafferAppleseed.so ImageEngine#16 0x00007fffd8a734b6 in GafferScene::InteractiveRender::stop() () from /disk1/john/dev/build/gaffer/lib/libGafferScene.so ImageEngine#17 0x00007fffd8a735bf in GafferScene::InteractiveRender::update() () from /disk1/john/dev/build/gaffer/lib/libGafferScene.so ``` The solution here is to return a new DisplayTileCallback every time `DisplayTileCallbackFactory::create()` is called, and to destroy those the regular way in `DisplayTileCallback::release()`. This works fine for interactive renders, because Appleseed will only ever make a single DisplayTileCallback at a time. For batch renders it is completely broken though, because Appleseed will call `DisplayTileCallbackFactory::create()` once per thread, and every thread will end up writing to a different display driver. So far we've only used the display driver for interactive renders though, so maybe this is OK.

…r is computing. I noticed crashes when the SceneCacheReader property panel was displayed and you would change frame or filter the scene tag. The following traceback indicates that it was related with refresh the SceneView widget. This is likely something fundamentally flawed with the node ( as Nuke often complains we are updating knob inside the _validate). Futhermore, we should be able to use native Nuke node once we start using Nuke 14+ and its USD support ( using our SceneCacheDataFormat AbstractFormat USD plugin ). The trick is to avoid updating the widget when it's already up to date. While testing,I couldnt reproduce the crash I saw when changing the filtering and the "loading" of location in the Nuke 3d system. ``` 0 0x0000000080000017 in ?? () 1 0x00007fa26c59fee2 in QObject::disconnect(QObject const*, char const*, QObject const*, char const*) () from /software/apps/nuke/13.2v8/cent7.x86_64/libQt5Core.so.5 2 0x00007fa26d341e90 in QTreeView::setModel(QAbstractItemModel*) () from /software/apps/nuke/13.2v8/cent7.x86_64/libQt5Widgets.so.5 3 0x00007fa25b1a089c in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libnuke-13.2.8.so 4 0x00007fa25aae2656 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libnuke-13.2.8.so 5 0x00007fa25aaec3f9 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libnuke-13.2.8.so 6 0x00007fa175dc56e9 in IECoreNuke::SceneCacheReader::filterScene(std::string const&, std::string const&, bool) () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 7 0x00007fa175dc6065 in IECoreNuke::SceneCacheReader::_validate(bool) () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 8 0x00007fa2597faf3b in DD::Image::Op::force_validate(bool) () from /software/apps/nuke/13.2v8/cent7.x86_64/libDDImage.so 9 0x00007fa175da9794 in IECoreNuke::LiveScene::geometryList(DD::Image::Op*, double const&) const () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 10 0x00007fa175da9dc9 in IECoreNuke::LiveScene::geometryList(double const&) const () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 11 0x00007fa175daa860 in IECoreNuke::LiveScene::cacheGeometryList(double const&) const () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 12 0x00007fa175dab8f2 in IECoreNuke::LiveScene::objectNum(double const*) const () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 13 0x00007fa175dacedb in IECoreNuke::LiveScene::childNames(std::vector<IECore::InternedString, std::allocator<IECore::InternedString> >&) const () from /software/apps/cortex/10.4.10.0/cent7.x86_64/nuke/13.2/lib/libIECoreNuke-10.4.so 14 0x00007fa1298fbca9 in CaribouNuke::CaribouSceneReader::computeChildNames(std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > const&, Gaffer::Context const*, GafferScene::ScenePlug const*) const () from /home/lucienf/apps/caribou/4.2.2dev/gaffer/1.2/tools/nuke/13.2/python/cent7.x86_64/cortex/10.4/CaribouNuke/_CaribouNuke.so 15 0x00007fa13530d444 in GafferScene::SceneNode::compute(Gaffer::ValuePlug*, Gaffer::Context const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGafferScene.so 16 0x00007fa148ec5340 in Gaffer::ValuePlug::ComputeProcess::ComputeProcess((anonymous namespace)::ComputeProcessKey const&) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 17 0x00007fa148ece618 in Gaffer::ValuePlug::ComputeProcess::value(Gaffer::ValuePlug const*, IECore::MurmurHash const*) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 18 0x00007fa148ec7f6a in Gaffer::ValuePlug::getValueInternal(IECore::MurmurHash const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 19 0x00007fa148eba781 in boost::intrusive_ptr<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > const> Gaffer::ValuePlug::getObjectValue<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > >(IECore::MurmurHash const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 20 0x00007fa148ebab7a in Gaffer::TypedObjectPlug<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > >::getValue(IECore::MurmurHash const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 21 0x00007fa148ebabbe in Gaffer::TypedObjectPlug<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > >::setFrom(Gaffer::ValuePlug const*) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 22 0x00007fa148e36256 in Gaffer::NameSwitch::compute(Gaffer::ValuePlug*, Gaffer::Context const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 23 0x00007fa148ec5340 in Gaffer::ValuePlug::ComputeProcess::ComputeProcess((anonymous namespace)::ComputeProcessKey const&) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 24 0x00007fa148ece618 in Gaffer::ValuePlug::ComputeProcess::value(Gaffer::ValuePlug const*, IECore::MurmurHash const*) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 25 0x00007fa148ec7f6a in Gaffer::ValuePlug::getValueInternal(IECore::MurmurHash const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 26 0x00007fa148eba781 in boost::intrusive_ptr<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > const> Gaffer::ValuePlug::getObjectValue<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > >(IECore::MurmurHash const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 27 0x00007fa148ebab7a in Gaffer::TypedObjectPlug<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > >::getValue(IECore::MurmurHash const*) const () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 28 0x00007fa1352c50ca in GafferScene::RenderController::SceneGraph::updateChildren(Gaffer::TypedObjectPlug<IECore::TypedData<std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > > > const*) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGafferScene.so 29 0x00007fa1352d9754 in GafferScene::RenderController::SceneGraph::update(std::vector<IECore::InternedString, std::allocator<IECore::InternedString> > const&, unsigned int, GafferScene::RenderController::SceneGraph::Type, GafferScene::RenderController*) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGafferScene.so 30 0x00007fa1352d9e1b in GafferScene::RenderController::SceneGraphUpdateTask::execute() () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGafferScene.so 31 0x00007fa2568544ef in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 ---Type <return> to continue, or q <return> to quit--- 32 0x00007fa256851b00 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 33 0x00007fa1352bb36e in GafferScene::RenderController::updateInternal(std::function<void (Gaffer::BackgroundTask::Status)> const&, IECore::PathMatcher const*, bool) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGafferScene.so 34 0x00007fa148e3b47b in std::_Function_handler<void (IECore::Canceller const&), Gaffer::ParallelAlgo::callOnBackgroundThread(Gaffer::Plug const*, std::function<void ()>)::{lambda(IECore::Canceller const&)#1}>::_M_invoke(std::_Any_data const&, IECore::Canceller const&) () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 35 0x00007fa148d83dd5 in _ZN3tbb8internal13function_taskIZN6Gaffer14BackgroundTaskC4EPKNS2_4PlugERKSt8functionIFvRKN6IECore9CancellerEEEEUlvE_E7executeEv () from /software/apps/gaffer/1.2.9.0/cent7.x86_64/cortex/10.4/nuke/13.2/lib/libGaffer.so 36 0x00007fa2568544ef in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 37 0x00007fa25684dae4 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 38 0x00007fa25684c603 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 39 0x00007fa256848846 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 40 0x00007fa256848a79 in ?? () from /software/apps/nuke/13.2v8/cent7.x86_64/libtbb.so.2 41 0x00007fa26bed3ea5 in start_thread () from /lib64/libpthread.so.0 42 0x00007fa26b308b0d in clone () from /lib64/libc.so.6 ```

danieldresser added 26 commits September 14, 2017 12:03

IE Config : Add solidangle_LICENSE to ENV_VARS_TO_IMPORT

ddad201

This is needed for some of the Arnold tests to pass.

SConstruct : Arnold tests depend on Arnold lib

ba7a2f9

IECoreArnold::PointsTest and RenderTest : This test could never have …

6e51598

…passed due to substantial watermarks baked into the bright values of the comparison image. We now assume tests will run with a license, and I've regenerated the comparison image

IECoreArnold::ProceduralTest : Need a tiny tolerance when checking va…

1d3e5df

…lues of rendered pixels, to account for floating point error in filter normalization. ( I couldn't figure out any easy way to swtich to a box filter )

IECoreArnold::RendererTest : With extreme motion blur, we need more s…

574ec8a

…amples to reliably converge ( in Arnold 4, the tests were passing due to good luck )

IECoreArnold tests : Don't dump render output during tests

cacc56c

IECoreArnold tests : When testing example parameters, use ones that s…

5320eae

…till exist in Arnold 5

IECoreArnold tests : Signature of AiNodeGetMatrix changed in Arnold 5

86fcb48

IECoreArnold::UniverseBlockTest : Accessing Metadata from Python now …

efa1b5d

…requires undocumented functions

IECoreArnold::RendererTest : Use correct types so we don't get warnings

4130ffa

IECoreArnold : Arnold 5 removes point types, and just uses vectors

73f9798

IECoreArnold : Arnold 5 replaces AtByte with uint8_t

f8d2571

IECoreArnold : Arnold 5 replaces AtColor with AtRGB

c146478

IECoreArnold : In Arnold 5, member variables of AtArray are not decla…

2f06053

…red in the header, and instead use accessor functions

IECoreArnold::OutputDriver : Many API changes in Arnold 5

e1cdb1e

IECoreAnold::Parameter Algo : Use AtString

5765d45

IECoreArnold : Update M44f to AtMatrix casts for Arnold 5

3ab0ea5

IECoreArnold::RendererImplementation : Updated visibility attributes …

4acc625

…for Arnold 5

IECoreArnold : Arnold 5 doesn't support non-uniform sample times

45a5156

IECoreArnold : Arnold 5 renamed aspect_ratio to pixel_aspect_ratio

dde6223

IECoreArnold::RenderImplemenation : Update for Arnold 5 procedural in…

a78e362

…terface

IECoreArnold::Procedural : Update for Arnold 5, remove obselete test

c3771b6

IECoreArnold::RendererImplemention : Remove duplicate AiEnd - the des…

7ad6595

…tructor for the writable UniverseBlock will always call it anyway

IECoreArnold::ParameterAlgo : Support V3i and V2i parameters by casti…

eeb3f92

…ng to float

IECoreArnold::AutomaticInstancingTest : Force on console output so we…

96b708e

… can check it ( There is no way to do this through Cortex 9 at the moment, so we use Arnold's python API )

johnhaddon reviewed Sep 27, 2017

View reviewed changes

Cortex 9, Arnold 5, exploratory PR #1

Are you sure you want to change the base?

Cortex 9, Arnold 5, exploratory PR #1

Uh oh!

Conversation

danieldresser-ie commented Sep 26, 2017

Uh oh!

johnhaddon left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants