profile
viewpoint
Romain Guy romainguy Google California https://www.curious-creature.com Android UI and graphics engineer at Google.

google/filament 9859

Filament is a real-time physically based rendering engine for Android, iOS, Windows, Linux, macOS and WASM/WebGL

android/android-ktx 7568

A set of Kotlin extensions for Android app development.

romainguy/ViewServer 2501

Local server for Android's HierarchyViewer

BinomialLLC/basis_universal 1504

Basis Universal GPU Texture Codec

romainguy/road-trip 1481

A sample Android application that demonstrates how to implement various visual effects.

romainguy/google-io-2014 1282

Demo for the Material Witness talk I gave at Google I/O.

romainguy/elegant-underline 821

Exploring possible implementations for better underline text decoration on Android

romainguy/kotlin-math 540

Set of Kotlin APIs to make graphics math easier to write

romainguy/sample-materials-shop 395

Sample app for Jetpack Compose Developer Preview 2

delete branch google/filament

delete branch : rg/maven

delete time in 15 hours

push eventgoogle/filament

Romain Guy

commit sha 50e5db4347d92accc7f711e8e67eb669c927daf5

Add tasks to manage Sonatype releases (#3127) * Add tasks to manage Sonatype releases * Edit comments

view details

push time in 15 hours

PR merged google/filament

Add tasks to manage Sonatype releases cla: yes
+30 -2

0 comment

2 changed files

romainguy

pr closed time in 15 hours

push eventgoogle/filament

Romain Guy

commit sha 726787e8fcf64b6b74346521f62355c07b442862

Edit comments

view details

push time in 15 hours

PR opened google/filament

Add tasks to manage Sonatype releases
+29 -3

0 comment

2 changed files

pr created time in 15 hours

create barnchgoogle/filament

branch : rg/maven

created branch time in 15 hours

push eventromainguy/sample-materials-shop

Romain Guy

commit sha 99a54fd6d30abd39fbcc54d2066a57fc8e737ed7

Update to the latest everything

view details

push time in 18 hours

PullRequestReviewEvent
PullRequestReviewEvent

issue commentgoogle/filament

Animations appear to not be sampled correctly

I can reproduce the bug in the latest gltf_viewer. Thanks!

Y0hy0h

comment created time in a day

PullRequestReviewEvent

issue commentgoogle/filament

Depth Support

Yes, Filament supports writing the depth buffer only. You can do this in your materials by disabling color writes. See the docs for materials: https://google.github.io/filament/Materials.html

HemanParbhakar

comment created time in 2 days

issue commentgoogle/filament

Consider adding stateful interface for glTF animation

@prideout A layer on top would probably be best indeed.

prideout

comment created time in 2 days

issue closedgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

Describe the bug After updating Filament 1.8.1 to 1.9.1 on Android, setting irradiance will cause a SIGSEGV. A/libc: Fatal signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 0x10007

To Reproduce Here is a sample project to reproduce the issue. https://github.com/zirman/arcore-filament-example-app/tree/irradience

Expected behavior Should set ambient spherical harmonics for render models.

Screenshots N/A

Desktop (please complete the following information): N/A

Smartphone (please complete the following information): Edit:

  • Device: Samsung S9

  • OS: Android 10

  • Device: Pixel 3

  • OS: Android 11

Additional context N/A

closed time in 2 days

zirman

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

Marking as closed since it's now fixed in main. We'll deliver the fix in 1.9.2.

zirman

comment created time in 2 days

PullRequestReviewEvent

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

Ah that would make sense, we call delete on those refs and they could sometimes be non-initialized.

zirman

comment created time in 2 days

PullRequestReviewEvent

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

CC @pixelflinger in case he has ideas of what could be going on here

zirman

comment created time in 2 days

issue closedgoogle/filament

Occluder shadow receiving issues

Hi, I have a problem using filament to achieve "shadows on invisible occluder blending with background" effects.

Expected: An invisible head, a visible eye-glasses in front it, casting shadows to the invisible head and blend with the background "camera photo".

What I tried with the invisible head before:

  1. Apply a material to the invisible head with color-write-off, depth-write-on. It can occlude eye-glasses, but can't show shadows since the color-write is off.
  2. Apply a material to the invisible head with unlit, color-write-on, depth-write-on, blend-on ("multiply"), output the shadow as base-color so it can mix with the 'real world' background. But the problem is: eye-glasses is opaque, invisible head is now transparent, and rendered after eye-glasses, so it didn't write depth firstly and cannot occlude the eye-glasses. occluder_with_shadows.mat.zip

My problem is, is there any way in filament to promote some specified 'transparent' mesh's rendering order before 'opaque' rendering stage? Or any other ways to accomplish such request? Thanks!

closed time in 2 days

vcxyk

issue closedgoogle/filament

missing vkInvalidateMappedMemoryRanges in filament-android-release.aar

I have an android app which calls filament directly from its native code (not through the java/kotlin wrappers). Built against 1.8.0 built from source, and doing System.loadLibrary("filament-jni") before loading my native code.

This worked with 1.6.0 and 1.7.0. With 1.8.0 I get a

java.lang.UnsatisfiedLinkError: dlopen failed: cannot locate symbol "vkInvalidateMappedMemoryRanges" referenced by ...

error at startup.

This happens both with the filament-android-release.aar built by me locally and with the binary release .aar.

Any ideas?

closed time in 2 days

kpeeters

issue commentgoogle/filament

missing vkInvalidateMappedMemoryRanges in filament-android-release.aar

The current Maven Central builds exclude Maven.

kpeeters

comment created time in 2 days

issue closedgoogle/filament

Math linking errors when building sample code on Ubuntu 20.04

Describe the bug I get linking errors related to math functions when following the instructions in README.md.

To Reproduce

  1. I downloaded filament-20200709-linux.tgz and extracted it.
  2. I read README.md in the directory (section "Linking against Filament") and created the main.cpp
#include <filament/FilamentAPI.h>
#include <filament/Engine.h>

using namespace filament;

int main(int argc, char** argv)
{
    Engine *engine = Engine::create();
    engine->destroy(&engine);
    return 0;
}
  1. I created the Makefile like instructed:
FILAMENT_LIBS=-lfilament -lbackend -lbluegl -lbluevk -lfilabridge -lfilaflat -lutils -lgeometry -lsmol-v -libl
CC=clang++

main: main.o
	$(CC) -Llib/x86_64/ main.o $(FILAMENT_LIBS) -lpthread -lc++ -ldl -o main

main.o: main.cpp
	$(CC) -Iinclude/ -std=c++14 -pthread -c main.cpp

clean:
	rm -f main main.o

.PHONY: clean
  1. I ran make

Expected behavior No compilation errors.

Actual behavior Linking errors related to math functions, eg.:

/usr/bin/ld: ColorGrading.cpp:(.text._ZN8filament7tonemap8UchimuraENS_4math7details5TVec3IfEE+0x176): undefined reference to `__expf_finite'
usr/bin/ld: Color.cpp:(.text._ZN8filament5Color20absorptionAtDistanceERKNS_4math7details5TVec3IfEEf+0x52): undefined reference to `__logf_finite'
Exposure.cpp:(.text._ZN8filament8Exposure5ev100Efff+0x15): undefined reference to `__log2f_finite'`
`/usr/bin/ld: Froxelizer.cpp:(.text._ZN8filament10Froxelizer6updateEv+0x300): undefined reference to `__exp2f_finite'

... and many more, but similar.

Desktop (please complete the following information):

  • OS: Ubuntu 20.04 64-bit
  • GPU: AMD Radeon R9 Nano
  • Backend: Vulkan

Additional context

  • clang version 10.0.0-4ubuntu1
  • gcc version 9.3.0 (Ubuntu 9.3.0-10ubuntu2)

closed time in 2 days

bioglaze

issue commentgoogle/filament

Tombstone in sample-gltf-viewer

Fixed in recent change.

daichen

comment created time in 2 days

issue closedgoogle/filament

Tombstone in sample-gltf-viewer

Describe the bug Tombstone in sample-gltf-viewer

To Reproduce Steps to reproduce the behavior:

  1. keep on double tabing the activity to reload the model
  2. after some times the app crashed SYSTEM_TOMBSTONE@1599705295475.txt

so-files.zip

closed time in 2 days

daichen

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

Looking into this further not holding onto the Stream object shouldn't cause an issue in your case.

@prideout @bejado can you take a close look at this issue? The crash occurs in nBeginFrame and if the stack trace is correct the crash occurs either because we are attempting to use a JNI ref that's already been destroyed. What's strange is that nBeginFrame does not use any JNI ref, but it could be something that nBeginFrames calls (updating external stream comes to mind).

zirman

comment created time in 2 days

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

I think I found the problem in your code in ArCore.kt:

                            setExternalStream(
                                filament.engine,
                                Stream
                                    .Builder()
                                    .stream(arCameraStreamTextureId1.toLong())
                                    .width(width)
                                    .height(height)
                                    .build(filament.engine)
                            )

You are not keeping a reference to the Stream instance. The crash is likely caused because that instance gets garbage collected. Texture doesn't keep a reference to the Java Stream instance.

zirman

comment created time in 2 days

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 causes JNI crash

I can't reproduce the issue with our samples. Is there any chance you could share the code of your app or at least an apk? Or explain a bit more what you are doing when rendering a frame? (uploading textures? using external textures? etc.)

zirman

comment created time in 2 days

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 gets SIGSEGV when setting irradiance

From what I can tell the crash has nothing to do with setting the irradiance. It's another JNI related issue.

zirman

comment created time in 2 days

issue closedgoogle/filament

How to render a video to a Node?

Is your feature request related to a problem? Please describe.

Hello every one ,Before sceneform 1.16.0 version ,i was used chroma_key_video.sfb to create a renderable for puting a video node. But in 1.16.0 version ,still use this format will crash.

the error trace is follows : 2020-09-18 17:36:34.033 15587-15587/com.whyhow.sucailib E/Filament: in static filament::MaterialParser

*filament::details::FMaterial::createParser(backend::Backend, const void *, size_t):493

reason: could not parse the material package

my code is : ModelRenderable.builder().setSource(activity, R.raw.chroma_key_video) .build() .thenAccept( { renderable -> renderable.material.setExternalTexture("videoTexture", texture) renderable.material.setFloat4("keyColor", Color(0.1843f, 1.0f,0.098f))if (!mediaPlayer.isPlaying) { mediaPlayer.start() texture.surfaceTexture.setOnFrameAvailableListener { surfaceTexture: SurfaceTexture -> texture.surfaceTexture.setOnFrameAvailableListener(null) this.renderable = renderable } } })

Describe the solution you'd like i want to know how to use filament rendering video to a node.

Describe alternatives you've considered A clear and concise description of any alternative solutions or features you've considered. Is it compatible with the sfb format rendering model? I would be very grateful to you for making these compatible OS and backend android 10 ,arcore v1.16.0 ,sceneform v1.16.0,filament v1.7.0

Looking forward to your reply

closed time in 2 days

steven-gao

issue commentgoogle/filament

How to render a video to a Node?

This is a Sceneform question, not Filament. It seems like sfb files are not supported anymore? Filament itself does not know what an sfb file is.

steven-gao

comment created time in 2 days

issue closedgoogle/filament

Is it possible to port the c ++ code samples to Android without much change?

Hello, I have a question about filament. I am trying to run an example of c ++ in android, I get error in the libraries; #include <filamentapp / Config.h> and #include <filamentapp / FilamentApp.h>. I do not know if this is not possible because it is android or there is an option that I am missing when I build filament or there is simply a way to obtain a similar functionality in android to be able to port these codes without many changes to Android.

Thx.

OS and backend osx 10.15.6 Android

closed time in 2 days

Nand2com

issue commentgoogle/filament

Is it possible to port the c ++ code samples to Android without much change?

Porting the samples wouldn't be difficult but you can't use FilamentApp as is. It's meant to be used on desktop. I recommend you start with our Android samples.

Nand2com

comment created time in 2 days

issue commentgoogle/filament

Updating Filament 1.8.1 to 1.9.1 gets SIGSEGV when setting irradiance

Please post the full stack trace/logs. I have not observed this issue in our own tests/samples.

zirman

comment created time in 2 days

PullRequestReviewEvent
PullRequestReviewEvent

issue commentgoogle/filament

Sceneform memory bug

You can use AssetLoader.createInstancedAsset() to generate multiple FilamentAsset instances. Each FilamentAsset gives you a hierarchy of entities you can add to a scene.

Consti10

comment created time in 3 days

issue closedgoogle/filament

Sceneform memory bug

Sceneform in its latest version does not properly release the gpu resources when switching the displayed renderable for a Node. AssetLoader.destroyAsset() is never called. I implemented a quick fix here: repo by calling this method when a RenderableInstance goes out of scope.

However, I am not sure if a FilamentAsset should be created for each RenderableInstance. Can you draw one filamentAsset multiple times in a scene (with different transforms) ?

If so, is there a filament example how to do so ?

closed time in 3 days

Consti10

issue commentgoogle/filament

Sceneform memory bug

I don't know about the specifics of Sceneform's implementation. What is your question/issue for Filament?

Consti10

comment created time in 3 days

issue commentgoogle/filament

How to Add "#version 430 compatibility" in opengl backend material ?

(If you want to do so look inside CodeGenerator.cpp in libs/filamat).

bitTalk1

comment created time in 4 days

issue commentgoogle/filament

How to Add "#version 430 compatibility" in opengl backend material ?

That's old-style GL and we don't support this. You can modify Filament yourself if you wish though.

bitTalk1

comment created time in 4 days

PullRequestReviewEvent

Pull request review commentgoogle/filament

imageio: add opt-in for ASTC / ETC support.

 if (WIN32)     target_link_libraries(${TARGET} PRIVATE wsock32) endif() +# ==================================================================================================+# Transitive macro definitions+# ==================================================================================================+target_compile_definitions(${TARGET} PUBLIC IMAGEIO_SUPPORTS_BLOCK_COMPRESSION)

That's my question, how does this make it optional? Our code doesn't check for this define in imageio.

prideout

comment created time in 4 days

PullRequestReviewEvent

Pull request review commentgoogle/filament

imageio: add opt-in for ASTC / ETC support.

 if (WIN32)     target_link_libraries(${TARGET} PRIVATE wsock32) endif() +# ==================================================================================================+# Transitive macro definitions+# ==================================================================================================+target_compile_definitions(${TARGET} PUBLIC IMAGEIO_SUPPORTS_BLOCK_COMPRESSION)

Shouldn't you not link against astcenc when IMAGEIO_SUPPORTS_BLOCK_COMPRESSION isn't set? cmgen will depend on imageio and imageio compils BlockCompression.cpp

prideout

comment created time in 4 days

PullRequestReviewEvent
PullRequestReviewEvent
PullRequestReviewEvent
PullRequestReviewEvent

Pull request review commentgoogle/filament

Flip the shading normal when det < 0.

 void FScene::updateUBOs(utils::Range<uint32_t> visibleRenderables, backend::Hand         mat3f m = mat3f::getTransformForNormals(model.upperLeft());         m *= mat3f(1.0f / std::sqrt(max(float3{length2(m[0]), length2(m[1]), length2(m[2])}))); +        // The shading normal must be flipped for mirror transformations.+        // Basically we're shading the other side of the polygon and therefore need to negate the+        // normal, similar to what we already do to support double-sided lighting.+        if (sceneData.elementAt<REVERSED_WINDING_ORDER>(i)) {+            m = -m;

That's with a negative scale. It reminds me of double sidedness where you can mirror or not mirror and I couldn't find a common solution (Unity HDRP lets you choose which behavior you want).

prideout

comment created time in 4 days

PullRequestReviewEvent

Pull request review commentgoogle/filament

gltfio: Fix "uniform not found" errors.

 bool operator==(const MaterialKey& k1, const MaterialKey& k2) {         (k1.unlit == k2.unlit) &&         (k1.hasVertexColors == k2.hasVertexColors) &&         (k1.hasBaseColorTexture == k2.hasBaseColorTexture) &&-        (k1.hasMetallicRoughnessTexture == k2.hasMetallicRoughnessTexture) &&         (k1.hasNormalTexture == k2.hasNormalTexture) &&         (k1.hasOcclusionTexture == k2.hasOcclusionTexture) &&         (k1.hasEmissiveTexture == k2.hasEmissiveTexture) &&+        (k1.useSpecularGlossiness == k2.useSpecularGlossiness) &&         (k1.alphaMode == k2.alphaMode) &&-        (k1.baseColorUV == k2.baseColorUV) &&+        (k1.hasMetallicRoughnessTexture == k2.hasMetallicRoughnessTexture) &&         (k1.metallicRoughnessUV == k2.metallicRoughnessUV) &&-        (k1.emissiveUV == k2.emissiveUV) &&-        (k1.aoUV == k2.aoUV) &&-        (k1.normalUV == k2.normalUV) &&-        (k1.hasClearCoat == k2.hasClearCoat) &&+        (k1.baseColorUV == k2.baseColorUV) &&         (k1.hasClearCoatTexture == k2.hasClearCoatTexture) &&+        (k1.clearCoatUV == k2.clearCoatUV) &&         (k1.hasClearCoatRoughnessTexture == k2.hasClearCoatRoughnessTexture) &&+        (k1.clearCoatRoughnessUV == k2.clearCoatRoughnessUV) &&         (k1.hasClearCoatNormalTexture == k2.hasClearCoatNormalTexture) &&+        (k1.clearCoatNormalUV == k2.clearCoatNormalUV) &&+        (k1.hasClearCoat == k2.hasClearCoat) &&         (k1.hasTransmission == k2.hasTransmission) &&+        (k1.hasTextureTransforms == k2.hasTextureTransforms) &&+        (k1.emissiveUV == k2.emissiveUV) &&+        (k1.aoUV == k2.aoUV) &&+        (k1.normalUV == k2.normalUV) &&         (k1.hasTransmissionTexture == k2.hasTransmissionTexture) &&-        (k1.clearCoatUV == k2.clearCoatUV) &&-        (k1.clearCoatRoughnessUV == k2.clearCoatRoughnessUV) &&-        (k1.clearCoatNormalUV == k2.clearCoatNormalUV);+        (k1.transmissionUV == k2.transmissionUV);

That's why I hate writing comparison functions for structs. I wish there was a safer way of doing this :/

prideout

comment created time in 4 days

PullRequestReviewEvent
PullRequestReviewEvent
PullRequestReviewEvent

issue commentgoogle/filament

How to Add "#version 430 compatibility" in opengl backend material ?

Why do you need to add this? You'd have to modify the filamat library to achieve this.

bitTalk1

comment created time in 4 days

delete branch google/filament

delete branch : rg/clearcoat

delete time in 4 days

push eventgoogle/filament

Romain Guy

commit sha 9830925a965320c199c87f21eef83a94664f2f68

Disable clear coat layer IOR change in glTF files (#3104) The glTF spec doesn't call for this and there are extensions to deal with this manually.

view details

push time in 4 days

PR merged google/filament

Disable clear coat layer IOR change in glTF files cla: yes

The glTF spec doesn't call for this and there are extensions to deal with this manually.

Screen Shot 2020-09-21 at 5 07 44 PM

+2 -0

0 comment

2 changed files

romainguy

pr closed time in 4 days

Pull request review commentgoogle/filament

Fix the normal matrix when det < 0.

 void FScene::updateUBOs(utils::Range<uint32_t> visibleRenderables, backend::Hand         mat3f m = mat3f::getTransformForNormals(model.upperLeft());         m *= mat3f(1.0f / std::sqrt(max(float3{length2(m[0]), length2(m[1]), length2(m[2])}))); +        // Since our "getTransformForNormals" matrix omits the divide-by-determinant step,+        // it has the wrong sign when the determinant is negative.+        if (sceneData.elementAt<REVERSED_WINDING_ORDER>(i)) {

Should we do this in getTransformForNormals?

prideout

comment created time in 5 days

PullRequestReviewEvent
PullRequestReviewEvent

pull request commentKhronosGroup/glTF-Sample-Models

Test model for KHR_materials_specular and KHR_materials_ior

I realized we were still using our option to change the base layer's color when clear coat is enabled (since enabling clear coat creates a transition from IOR=1.5 instead of IOR=1). With this fixed we're closer to the references (still missing support for the IOR/specular extensions though):

Screen Shot 2020-09-21 at 5 07 44 PM

proog128

comment created time in 5 days

PR opened google/filament

Reviewers
Disable clear coat layer IOR change in glTF files

The glTF spec doesn't call for this and there are extensions to deal with this manually.

Screen Shot 2020-09-21 at 5 07 44 PM

+2 -0

0 comment

2 changed files

pr created time in 5 days

create barnchgoogle/filament

branch : rg/clearcoat

created branch time in 5 days

pull request commentKhronosGroup/glTF-Sample-Models

Test model for KHR_materials_specular and KHR_materials_ior

What is the IBL used in all the reference images btw? I'd like to replicate the same lighting.

proog128

comment created time in 5 days

issue closedgoogle/filament

error: use of undeclared identifier 'SREV_TAM'

Describe the bug I try to load a material but I get an error with the following expression in android studio:

static constexpr uint8_t DEFAULT_MATERIAL_PACKAGE [] = {

 #include "Assets/materials/aiDefaultMat.filamat"

};

The error I get is: "error: use of undeclared identifier 'SREV_TAM'" SREV_TAM is the first string that appears in the aiDefaultMat.filamat file.

To Reproduce

I use the matc tool and the aiDefaultMat.mat material that comes in the examples to create aiDefaultMat.filamat, I try to load the material and it gives me an error. I am using c ++("-std=c++17").

Expected behavior I think the error is due to the fact that the version of c ++ that I am using does not recognize the expression "#include" to load the material file.

Additional context

The operating system is OSX Catalina

closed time in 5 days

Nand2com

issue commentgoogle/filament

error: use of undeclared identifier 'SREV_TAM'

You can't do it this way. When you invoke matc you need to use -f header to produce a file that can be included in a C++ file. This is not the preferred way of loading materials. Instead you should just read them like normal binary files.

Nand2com

comment created time in 5 days

pull request commentKhronosGroup/glTF-Sample-Models

Test model for KHR_materials_specular and KHR_materials_ior

Thanks @proog128! Here is the result in Filament:

Screen Shot 2020-09-21 at 9 05 39 AM

Incidentally my changes to support KHR_materials_ior and KHR_materials_specular have been merged to clgtf so I should be able to show better results for this model fairly soon.

proog128

comment created time in 5 days

PullRequestReviewEvent
PullRequestReviewEvent

issue closedgoogle/filament

Depth Support

Does Filament support object occlusion when combined with arcore?

closed time in 6 days

HemanParbhakar

issue commentgoogle/filament

Depth Support

Yes you can render to the depth buffer or do the occlusion yourself in your materials.

HemanParbhakar

comment created time in 6 days

PullRequestReviewEvent

issue closedgoogle/filament

libfilament-jni.so', missing and no known rule to make it

Describe the bug A clear and concise description of what the bug is.

I run the sample-hello-camera occurs error,the log is ,please help me ,thank you Build command failed. Error while executing process /Users/wh-gaotaiwen/Library/Android/sdk/cmake/3.10.2.4988404/bin/ninja with arguments {-C /Users/wh-gaotaiwen/Desktop/ar-pro/filament-main/android/filament-android/.cxx/cmake/debug/arm64-v8a filament-jni} ninja: Entering directory `/Users/wh-gaotaiwen/Desktop/ar-pro/filament-main/android/filament-android/.cxx/cmake/debug/arm64-v8a'

ninja: error: '/Users/wh-gaotaiwen/Desktop/ar-pro/filament-main/out/android-release/filament/lib/arm64-v8a/libfilament.a', needed by '../../../../build/intermediates/cmake/debug/obj/arm64-v8a/libfilament-jni.so', missing and no known rule to make it

To Reproduce Steps to reproduce the behavior: 1.cd filament-main/build/android 2. invoke "./build.sh release" 3 run sample-hello-camera

Expected behavior A clear and concise description of what you expected to happen. the app is run successfuly Screenshots If applicable, add screenshots to help explain your problem. 截屏2020-09-19 下午11 04 33

Desktop (please complete the following information):

  • OS: [e.g. iOS] macOs 10.15.6 (19G2021)
  • GPU: [e.g. NVIDIA GTX 1080] Intel HD Graphics 6000 1536 MB
  • Backend: [OpenGL/Vulkan]

Smartphone (please complete the following information):

  • Device: [e.g. Pixel 2] huawei mate20
  • OS: [e.g. Android Pie 9.0] android 10 Additional context Add any other context about the problem here.

closed time in 7 days

steven-gao

issue commentgoogle/filament

libfilament-jni.so', missing and no known rule to make it

Please refer to BUILDING.md for instructions. This script is for CI only. Use the root build.sh instead.

steven-gao

comment created time in 7 days

PullRequestReviewEvent

issue closedgoogle/filament

When using filament to draw a line, it will flash and you can check the effect in the video。

When using Filament to draw a line, the line will flash。 ezgif-1-34f9b442427f

closed time in 8 days

shlzxjp

issue commentgoogle/filament

When using filament to draw a line, it will flash and you can check the effect in the video。

That’s because of dithering and fxaa together. You’ll need to use TAA instead or disable dithering.

shlzxjp

comment created time in 8 days

push eventgoogle/filament

Romain Guy

commit sha 60ef6581c1f1c84e183f770eece0fc2517815035

Fix compilation issue with clang 12

view details

push time in 9 days

issue commentgoogle/filament

range-loop-analysis error on Apple Clang 12

clang 12 is out and I can reproduce the error. I will submit a fix.

alexey-pelykh

comment created time in 9 days

issue commentgoogle/filament

Minimum graphics card version for Filament library - macOS

Yes it should work. Our target version is macOS 10.14.

Pankaj003

comment created time in 9 days

issue commentgoogle/filament

Callback reference delete errors

Fixed, thank you!

rawnsley

comment created time in 9 days

issue closedgoogle/filament

Callback reference delete errors

I'm getting some jobject / DeleteGlobalRef related errors that I think are due to the mHandler and mCallback fields not being initialized as they are here.

I think it falls over when it calls DeleteGlobalRef on the uninitialized pointers.

I can put more detail around this if required and I'm waiting for a compilation to confirm, but this looks wrong to me.

closed time in 9 days

rawnsley

push eventgoogle/filament

Romain Guy

commit sha 2708b3d0766826a326912183d8e5f419aba05a24

Initialize fields to nullptr to avoid crashes This happens when a Texture with no callback is created.

view details

push time in 9 days

issue commentgoogle/filament

Minimum graphics card version for Filament library - macOS

Why do you need setClearColor? We have equivalent APIs in newer versions. We've made many fixes to the Metal backend since 1.5.

Pankaj003

comment created time in 9 days

push eventgoogle/filament

Romain Guy

commit sha 6dc278c5511cd959bb06edab72a23783993c9dd1

Update cgltf (#3092) Adds support for glTF extensions we want to support: KHR_materials_ior KHR_materials_specular

view details

push time in 9 days

delete branch google/filament

delete branch : rg/cgltf

delete time in 9 days

PR merged google/filament

Reviewers
Update cgltf cla: yes

Adds support for glTF extensions we want to support: KHR_materials_ior KHR_materials_specular

+155 -7

0 comment

4 changed files

romainguy

pr closed time in 9 days

issue commentgoogle/filament

Callback reference delete errors

Right that makes more sense. I'll prepare a fix, thanks!

rawnsley

comment created time in 9 days

PR opened google/filament

Reviewers
Update cgltf

Adds support for glTF extensions we want to support: KHR_materials_ior KHR_materials_specular

+155 -7

0 comment

4 changed files

pr created time in 10 days

create barnchgoogle/filament

branch : rg/cgltf

created branch time in 10 days

pull request commentjkuhlmann/cgltf

Add support for KHR_materials_specular

No need to apologize, I appreciate you taking the time to take a look. Thank you!

romainguy

comment created time in 10 days

PullRequestReviewEvent
PullRequestReviewEvent

issue commentgoogle/filament

Callback reference delete errors

They are not initialized but the only constructor sets them. It could be a copy/move gone wrong (but we delete those constructors) or a caller passing bogus data.

rawnsley

comment created time in 10 days

issue commentgoogle/filament

Minimum graphics card version for Filament library - macOS

These cards are supported. Actually filament should support all cards on macos going back to 6 years ago or so at least.

Pankaj003

comment created time in 10 days

Pull request review commentgoogle/filament

several fog fixes

         public float inScatteringStart = 0.0f;          /**-         * size of in-scattering (>=0 to activate). Good values are >> 1 (e.g. ~10 - 100)+         * size of in-scattering (>0 to activate). Good values are >> 1 (e.g. ~10 - 100)          */-        public float inScatteringSize = 0.0f;+        public float inScatteringSize = -1.0f;

0.0 would work with the new shader right?

pixelflinger

comment created time in 10 days

PullRequestReviewEvent
more