Page MenuHome

Ha Hyung-jin (robobeg)
User

Projects

User does not belong to any projects.

User Details

User Since
Jan 4 2018, 7:57 AM (151 w, 3 d)

Recent Activity

Oct 27 2020

Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Sorry, I left this thread for a while because there were no responses. Are reviewers still interested in this patch? Are there any technical issues other than performance?
I am currently busy in finishing a company project, so I may get some free time only after a couple of months.

Oct 27 2020, 6:22 AM · Render & Cycles, BF Blender (2.83), Cycles

Sep 27 2020

Ilia (reightar) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Sep 27 2020, 12:12 AM · Render & Cycles, BF Blender (2.83), Cycles

Sep 26 2020

kaiwas (kaiwas) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Sep 26 2020, 5:45 PM · Render & Cycles, BF Blender (2.83), Cycles
Nikita (Nikita) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Like token.
Sep 26 2020, 5:31 PM · Render & Cycles, BF Blender (2.83), Cycles
Paul Kotelevets (1D_Inc) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Sep 26 2020, 3:54 PM · Render & Cycles, BF Blender (2.83), Cycles
Viktor Mukayev (vitos1k) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Sep 26 2020, 2:58 PM · Render & Cycles, BF Blender (2.83), Cycles

Jun 19 2020

Alaska (Alaska) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Jun 19 2020, 9:22 AM · Render & Cycles, BF Blender (2.83), Cycles

Apr 19 2020

surya kiran reddy (Dir-Surya) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Burninate token.
Apr 19 2020, 1:09 PM · Render & Cycles, BF Blender (2.83), Cycles

Feb 13 2020

Metin Seven (MetinSeven) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Like token.
Feb 13 2020, 2:50 PM · Render & Cycles, BF Blender (2.83), Cycles

Feb 11 2020

Laurent Noel (c2ba) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Feb 11 2020, 9:57 AM · Render & Cycles, BF Blender (2.83), Cycles

Jan 29 2020

Bintang Senja Pratama (bintang) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Jan 29 2020, 12:58 AM · Render & Cycles, BF Blender (2.83), Cycles

Dec 16 2019

Juan Gea (juang3d) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a Love token.
Dec 16 2019, 10:41 PM · Render & Cycles, BF Blender (2.83), Cycles

Dec 15 2019

Nabil Adel (mywa880) awarded D6250: Fix T43835, T54284: Cycles with no ray offsetting a 100 token.
Dec 15 2019, 10:38 AM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Update the patch to be appliable to the latest revision.

Dec 15 2019, 9:08 AM · Render & Cycles, BF Blender (2.83), Cycles

Dec 9 2019

Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Dec 9 2019, 12:34 PM · Render & Cycles, BF Blender (2.83), Cycles

Dec 3 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

For external path tracing libraries not supporting a feature for estimation of a tight ray start time at the moment, ray_offset() is restored as a fallback. However, self-intersection is still guaranteed to be prevented by this patch for them. Only functions ray_offset() and ray_update_tstart() need to be modified when those libraries are ready to support the feature.

Dec 3 2019, 10:51 PM · Render & Cycles, BF Blender (2.83), Cycles

Dec 2 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

In addition to triangle_update_tstart(), motion_triangle_update_tstart() is implemented to calculate a tight ray start time for motion triangle primitives.
I have completely forgotten triangle vertex motion.

Dec 2 2019, 8:44 PM · Render & Cycles, BF Blender (2.83), Cycles

Dec 1 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

For an AVX2-optimized CPU kernel, bvh traversal routines are set to use triangle_intersect() temporarily instead of triangle_intersect8() for rays with tstart updated.

Dec 1 2019, 3:35 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 30 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

External path tracing libraries now skip triangle_update_tstart() because the function does not work with them at all.
Some OptiX code is modified not to check unnecessary conditions. Optimization for OptiX is still left to experts.

Nov 30 2019, 3:14 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 29 2019

Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 29 2019, 10:29 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Thank @Patrick Mours (pmoursnv) for your advice. I have practically zero knowledge on OptiX.

Nov 29 2019, 4:03 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 29 2019, 1:29 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 29 2019, 1:13 AM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 29 2019, 1:06 AM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 29 2019, 1:04 AM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 29 2019, 1:02 AM · Render & Cycles, BF Blender (2.83), Cycles

Nov 28 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Artifacts caused by mesh overlapping are resolved.

Nov 28 2019, 10:14 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 25 2019

Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Recently I encountered an application which requires not only to render a scene filled with instanced objects correctly with arbitrary scaling but also to deal with objects of overlapped meshes without noticeable artifacts. For that application, my solution was to use this patch's method ( 'primitive id checking' ) with providing 'ray offsetting' as an additional option. I believe that we should put first priority on rendering a normal scene without glitches, but I admit that I am also distracted by mesh overlapping artifacts.

Nov 25 2019, 5:11 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 21 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Cycles object property 'offset_bouncing_ray' and light property 'offset_emitting_ray' are added. Their default values are 'False' but if given 'True' for older blender files after the blender minor version number being raised, all previous versions of blender files would be rendered as before with the mesh overlap issue as less noticed as before. ( Self-intersection avoidance is guaranteed now with this patch, but brightness of hair geom test results will become darken a little. ) I believe this is how blender has evolved. The related UI change is not included, and it seems better for blender staffs to consider the UI layout themselves.

Nov 21 2019, 8:41 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 19 2019

Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Thanks for spotting my last-minute error! I had to re-render benchmark scenes. Here are the results:

Nov 19 2019, 1:40 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 18 2019

Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Logic for preventing self-intersection within a cardinal curve is revised to be consistent with logic for other primitives.
Epsilon check is introduced between adjacent line segments to make the algorithm more numerically robust. Since this epsilon is defined in the curve parameter space [0,1], it will not cause any harm unlike those defined in the world space or the object space.

Nov 18 2019, 6:51 AM · Render & Cycles, BF Blender (2.83), Cycles

Nov 17 2019

Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 17 2019, 6:21 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Now I claim that this patch passed all the cycles regression tests.

Nov 17 2019, 5:56 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 15 2019

Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 15 2019, 5:03 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the diff for D6250: Fix T43835, T54284: Cycles with no ray offsetting.

Many bugs are fixed.

Nov 15 2019, 3:06 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 14 2019

Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.

8>>! In D6250#144466, @Sergey Sharybin (sergey) wrote:

That is exciting to see movement on this front!

Unfortunately, with this change there are some real regressions caught by out regression suit. Just some of them: visibility_particles, ambient_occlusion_only_local, point density_vertices_world, volume_overlap, volume_scatter_albedo.

Nov 14 2019, 3:49 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) added a comment to D6250: Fix T43835, T54284: Cycles with no ray offsetting.

That is exciting to see movement on this front!

Unfortunately, with this change there are some real regressions caught by out regression suit. Just some of them: visibility_particles, ambient_occlusion_only_local, point density_vertices_world, volume_overlap, volume_scatter_albedo.

Nov 14 2019, 11:26 AM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 14 2019, 4:16 AM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 14 2019, 1:47 AM · Render & Cycles, BF Blender (2.83), Cycles

Nov 13 2019

Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 7:41 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 7:21 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 7:17 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 7:16 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 7:03 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) retitled D6250: Fix T43835, T54284: Cycles with no ray offsetting from Fix T43835, T54284: Cycles with no ray offsetting and epsilon to Fix T43835, T54284: Cycles with no ray offsetting and epsilon testing.
Nov 13 2019, 6:35 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 6:27 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) updated the summary of D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 6:12 PM · Render & Cycles, BF Blender (2.83), Cycles
Ha Hyung-jin (robobeg) created D6250: Fix T43835, T54284: Cycles with no ray offsetting.
Nov 13 2019, 6:08 PM · Render & Cycles, BF Blender (2.83), Cycles

Nov 6 2019

Ha Hyung-jin (robobeg) added a comment to T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

Hi robobeg
Yes. Using the the "spring" scene as test again, the same error occures with only the screen driving gpu should render.
So the issue seems not to be the balancing between differently VRAM equipped gpu's, your fix looks okay here.
I now follow your assumption that its another issue.
I can get the scene to render by either reducing the haircount or other simplifications that reduce memory usage.
I think now that the memory used for the attched screens is not taken into account. ( ? )

You might end with the same result as me, is it okay you make a new bugreport out of this find yourself ?
You are the specialist here.

Greetings ... Jens

Nov 6 2019, 2:04 PM · Cycles, Render & Cycles, BF Blender
Ha Hyung-jin (robobeg) added a comment to T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

ㅡ>>! In T71071#806412, @jens verwiebe (jensverwiebe) wrote:

Unfortunately this does not fix the Optix render if one of the gpu's is driving the screen.
I get with 2 exact same RTX 2080 and the spring test scene:

OptiX error OPTIX_ERROR_LAUNCH_FAILURE in optixLaunch(pipelines[PIP_PATH_TRACE], cuda_stream[thread_index], launch_params_ptr, launch_params.data_elements, &sbt_params, wtile.w * wtile.num_samples, wtile.h, 1), line 668

and then only the not screen driving device renders the scene. Cuda seems fine now.

Jens

Nov 6 2019, 1:45 AM · Cycles, Render & Cycles, BF Blender

Nov 5 2019

Ha Hyung-jin (robobeg) updated the diff for D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

Parentheses are added around ‘&&’ within ‘||’ to make it easier to understand and silence strict compiler warning.

Nov 5 2019, 7:32 AM · BF Blender, Render & Cycles

Oct 29 2019

Ha Hyung-jin (robobeg) updated the diff for D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

The style of comments is modified.

Oct 29 2019, 2:32 AM · BF Blender, Render & Cycles

Oct 28 2019

Ha Hyung-jin (robobeg) added a comment to D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

This isn't really a higher level overview of the problem.

Am i right that the issue is that multiple compute devices were using same device pointer, while the code was relying on the fact that it never happens?

Oct 28 2019, 10:41 PM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) added a comment to D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

This isn't really a higher level overview of the problem.

Am i right that the issue is that multiple compute devices were using same device pointer, while the code was relying on the fact that it never happens?

Oct 28 2019, 2:44 PM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) added a comment to D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

The code related to 'device_memory' does not handle this situation properly.

Can you elaborate a bit more about it (on a higher level) ?

Oct 28 2019, 1:32 PM · BF Blender, Render & Cycles

Oct 25 2019

Ha Hyung-jin (robobeg) added projects to T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render: Render & Cycles, Cycles.
Oct 25 2019, 4:21 AM · Cycles, Render & Cycles, BF Blender
Ha Hyung-jin (robobeg) updated the task description for T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 25 2019, 3:48 AM · Cycles, Render & Cycles, BF Blender
Ha Hyung-jin (robobeg) updated the task description for T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 25 2019, 3:47 AM · Cycles, Render & Cycles, BF Blender
Ha Hyung-jin (robobeg) updated the summary of D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 25 2019, 2:11 AM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) updated the diff for D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.

Tab spacings were modified according to surrounding codes.

Oct 25 2019, 2:09 AM · BF Blender, Render & Cycles

Oct 24 2019

Ha Hyung-jin (robobeg) retitled D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render from Fix T71071: Using only one GPU will render successfully, but multiple GPUs fail to render. to Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 2:15 PM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) retitled D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render from Bug fix: Using only one GPU will render successfully, but multiple GPUs fails to render. to Fix T71071: Using only one GPU will render successfully, but multiple GPUs fail to render..
Oct 24 2019, 2:14 PM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) created T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 2:13 PM · Cycles, Render & Cycles, BF Blender
Ha Hyung-jin (robobeg) updated the summary of D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 1:56 PM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) updated the summary of D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 12:42 PM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) updated the summary of D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 9:18 AM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) updated the summary of D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 9:07 AM · BF Blender, Render & Cycles
Ha Hyung-jin (robobeg) created D6126: Fix T71071: Complex scene, one GPU renders successfully, but multiple GPUs fail to render.
Oct 24 2019, 8:50 AM · BF Blender, Render & Cycles

Jan 31 2018

Ha Hyung-jin (robobeg) added a comment to rB7261d675e6ae: Fix Cycles incorrect principled BSDF mixing with some parameter variatons..

Thanks for a quick fix. In the above patch, I found one initialization was missed:
(svm_closure.h, line 336)
bsdf->T = make_float3(0.0f, 0.0f, 0.0f);
and it is not in function 'bsdf_microfacet_ggx_refraction_setup()', either.

Jan 31 2018, 9:02 AM

Jan 30 2018

Ha Hyung-jin (robobeg) added a comment to rB8825a8e951c1: Squashed commit of the following:.

This patch might result in wrong merges of MicrofacetBsdf closures.

Jan 30 2018, 4:40 AM

Jan 28 2018

Ha Hyung-jin (robobeg) added a comment to rC25f9fc71ce53: Code refactor: move more memory allocation logic into device API..

I found that this patch made 'device_only_memory<T>' over-allocate device memory by a factor of sizeof(T). I think that line 248 of 'device_memory.h' should be revised from

data_size = num*sizeof(T);

to

data_size = num;
Jan 28 2018, 3:34 PM

Jan 27 2018

Ha Hyung-jin (robobeg) added a comment to rB095a01a73a35: Cycles: slightly improve BSDF sample stratification for path tracing..

I made a simple example showing the bug.
An emissive sphere is reflected on a floor with two different principled bsdfs. The left half has 'subsurface: 0.000' while the right half has 'subsurface: 0.300', so the latter is expected to look fuzzier.

Jan 27 2018, 12:09 PM

Jan 26 2018

Ha Hyung-jin (robobeg) added a comment to rBd611cf923358: Fix mixed BSDF + BSSRDF sampling bug in path tracing, after 095a01a73a35..

That's it! Thanks for a quick fix!
Although I cannot test the patch at the moment, I am certain that it will work for the 'path tracing' mode because 'shader_bsdf_pick' and 'shader_bssrdf_pick' now perform sampling closures in the same way with the same 'randu' ( = PRNG_BSDF_U ). ( ... besides numerical issues. How about sharing the code body as a common function? )

Jan 26 2018, 8:06 PM
Ha Hyung-jin (robobeg) added a comment to rB095a01a73a35: Cycles: slightly improve BSDF sample stratification for path tracing..

Yes. I will upload the .blend file after this weekend.

Jan 26 2018, 6:14 PM
Ha Hyung-jin (robobeg) added a comment to rB095a01a73a35: Cycles: slightly improve BSDF sample stratification for path tracing..

I encountered an issue related to this patch. Using blender v2.79 I made a test scene having a floor with rough reflection. However, the floor was rendered like a mirror with the blender executable built with the latest master branch. I found that the phenomenon occurred when a single shader generated multiple BSDF closures along with BSSRDF closures.

Jan 26 2018, 5:27 PM

Jan 17 2018

Ha Hyung-jin (robobeg) added a comment to rBfa3d50af95fd: Cycles: Improve denoising speed on GPUs with small tile sizes.

Oh... I've always taken a snapshot of projects/blender.git at the git://git.blender.org/blender.git but it has given me mixed files.
The master version contains the atomic version of util_math_matrix.h, as you said.
Thanks.

Jan 17 2018, 5:38 PM
Ha Hyung-jin (robobeg) added a comment to T52802: Denoise artifacts with Direct Glossy.

I also encountered this problem and found a remedy.

Jan 17 2018, 5:09 PM · Tracker Curfew, Cycles, BF Blender
Ha Hyung-jin (robobeg) added a comment to rBfa3d50af95fd: Cycles: Improve denoising speed on GPUs with small tile sizes.

I have been working on a rendering project, which heavily relies on the cycles denoising feature. ( Thanks a lot~! )

Jan 17 2018, 1:04 PM