Multiscatter GGX converges to different result (Path vs Branched Path) #48819

New Issue

Robert M · 2016-07-09T11:16:24+02:00

Robert M commented

2016-07-09 11:16:24 +02:00

System:
Debian, amd64

Broken: current master

Description:
Multiscatter GGX converges to a different result in Pathtracing vs branched Pathtracing.
It seems that MIS and "Sample all Lights" don't influence the final outcome but it is only related to Multiscatter GGX.
Here are the results, both rendered with 100000 samples.
{F320573}

Steps to reproduce:
Here is a .blend file:
Multiscatter_GGX.blend
Make one render with "Path Tracing" and one render with "Branched Path Tracing", then compare both.

Thank you!

**System:** Debian, amd64 **Broken:** current master **Description:** Multiscatter GGX converges to a different result in Pathtracing vs branched Pathtracing. It seems that MIS and "Sample all Lights" don't influence the final outcome but it is only related to Multiscatter GGX. Here are the results, both rendered with 100000 samples. {[F320573](https://archive.blender.org/developer/F320573/path.png)}![branched_path.png](https://archive.blender.org/developer/F320575/branched_path.png) **Steps to reproduce:** Here is a .blend file: [Multiscatter_GGX.blend](https://archive.blender.org/developer/F320579/Multiscatter_GGX.blend) Make one render with "Path Tracing" and one render with "Branched Path Tracing", then compare both. Thank you!

Robert M commented

2016-07-09 11:16:24 +02:00

Changed status to: 'Open'

Robert M commented

2016-07-09 11:16:24 +02:00

Added subscriber: @robertm

Lukas Stockner was assigned by Sergey Sharybin

2016-07-11 11:27:23 +02:00

Steffen Dünner commented

2017-06-09 17:49:54 +02:00

Added subscriber: @SteffenD

Brecht Van Lommel commented

2017-07-03 05:37:49 +02:00

Added subscriber: @brecht

Brecht Van Lommel commented

2017-07-03 05:37:49 +02:00

Changed status from 'Open' to: 'Resolved'

Brecht Van Lommel closed this issue

2017-07-03 05:37:49 +02:00

Brecht Van Lommel commented

2017-07-03 05:37:49 +02:00

This was fixed in 8cb741a598.

This was fixed in 8cb741a598.

Robert M commented

2017-07-14 17:32:32 +02:00

The issue still persists here with build 0f793ee. @brecht did you render my example .blend file before closing this issue?

The issue still persists here with build 0f793ee. @brecht did you render my example .blend file before closing this issue?

Brecht Van Lommel commented

2017-07-20 15:55:21 +02:00

Changed status from 'Resolved' to: 'Open'

Brecht Van Lommel reopened this issue

2017-07-20 15:55:21 +02:00

Brecht Van Lommel commented

2017-07-20 15:55:21 +02:00

Tested on Linux now and it does still converge to a different result, maybe something compiler specific. With this many samples to resolve noise I wouldn't be surprised if float precision issues have an effect.

Robert M commented

2017-07-20 17:52:49 +02:00

I'm not sure if it is float precision issue, but one can see pretty quickly (after 4000 or 8000 samples) that it will converge to a different result.

Robert M commented

2018-03-16 01:57:49 +01:00

I'm happy to say that with current builds (b0823962e9) this issue is solved. Both methods converge now. The final picture is a bit different from both pictures I originally posted:

I don't know which commit solved this issue, but thank you very much!

I'm happy to say that with current builds (b0823962e9b) this issue is solved. Both methods converge now. The final picture is a bit different from both pictures I originally posted: ![path_new.png](https://archive.blender.org/developer/F2445446/path_new.png) ![branched_path_new.png](https://archive.blender.org/developer/F2445448/branched_path_new.png) I don't know which commit solved this issue, but thank you very much!

Brecht Van Lommel commented

2018-03-16 02:06:27 +01:00

Changed status from 'Open' to: 'Resolved'

Brecht Van Lommel closed this issue

2018-03-16 02:06:27 +01:00

Brecht Van Lommel commented

2018-03-16 02:06:27 +01:00

Thanks for testing, my best guess is d454a44e9, but doesn't really matter.

Thanks for testing, my best guess is d454a44e9, but doesn't really matter.

Sign in to join this conversation.

No Label

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

Multiscatter GGX converges to different result (Path vs Branched Path) #48819