Volume Emission Material Overflow(?) on Hard Edges (Cycles) #86742

Open
opened 2021-03-19 16:59:05 +01:00 by Jon Tetrino · 9 comments

System Information
Operating system: Windows-10-10.0.19041-SP0 64 Bits
Graphics card: GeForce RTX 3090/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 461.92

Blender Version
Broken: version: 2.92.0, branch: master, commit date: 2021-02-24 16:25, hash: 02948a2cab
Worked: Unknown, is present in versions back to 2.8

Short description of error
When creating a volume emission material, applying it to an object, and placing said material on an object, fireflies will form on hard mesh edges when there is geometry behind it.

Exact steps for others to reproduce the error
It is difficult to reproduce accurately, and will sometimes not occur. It also seems to be vaguely dependent on distance from origin point of the scene, but I have seen this at close to 0,0,0 and distance is simply more likely to show it, indicating floating point math errors? (Also more visible when scene is scaled to micro levels).

1: Create a primitive or use the default cube, and in its material, create a single Emission node and apply it to the Volume node of the default output.
2: Create a second primitive larger than the first, place it such that it is behind the emission object in your perspective.

If the bug is triggered you will see white fireflies at the hard edges of the silhouette of the emission object as seen in render view and final render, as seen here:
image.png

Bug is also visible in cases of semi-transparency, such as this example with a gradient + noise applied:
image.png

I have included the .blend file these images are from, with a stripped out version of the scene. I have left the more advanced world and emission materials in the hope it helps find the issue.
It can be mitigated by applying a standard Laplacian Smooth modifier to the primitive (and changing no values, probably fixing the math??), however for more advanced objects this isn't feasible.
It seems related to the density of the volume, for if you turn up the density the fireflies lessen or disappear, even by applying a multiplier to the gradient example shown above.
Strength of the fireflies is directly proportional to the strength of the emission of the volume, as seen here:
image.png

What it is not:

  • Mitigated by any other firefly handling techniques
  • Mitigated by using other Volumetric shader options (such as Principled Volume).
  • Related to any post-processing or motion blur as it's visible in Rendered view
  • Related to scene scale
  • Directly related to distance from origin, but it can be more prevalent with distance.

New as of 2021/05/30

  • Volumes showing this issue to not show this issue in direct or indirect reflections (tested with Principled BCDF set for metallic 1.0, roughness zero)
  • Volumes showing this issue still show this issue if the object behind them is set to holdout, either through shader or view layer, though this may not be that surprising depending on how the holdout backend works with geometry.

I surmise there is some kind of overflow happening producing these very specific edge fireflies on object. Whatever it is, it's a disruption to the use of these materials.

See blend file: volume_assistance.blend

Excuse the number of edits, I'm trying to do this right.

**System Information** Operating system: Windows-10-10.0.19041-SP0 64 Bits Graphics card: GeForce RTX 3090/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 461.92 **Blender Version** Broken: version: 2.92.0, branch: master, commit date: 2021-02-24 16:25, hash: `02948a2cab` Worked: *Unknown, is present in versions back to 2.8* **Short description of error** When creating a volume emission material, applying it to an object, and placing said material on an object, fireflies will form on hard mesh edges when there is geometry behind it. **Exact steps for others to reproduce the error** It is difficult to reproduce accurately, and will sometimes not occur. It also seems to be vaguely dependent on distance from origin point of the scene, but I have seen this at close to 0,0,0 and distance is simply more likely to show it, indicating floating point math errors? (Also more visible when scene is scaled to micro levels). 1: Create a primitive or use the default cube, and in its material, create a single Emission node and apply it to the Volume node of the default output. 2: Create a second primitive larger than the first, place it such that it is behind the emission object in your perspective. If the bug is triggered you will see white fireflies at the hard edges of the silhouette of the emission object as seen in render view and final render, as seen here: ![image.png](https://archive.blender.org/developer/F9899693/image.png) Bug is also visible in cases of semi-transparency, such as this example with a gradient + noise applied: ![image.png](https://archive.blender.org/developer/F9899685/image.png) I have included the .blend file these images are from, with a stripped out version of the scene. I have left the more advanced world and emission materials in the hope it helps find the issue. It can be mitigated by applying a standard Laplacian Smooth modifier to the primitive (and changing no values, probably fixing the math??), however for more advanced objects this isn't feasible. It seems related to the density of the volume, for if you turn up the density the fireflies lessen or disappear, **even by applying a multiplier to the gradient example shown above.** Strength of the fireflies is directly proportional to the strength of the emission of the volume, as seen here: ![image.png](https://archive.blender.org/developer/F9899697/image.png) **What it is not:** - Mitigated by any other firefly handling techniques - Mitigated by using other Volumetric shader options (such as Principled Volume). - Related to any post-processing or motion blur as it's visible in Rendered view - Related to scene scale - Directly related to distance from origin, but it can be more prevalent with distance. **New as of 2021/05/30** - Volumes showing this issue to not show this issue in direct or indirect reflections (tested with Principled BCDF set for metallic 1.0, roughness zero) - Volumes showing this issue still show this issue if the object behind them is set to holdout, either through shader or view layer, though this may not be that surprising depending on how the holdout backend works with geometry. I surmise there is some kind of overflow happening producing these very specific edge fireflies on object. Whatever it is, it's a disruption to the use of these materials. See blend file: [volume_assistance.blend](https://archive.blender.org/developer/F9899705/volume_assistance.blend) Excuse the number of edits, I'm trying to do this right.
Author

Added subscriber: @Tetrino

Added subscriber: @Tetrino

#89814 was marked as duplicate of this issue

#89814 was marked as duplicate of this issue
Member

Added subscriber: @lichtwerk

Added subscriber: @lichtwerk
Member

Changed status from 'Needs Triage' to: 'Confirmed'

Changed status from 'Needs Triage' to: 'Confirmed'
Member

Can confirm, also present in 2.93 074b0b6da0.

Can confirm, also present in 2.93 074b0b6da0.
Author

I'm adding an extra to the topic, I tried to find a workaround with holdout layers and it turns out that even when an object behind the volume is is not being rendered, the mere existence of it is enough to trigger this issue. Oddly, it does not occur in any reflections, which work as intended.

I know this isn't likely to be solved until Cycles X gets to volumes, but it's a real ballache in the mean time!

I'm adding an extra to the topic, I tried to find a workaround with holdout layers and it turns out that even when an object behind the volume is is not being rendered, the mere existence of it is enough to trigger this issue. Oddly, it does not occur in any reflections, which work as intended. I know this isn't likely to be solved until Cycles X gets to volumes, but it's a real ballache in the mean time!
Author

Added subscriber: @brecht

Added subscriber: @brecht
Author

Now Cycles X has basic support for this, I loaded the file into that experimental build (committed 09 Jul) and saw the same behaviour.

Tagging @brecht in this only because he's been working in this area of code very recently with Cycles X, so would have a better potential idea on where the issue may lie for a fix during the cycles rewrite. Sorry Brecht I know you're busy as hell, feel free to ignore.

Now Cycles X has basic support for this, I loaded the file into that experimental build (committed 09 Jul) and saw the same behaviour. Tagging @brecht in this only because he's been working in this area of code very recently with Cycles X, so would have a better potential idea on where the issue may lie for a fix during the cycles rewrite. Sorry Brecht I know you're busy as hell, feel free to ignore.

Added subscriber: @SimonThommes

Added subscriber: @SimonThommes
Philipp Oeser removed the
Interest
Render & Cycles
label 2023-02-09 14:02:32 +01:00
Sign in to join this conversation.
No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Interest: X11
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
No Milestone
No project
No Assignees
4 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#86742
No description provided.