Liquid simulation, cache type "Replay", doesn't recognize objects created after Domain creation, doesn't update #75850

Closed
opened 2020-04-17 22:44:05 +02:00 by Nicola De Mitri · 22 comments

System Information
Operating system: Ubuntu 18.04
Graphics card: irrelevant

Blender Version
Broken: 2.83 Beta 8cb10c124e 17 Apr 2020

Short description of error
Under cache type "Replay" (which appears to be the new default) the domain only takes into account fluid objects that were created before creating the domain!

Exact steps for others to reproduce the error

Creating domain first: Doesn't work

  • Default file. (Activate X-Ray)
  • Default cube: Scale up by 2. Add Physics > Fluid > Domain > Type: Liquid
  • Add new cube at origin (inside the initial cube).
  • Add: Fluid > Flow > Type Liquid
  • Right arrow 10 times
    • Expected: FLIP particles pouring from the inner cube
    • Actual: Nothing happens

Creating flow/effectors first: Kind of works

  • Default file. (Activate X-Ray)
  • Default cube: Add: Fluid > Flow > Type Liquid
  • Add new cube at origin. Scale up by 2.
  • Add: Fluid > Domain > Type Liquid
  • Right arrow 10 times (works)

However in this situation I can't make use of the "Replay" advantages because any change I make to the scene won't be taken into account (unless I change cache location)

**System Information** Operating system: Ubuntu 18.04 Graphics card: irrelevant **Blender Version** Broken: 2.83 Beta 8cb10c124efe 17 Apr 2020 **Short description of error** Under cache type "Replay" (which appears to be the new default) the domain only takes into account fluid objects that were created before creating the domain! **Exact steps for others to reproduce the error** Creating domain first: Doesn't work - Default file. (Activate X-Ray) - Default cube: Scale up by 2. Add Physics > Fluid > Domain > Type: Liquid - Add new cube at origin (inside the initial cube). - Add: Fluid > Flow > Type Liquid - Right arrow 10 times - Expected: FLIP particles pouring from the inner cube - Actual: Nothing happens Creating flow/effectors first: Kind of works - Default file. (Activate X-Ray) - Default cube: Add: Fluid > Flow > Type Liquid - Add new cube at origin. Scale up by 2. - Add: Fluid > Domain > Type Liquid - Right arrow 10 times (works) However in this situation I can't make use of the "Replay" advantages because any change I make to the scene won't be taken into account (unless I change cache location)

Added subscriber: @NicolaSap

Added subscriber: @NicolaSap
Member

Added subscriber: @ankitm

Added subscriber: @ankitm
Member

Changed status from 'Needs Triage' to: 'Needs User Info'

Changed status from 'Needs Triage' to: 'Needs User Info'
Member

Creating domain first: Doesn't work

I can see the particles, & I followed the steps in order. I'm using 221a7002a9
Please share a blend file for both cases.

> Creating domain first: Doesn't work I can see the particles, & I followed the steps in order. I'm using 221a7002a9 Please share a blend file for both cases.

I don't have my building environment ready, so I downloaded last night's builds (76b3aac802). I still reproduce the error.

However, if I save and reopen, the particles now show correctly (but the cache is empty). This is an interesting finding, by the way. However, it means that I can't share a file to demonstrate, because opening from file is effectively a workaround.

I'll show a video instead:
Peek 2020-04-18 19-15.mp4

I don't have my building environment ready, so I downloaded last night's builds (76b3aac802f9). I still reproduce the error. However, if I save and reopen, the particles now show correctly (but the cache is empty). This is an interesting finding, by the way. However, it means that I can't share a file to demonstrate, because opening from file is effectively a workaround. I'll show a video instead: [Peek 2020-04-18 19-15.mp4](https://archive.blender.org/developer/F8482152/Peek_2020-04-18_19-15.mp4)

PS: saving alone doesn't help, nor does saving at the very beginning (which would be a problem if cache was somewhere in //, which it isn't). I have to save then re-open to obtain a workaround

PS: saving alone doesn't help, nor does saving at the very beginning (which would be a problem if cache was somewhere in `//`, which it isn't). I have to save then re-open to obtain a workaround
Member

Please try after moving the blender preferences folder outside so that blender starts fresh & you have your preferences safe.
I used 1f1520a045.
Still cannot redo. Making the inner cube flow | liquid shows particles.

Screen Recording 2020-04-19 at 11.47.54.mov

Please try after moving the blender preferences folder outside so that blender starts fresh & you have your preferences safe. I used 1f1520a045. Still cannot redo. Making the inner cube flow | liquid shows particles. [Screen Recording 2020-04-19 at 11.47.54.mov](https://archive.blender.org/developer/F8482926/Screen_Recording_2020-04-19_at_11.47.54.mov)

I've repeated with fresh preferences and still no luck :(

I've repeated with fresh preferences and still no luck :(
Member

Changed status from 'Needs User Info' to: 'Needs Triage'

Changed status from 'Needs User Info' to: 'Needs Triage'

Added subscriber: @iss

Added subscriber: @iss

@NicolaSap does this happen if you modify domain content and press shift + right key?

It seems that if you move playhead my mouse, domain won't update

@NicolaSap does this happen if you modify domain content and press shift + right key? It seems that if you move playhead my mouse, domain won't update

Added subscriber: @mano-wii

Added subscriber: @mano-wii

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

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

I can confirm that there is some strange behavior.
Although there are workarounds (for example updating the cache) this should not happen.

So I am confirming it as a bug.

I can confirm that there is some strange behavior. Although there are workarounds (for example updating the cache) this should not happen. So I am confirming it as a bug.

The issue I was encountering seems to be fixed in the latest builds (d712f1f83a)

The issue I was encountering seems to be fixed in the latest builds (d712f1f83af8)

What is still true, however, is that there isn't a way to invalidate a "Replay" cache once the simulation has started. If I cache the first 3 frames, then I add/remove/move some element, playing the animation still works using the original flows and effectors. In the other simulation engines that allow "Replay" modes (e.g. rigid body, cloth) going back to frame 0 and re-starting the animation would invalidate and recompute the cache, if something has changed in the meanwhile.

I don't know if this (↑) is the "strange behavior" Germano is talking about. If it is, then this Task should probably stay open. If it's not, then probably this Task can be closed and my other concern can be ticketed separately (can/should I do it or is it considered a feature request?)

What is still true, however, is that there isn't a way to invalidate a "Replay" cache once the simulation has started. If I cache the first 3 frames, then I add/remove/move some element, playing the animation still works using the original flows and effectors. In the other simulation engines that allow "Replay" modes (e.g. rigid body, cloth) going back to frame 0 and re-starting the animation would invalidate and recompute the cache, if something has changed in the meanwhile. I don't know if this (↑) is the "strange behavior" Germano is talking about. If it is, then this Task should probably stay open. If it's not, then probably this Task can be closed and my other concern can be ticketed separately (can/should I do it or is it considered a feature request?)
Member

In the other simulation engines that allow "Replay" modes (e.g. rigid body, cloth) going back to frame 0 and re-starting the animation would invalidate and recompute the cache, if something has changed in the meanwhile.

As @iss also said above, shift right or going to the starting frame updates the cache.
If there's isn't anything else @mano-wii you noticed, closing it should be fine.

> In the other simulation engines that allow "Replay" modes (e.g. rigid body, cloth) going back to frame 0 and re-starting the animation would invalidate and recompute the cache, if something has changed in the meanwhile. As @iss also said above, shift right or going to the starting frame updates the cache. If there's isn't anything else @mano-wii you noticed, closing it should be fine.

going to the starting frame updates the cache

This is not true for me!

Peek 2020-04-25 12-24.mp4

shift right

This neither

Peek 2020-04-25 12-25.mp4

(d712f1f83a builds from yesterday)

> going to the starting frame updates the cache This is not true for me! [Peek 2020-04-25 12-24.mp4](https://archive.blender.org/developer/F8494272/Peek_2020-04-25_12-24.mp4) > shift right This neither [Peek 2020-04-25 12-25.mp4](https://archive.blender.org/developer/F8494273/Peek_2020-04-25_12-25.mp4) (d712f1f83af8 builds from yesterday)
Member

Does the effector have some thickness ?

Does the effector have some thickness ?

Yes, it's a scaled cube. It doesn't have "Surface thickness" but that's not required. Indeed, it works just fine if I manually delete the cache and restart

image.png

But Blender won't invalidate the cache on its own. The problem is with the cache, not with my scene settings.

Yes, it's a scaled cube. It doesn't have "Surface thickness" but that's not required. Indeed, it works just fine if I manually delete the cache and restart ![image.png](https://archive.blender.org/developer/F8494296/image.png) But Blender won't invalidate the cache on its own. The problem is with the cache, not with my scene settings.

Changed status from 'Confirmed' to: 'Resolved'

Changed status from 'Confirmed' to: 'Resolved'
Nicola De Mitri self-assigned this 2020-04-25 14:33:52 +02:00

It's been pointed out, in the chat, that the shift in focus of this Task may be confusing.

Since the issue that gives the title to this Task is now solved, this task can be closed as Resolved.

I've opened another Task (#76095) for the bugs that emerged after the original bug.

It's been pointed out, in the chat, that the shift in focus of this Task may be confusing. Since the issue that gives the title to this Task is now solved, this task can be closed as Resolved. I've opened another Task (#76095) for the bugs that emerged after the original bug.
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#75850
No description provided.