Filmic default view isn't working correctly #51896

Closed
opened 2017-06-26 03:37:57 +02:00 by Greg · 13 comments

System Information
Windows 7 x 64, Titan X (display+render) + 780ti (render)

Blender Version
Broken: 2.78.5 e0bc5b5 (2.79 test build)
Worked: 2.78c with git hub Filmic color management

Short description of error
When the color management Display device is set to sRGB, the view to filmic and the look to none (basic filmic log transfer curve without look shaper), the result is a filmic log transfer curve with the base contrast look active.

2.78c behavior:
https://i.stack.imgur.com/WvSJa.jpg
2.78.5 test build:
https://i.stack.imgur.com/pefmg.jpg

Exact steps for others to reproduce the error
Open Blender, use default scene, switch to Cycles render (or not), hit F12.
In color manager change view to Filmic, look at the render when you change look from none to base contrast, there is no change at all.
It means that the default Filmic view have a base contrast, it's not just the filmic log transfer curve.

Why it is important
It's important to have a defaut filmic view without look shaper, this way, when we save an image in 16bits tiff with filmic, we are still able to change back to linear space, by importing the image and set the image color space to filmic log in the N panel.
It means you could encode decent amount of data in a 16bits image, you couldn't do this with an image saved in sRGB EOTF view.

I noticed something too, the defaut view without any information on the transformation (sRGB EOTF) doesn't mean a lot. In the not so friendly color management topic it's better to have a minimum of informations to avoid confusion.

**System Information** Windows 7 x 64, Titan X (display+render) + 780ti (render) **Blender Version** Broken: 2.78.5 e0bc5b5 (2.79 test build) Worked: 2.78c with git hub Filmic color management **Short description of error** When the color management Display device is set to sRGB, the view to filmic and the look to none (basic filmic log transfer curve without look shaper), the result is a filmic log transfer curve with the base contrast look active. 2.78c behavior: https://i.stack.imgur.com/WvSJa.jpg 2.78.5 test build: https://i.stack.imgur.com/pefmg.jpg **Exact steps for others to reproduce the error** Open Blender, use default scene, switch to Cycles render (or not), hit F12. In color manager change view to Filmic, look at the render when you change look from none to base contrast, there is no change at all. It means that the default Filmic view have a base contrast, it's not just the filmic log transfer curve. **Why it is important** It's important to have a defaut filmic view without look shaper, this way, when we save an image in 16bits tiff with filmic, we are still able to change back to linear space, by importing the image and set the image color space to filmic log in the N panel. It means you could encode decent amount of data in a 16bits image, you couldn't do this with an image saved in sRGB EOTF view. I noticed something too, the defaut view without any information on the transformation (sRGB EOTF) doesn't mean a lot. In the not so friendly color management topic it's better to have a minimum of informations to avoid confusion.
Author

Changed status to: 'Open'

Changed status to: 'Open'
Author

Added subscriber: @mareck

Added subscriber: @mareck

Added subscriber: @brecht

Added subscriber: @brecht

Changed status from 'Open' to: 'Archived'

Changed status from 'Open' to: 'Archived'
Brecht Van Lommel self-assigned this 2017-06-26 04:03:05 +02:00

These changes compared to the original filmic config are intentional.

  • No look indeed corresponds to the Base Contrast, one fewer setup step learn or forget. But also, in OpenColorIO design looks are meant to modify the image in a creative manner, not affect the output color space.
  • For digital intermediates, 16 bit half float EXR is superior to TIFF, so I suggest to use that. If there is a good reason to support writing files in filmic log, it should be implemented in another way.
  • We can't rename "Default" or "sRGB" to "sRGB EOTF" in 2.79 for backwards/forwards compatibility reasons. Further, in the filmic config it is used to name a color space, while it actually is a transformation between color spaces.
These changes compared to the original filmic config are intentional. * No look indeed corresponds to the Base Contrast, one fewer setup step learn or forget. But also, in OpenColorIO design looks are meant to modify the image in a creative manner, not affect the output color space. * For digital intermediates, 16 bit half float EXR is superior to TIFF, so I suggest to use that. If there is a good reason to support writing files in filmic log, it should be implemented in another way. * We can't rename "Default" or "sRGB" to "sRGB EOTF" in 2.79 for backwards/forwards compatibility reasons. Further, in the filmic config it is used to name a color space, while it actually is a transformation between color spaces.
Author

I'll not tell you how to make Blender, I haven't got the knowlege for that.
If it's intentional it's not a bug, but it doesn't change the fact that it's confusing and incorrect.
For me it's not a big deal, I can go 3rd before 2nd if I know it's this way, in this case even making without 2nd.
Being good in the colormanagement topic can profit Blender to be looked like a great tool in the CG industrie.

But also, in OpenColorIO design looks are meant to modify the image in a creative manner, not affect the output color space.

Sorry to bother again but a color space is defined by 3 things, the 3 primaries color, the white point, and the transfer curve.
So if the transfer curve is modified in fact, it is an other color space.

EXR is better, I can't tell if people making animation would profite of having a reasonable weight image that can be reconvert back to linear space.

For the last point, perhaps it can be one or two words in the tool tip, such a hard topic because of years of misunderstanding.

Thanks for your time and for making Blender better day after day.

I'll not tell you how to make Blender, I haven't got the knowlege for that. If it's intentional it's not a bug, but it doesn't change the fact that it's confusing and incorrect. For me it's not a big deal, I can go 3rd before 2nd if I know it's this way, in this case even making without 2nd. Being good in the colormanagement topic can profit Blender to be looked like a great tool in the CG industrie. > But also, in OpenColorIO design looks are meant to modify the image in a creative manner, not affect the output color space. Sorry to bother again but a color space is defined by 3 things, the 3 primaries color, the white point, and the transfer curve. So if the transfer curve is modified in fact, it is an other color space. EXR is better, I can't tell if people making animation would profite of having a reasonable weight image that can be reconvert back to linear space. For the last point, perhaps it can be one or two words in the tool tip, such a hard topic because of years of misunderstanding. Thanks for your time and for making Blender better day after day.

In #51896#442530, @mareck wrote:
If it's intentional it's not a bug, but it doesn't change the fact that it's confusing and incorrect.
For me it's not a big deal, I can go 3rd before 2nd if I know it's this way, in this case even making without 2nd.
Being good in the colormanagement topic can profit Blender to be looked like a great tool in the CG industrie.

Right, but these changes mostly keep the configuration more in line with CG industry standards in my opinion. I'm not aware of anything incorrect in the naming.

Sorry to bother again but a color space is defined by 3 things, the 3 primaries color, the white point, and the transfer curve.
So if the transfer curve is modified in fact, it is an other color space.

We can manipulate renders in the compositor with brightness/contrast, RGB curves or other nodes, but we don't consider the output to be in a different color space because of that. The same goes for looks, the resulting render is still displayed on a monitor that expects the same color space, or written to a file as having the same color space.

EXR is better, I can't tell if people making animation would profite of having a reasonable weight image that can be reconvert back to linear space.

Half float EXRs with DWAA compression are usually 2-3x smaller than 16 bit TIFF in my experience, if file size is a concern.

> In #51896#442530, @mareck wrote: > If it's intentional it's not a bug, but it doesn't change the fact that it's confusing and incorrect. > For me it's not a big deal, I can go 3rd before 2nd if I know it's this way, in this case even making without 2nd. > Being good in the colormanagement topic can profit Blender to be looked like a great tool in the CG industrie. Right, but these changes mostly keep the configuration more in line with CG industry standards in my opinion. I'm not aware of anything incorrect in the naming. > Sorry to bother again but a color space is defined by 3 things, the 3 primaries color, the white point, and the transfer curve. > So if the transfer curve is modified in fact, it is an other color space. We can manipulate renders in the compositor with brightness/contrast, RGB curves or other nodes, but we don't consider the output to be in a different color space because of that. The same goes for looks, the resulting render is still displayed on a monitor that expects the same color space, or written to a file as having the same color space. > EXR is better, I can't tell if people making animation would profite of having a reasonable weight image that can be reconvert back to linear space. Half float EXRs with DWAA compression are usually 2-3x smaller than 16 bit TIFF in my experience, if file size is a concern.

Added subscriber: @troy_s

Added subscriber: @troy_s

No look indeed corresponds to the Base Contrast, one fewer setup step learn or forget. But also, in OpenColorIO design looks are meant to modify the image in a creative manner, not affect the output color space.

Sorry, this isn't not what OCIO's design is based on, as discussed with the folks that both created it and the folks that are currently maintaining it.

So please don't repeat incorrect information here. This choice is entirely Blender. Valid, but not adhering to some incorrect interpretation of the OpenColorIO design. Using the creative looks is precisely appropriate given Filmic's design.

Regarding OP, the proper file format for output / display referred encodes is DPX, but DPX in Blender has been broken forever, so there are no real options to encoding 10, 12, etc. bit depth files in Blender. Half float EXRs, as terrific as they are, are sub-optimal in a number of display referred pipelines.

> No look indeed corresponds to the Base Contrast, one fewer setup step learn or forget. But also, in OpenColorIO design looks are meant to modify the image in a creative manner, not affect the output color space. Sorry, this isn't not what OCIO's design is based on, as discussed with the folks that both created it and the folks that are currently maintaining it. So please don't repeat incorrect information here. This choice is entirely Blender. Valid, but not adhering to some incorrect interpretation of the OpenColorIO design. Using the creative looks is precisely appropriate given Filmic's design. Regarding OP, the proper file format for output / display referred encodes is DPX, but DPX in Blender has been broken forever, so there are no real options to encoding 10, 12, etc. bit depth files in Blender. Half float EXRs, as terrific as they are, are sub-optimal in a number of display referred pipelines.

Can you point me to the docs or emails explaining that looks are indeed intended to affect the output color space?

Can you point me to the docs or emails explaining that looks are indeed intended to affect the output color space?

Unless your point is that the None and Base Contrast in the original config result in an image in the same color space, which I see differently, but I can see it as a valid interpretation. In this particular bug report though it seems the None look is used as a compression scheme for saving digital intermediates as a 16 bit TIFF using log encoding.

If that is what the user wants to do, then I think the better solution to that would be to have a color space setting in the file Output panel where you can select Filmic Log, instead of coupling it to the display settings.

Unless your point is that the None and Base Contrast in the original config result in an image in the same color space, which I see differently, but I can see it as a valid interpretation. In this particular bug report though it seems the None look is used as a compression scheme for saving digital intermediates as a 16 bit TIFF using log encoding. If that is what the user wants to do, then I think the better solution to that would be to have a color space setting in the file Output panel where you can select Filmic Log, instead of coupling it to the display settings.

Added subscriber: @RomboutVersluijs

Added subscriber: @RomboutVersluijs

Im not sure i followed along properly. BUt when saving filmic look in EXR it does not show or save the FILMIC look into it. I tried both Full and Half Float as well all compression methods. It does get save to 8/16 TIFF though?

Reading this post, i believe EXR should save the FILMIC look in EXR right? I also tried "Save Render" but that doesnt help neither.

Im not sure i followed along properly. BUt when saving filmic look in EXR it does not show or save the FILMIC look into it. I tried both Full and Half Float as well all compression methods. It does get save to 8/16 TIFF though? Reading this post, i believe EXR should save the FILMIC look in EXR right? I also tried "Save Render" but that doesnt help neither.
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#51896
No description provided.