This is a complete refactor over the old system. The goal was to increase quality
first and then have something more flexible and optimised.
|{F9603145} | {F9603142}|{F9603147}|
This fixes issues we had with the old system which were:
- Too much overdraw (low performance).
- Not enough precision in render targets (hugly color banding/drifting).
- Poor resolution near in-focus regions.
- Wrong support of orthographic views.
- Missing alpha support in viewport.
- Missing bokeh shape inversion on foreground field.
- Issues on some GPUs. (see T72489) (But I'm sure this one will have other issues as well heh...)
- Fix T81092
I chose Unreal's Diaphragm DOF as a reference / goal implementation.
It is well described in the presentation "A Life of a Bokeh" by Guillaume Abadie.
You can check about it here https://epicgames.ent.box.com/s/s86j70iamxvsuu6j35pilypficznec04
Along side the main implementation we provide a way to increase the quality by jittering the
camera position for each sample (the ones specified under the Sampling tab).
The jittering is dividing the actual post processing dof radius so that it fills the undersampling.
The user can still add more overblur to have a noiseless image, but reducing bokeh shape sharpness.
Effect of overblur (left without, right with):
| {F9603122} | {F9603123}|
The actual implementation differs a bit:
- Foreground gather implementation uses the same "ring binning" accumulator as background
but uses a custom occlusion method. This gives the problem of inflating the foreground elements
when they are over background or in-focus regions.
This is was a hard decision but this was preferable to the other method that was giving poor
opacity masks for foreground and had other more noticeable issues. Do note it is possible
to improve this part in the future if a better alternative is found.
- Use occlusion texture for foreground. Presentation says it wasn't really needed for them.
- The TAA stabilisation pass is replace by a simple neighborhood clamping at the reduce copy
stage for simplicity.
- We don't do a brute-force in-focus separate gather pass. Instead we just do the brute force
pass during resolve. Using the separate pass could be a future optimization if needed but
might give less precise results.
- We don't use compute shaders at all so shader branching might not be optimal. But performance
is still way better than our previous implementation.
- We mainly rely on density change to fix all undersampling issues even for foreground (which
is something the reference implementation is not doing strangely).
Remaining issues (not considered blocking for me):
- Slight defocus stability: Due to slight defocus bruteforce gather using the bare scene color,
highlights are dilated and make convergence quite slow or imposible when using jittered DOF
(or gives )
- ~~Slight defocus inflating: There seems to be a 1px inflation discontinuity of the slight focus
convolution compared to the half resolution. This is not really noticeable if using jittered
camera.~~ Fixed
- Foreground occlusion approximation is a bit glitchy and gives incorrect result if the
a defocus foreground element overlaps a farther foreground element. Note that this is easily
mitigated using the jittered camera position.
|{F9603114}|{F9603115}|{F9603116}|
- Foreground is inflating, not revealing background. However this avoids some other bugs too
as discussed previously. Also mitigated with jittered camera position.
|{F9603130}|{F9603129}|
- Sensor vertical fit is still broken (does not match cycles).
- Scattred bokeh shapes can be a bit strange at polygon vertices. This is due to the distance field
stored in the Bokeh LUT which is not rounded at the edges. This is barely noticeable if the
shape does not rotate.
- ~~Sampling pattern of the jittered camera position is suboptimal. Could try something like hammersley
or poisson disc distribution.~~Used hexaweb sampling pattern which is not random but has better
stability and overall coverage.
- Very large bokeh (> 300 px) can exhibit undersampling artifact in gather pass and quite a bit of
bleeding. But at this size it is preferable to use jittered camera position.
Codewise the changes are pretty much self contained and each pass are well documented.
However the whole pipeline is quite complex to understand from bird's-eye view.
Notes:
- There is the possibility of using arbitrary bokeh texture with this implementation.
However implementation is a bit involved.
- Gathering max sample count is hardcoded to avoid to deal with shader variations. The actual
max sample count is already quite high but samples are not evenly distributed due to the
ring binning method.
- While this implementation does not need 32bit/channel textures to render correctly it does use
many other textures so actual VRAM usage is higher than previous method for viewport but less
for render. Textures are reused to avoid many allocations.
- Bokeh LUT computation is fast and done for each redraw because it can be animated. Also the
texture can be shared with other viewport with different camera settings.
This commit adds a drop-down to the attribute randomize node to support
a few operations on the values of existing attributes: "Replace/Create"
(the existing behavior), "Add", "Subtract", and "Multiply".
At this point, the operations are limited by what is simple to implement.
More could be added in the future, but there isn't a strong use case
for more complex operations anyway, and a second math node can be used.
Differential Revision: https://developer.blender.org/D10269
This patch will share the AOV settings between Cycles and Eevee.
It enable using the AOV name conflict detection of Blender. This
means that unlike how Cycles used to work it isn't possible to add an
AOV with a similar name. Conflicts with internal render pass names will
be indicated with an Warning icon.
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D9774
This updates the join node to use the new multi-input socket from D10067.
The change just requires slightly changing the arguments of the exec
function, and changing the socket input list.
Note that this commit does not contain the UI changes, it only allows for
more input links, and combines the two sockets into one. The UI changes
will come next.
Reviewed By: Hans Goudey
Differential Revision: https://developer.blender.org/D10069
The render visibility column in the outliner should be enabled by
default. This change makes it more obvious which objects will be
rendered which can be misleading when only viewport visibility toggles
are shown by default. This commit enables the render visibility column
in all existing and new Outliner editors.
Differential Revision: https://developer.blender.org/D10365
This patch will show textures in the image editor with the maximum
available resolution determined by the GPU Hardware/Driver.
Currently the size is limited by the user preference texture size limit.
An image user can set the `IMA_SHOW_MAX_RESOLUTION` flag to request
gpu textures in the max supported resolution. When this flag isn't
set the gpu texture is limited by the user preference setting.
When the gl resolution limit is disabled the GPU texture is always
created for the max supported resolution.
Reviewed By: Clément Foucault
Maniphest Tasks: T81206
Differential Revision: https://developer.blender.org/D9160
This patch will show textures in the image editor with the maximum
available resolution determined by the GPU Hardware/Driver.
Currently the size is limited by the user preference texture size limit.
An image user can set the `IMA_SHOW_MAX_RESOLUTION` flag to request
gpu textures in the max supported resolution. When this flag isn't
set the gpu texture is limited by the user preference setting.
When the gl resolution limit is disabled the GPU texture is always
created for the max supported resolution.
Reviewed By: Clément Foucault
Maniphest Tasks: T81206
Differential Revision: https://developer.blender.org/D9160
This adds the ability to use all the math operations in the regular
utility "Math" node. The code is quite similar to the attribute vector
math node, with the simplification that the result is always a float.
Differential Revision: https://developer.blender.org/D10199
The versioning code introduced in rB38df935c0985 skips the composite node
tree that is used by the scene thus not correctly versioning Set Alpha
nodes outside of node groups. This fix iterates through all node trees to
version all Set Alpha nodes.
Reviewed By: JacquesLucke
Differential Revision: https://developer.blender.org/D10110
We will need to expand this node soon to add weight/count for different
elements inside the collection. For that it is better to have the node
to use its own DNA.
By design the modified object transformations should still work and
affect the geometry nodes results. The current behaviour, however, would
make the geometry from the object info to not be affected by the
modified object transformations. This patch changes that by default.
In a similar fashion the Location, Rotation and Scale sockets outputs
should be aware of whether the output should be in the global space or in the
space of the nodetree.
To solve this, the patch introduces a new transformation space "enum"
where users can pick "Original" or "Relative" space.
Original
--------
Output the geometry relative to the input object transform, and the
location, rotation and scale relative to the world origin.
Relative
--------
Bring the input object geometry, location, rotation and scale into the
modified object maintaining the relative position between the two objects in
the scene.
Relative space violates a bit the design of the nodetree. The geometry in this
case is transformed so that moving the modified object doesn't interfere with
the geometry. This is particularly useful for the boolean node for instance.
"Original" is the default space, but old files are set to "Relative" for
backwards compatibility.
Differential Revision: https://developer.blender.org/D10124
When using grease pencil for drawing Storyboards, it's very common to require a transform of the layers. This transform can be done using the offset modifier, but in some cases, the scene requires a lot of modifiers and makes the file hard to work.
This new feature adds a transforms Location, Rotation and Scale at Layer level, and allows to transform the layer without using a modifier, keeping the scene more clean.
{F9480695}
This feature was suggested by @pepeland after receiving feedback from several artists.
Also, done some code cleanup and rename some functions to get a better naming.
Maniphest Tasks: T83660
Differential Revision: https://developer.blender.org/D9761
This commit adds the "Save as Render" feature to the file output node,
that allows to disable the application of the configured view transform
and other color management settings on the image. If disable it uses
the standard view transform instead. This feature was originally
suggested in T83842 and is part of the color management improvements
task in T68926. With this feature it is possible to toggle the
application of the color management settings for each input socket
of the File Output node individually.
Reviewed By: brecht
Maniphest Tasks: T68926, T83842
Differential Revision: https://developer.blender.org/D9921
Adjust default cache settings for all files to store raw and final
images.
All settings are still available when developer extras option is
enabled in user preferences.
This is part of design task T80278
Differential Revision: https://developer.blender.org/D9745
{D9211} introduced pre-multiplying the color for the keying node. This
pre-multiplication should also be done by other keying nodes and should be
the default operation for alpha node.
This patch will change the logic of keying nodes (Cryptomatte Node,
Channel Matte, Chroma Matte, Color Matte, Difference Matte, Distance
Matte, Luminance Matte) and breaks old files.
The Set alpha node has a mode parameter. This parameter changes
the logic to `Apply Mask` the alpha on the RGBA channels of the input color
or only replace the alpha channel (old behavior).
The replace mode is automatically set for older files. When adding
new files the the multiply mode is set.
Reviewed By: Sergey Sharybin
Differential Revision: https://developer.blender.org/D9630
Mainly updated the Mantaflow version. It includes the new viscosity solver plugin based on the method from 'Accurate Viscous Free Surfaces for Buckling, Coiling, and Rotating Liquids' (Batty & Bridson).
In the UI, this update adds a new 'Viscosity' section to the fluid modifier UI (liquid domains only). For now, there is a single 'strength' value to control the viscosity of liquids.
`seq_convert_transform_animation()` converted only keyframed value, but
when bezier interpolation is used, posotion of handles was unchanged.
This caused significant difference in animation.
I have checked only linear interpolation when testing versioning originally.
This adds a popover to the properties editor. Currently the only setting
is for controlling outliner to properties syncing.
Because we cannot define a perfect heuristic to determine when
properties editors should change tabs based on outliner icon selection,
we need an option to enable or disable this behavior per properties
editor.
There are 3 options for controlling the syncing. Auto uses the heuristic
to only allow tab switching when a properties editor and outliner share
a border. On and off enable and disable syncing respectively.
Differential Revision: https://developer.blender.org/D9758
Code was removed in 247b10e6a2 but it was incorrect in first place.
Conversion was done for `offset_x` and `offset_y` channel originally,
but it should be done for `scale_x` and `scale_y`
Since creating the attribute node, a helper function has been added to
automatically get the input attribute or a constant value, depending on
the "input type" values for the node. This commit replaces the specific
implementation of that behavior with the new helper function.
The versioning is necessary since the node now has a "storage" struct.
This commit resolves problem introduced in e1665c3d31 - it was
difficult to import media at their original resolution.
This is done by using original resolution as reference for scale.
All crop and strip transform values and their animation is converted
form old files.
To make both workflows easy to use, sequencer tool settings have been
created with preset for preffered scaling method. This setting is in
sequencer timeline header and add image or movie strip operator
properties.
Two new operators have been added:
`sequencer.strip_transform_fit` operator with 3 options: Scale To Fit,
Scale to Fill and Stretch To Fill.
Operator can fail if strip image or video is not loaded currently, this
case should be either sanitized or data loaded on demand.
`sequencer.strip_transform_clear` operator with 4 options:
Clear position, scale, rotation and all (previous 3 options combined).
Reviewed By: sergey, fsiddi
Differential Revision: https://developer.blender.org/D9582
Add panels with overlay settings for strips and preview and overlay
enable/disable button.
Entries from the View menus moved to the overlay panels, which will
simplify cluttered View menus.
Additional options have been added:
- Strip Name
- Strip Source(ex. path)
- Strip Duration
So users can now select what info they need to see on the strips. When
No text is displayed, waveforms are drawn in full height.
Reviewed By: ISS, HooglyBoogly, pablovazquez
Differential Revision: https://developer.blender.org/D9751
Remove DNA headers, using forward declarations where possible.
Also removed duplicate header, header including it's self
and unnecessary inclusion of libc system headers from BKE header.
This changes the way how the mattes are stored in the compositor node. This used to
be a single string what was decoded/encoded when needed. The new data structure
stores all entries in `CryptomatteEntry` and is converted to the old `matte_id`
property on the fly.
This is done for some future changes in the workflow where a more structured
approach leads to less confusing and easier to read code.
In Cycles the volume transmittance is already composited into the color
passes. In Eevee the volume transmittance pass was separate and needed
to be composited in the compositor. This patch adds the volume
transmittance pass direct in the next render passes:
* Diffuse Color
* Specular Color
* Emission
* Environment
This patch includes the removal of the volume transmittance render pass.
It also renames the volume render passes to match Cycles. The setting
themselves aren't unified.
Maniphest Tasks: T81134
