2018-07-18 13:03:09 +02:00
|
|
|
/*
|
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
|
* as published by the Free Software Foundation; either version 2
|
|
|
|
|
* of the License, or (at your option) any later version.
|
|
|
|
|
*
|
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
|
*
|
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
|
* along with this program; if not, write to the Free Software Foundation,
|
|
|
|
|
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
|
2019-02-18 07:21:50 +11:00
|
|
|
*
|
|
|
|
|
* Copyright 2018, Blender Foundation.
|
2018-07-18 13:03:09 +02:00
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#include "COM_CryptomatteNode.h"
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
#include "BKE_node.h"
|
2020-12-14 16:14:21 +01:00
|
|
|
#include "BLI_assert.h"
|
|
|
|
|
#include "BLI_hash_mm3.h"
|
|
|
|
|
#include "BLI_listbase.h"
|
|
|
|
|
#include "BLI_string.h"
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
#include "COM_ConvertOperation.h"
|
|
|
|
|
#include "COM_CryptomatteOperation.h"
|
|
|
|
|
#include "COM_MultilayerImageOperation.h"
|
|
|
|
|
#include "COM_RenderLayersProg.h"
|
2021-01-05 16:16:18 +01:00
|
|
|
#include "COM_SetAlphaMultiplyOperation.h"
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
#include "COM_SetColorOperation.h"
|
2018-07-18 13:03:09 +02:00
|
|
|
#include <iterator>
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
#include <string>
|
|
|
|
|
|
2021-03-23 17:12:27 +01:00
|
|
|
namespace blender::compositor {
|
|
|
|
|
|
2021-09-15 10:50:33 +10:00
|
|
|
/* -------------------------------------------------------------------- */
|
|
|
|
|
/** \name Cryptomatte Base
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
* \{ */
|
2018-07-18 13:03:09 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
void CryptomatteBaseNode::convertToOperations(NodeConverter &converter,
|
|
|
|
|
const CompositorContext &context) const
|
2018-07-18 13:03:09 +02:00
|
|
|
{
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
NodeOutput *output_image_socket = this->getOutputSocket(0);
|
|
|
|
|
|
|
|
|
|
bNode *node = this->getbNode();
|
|
|
|
|
NodeCryptomatte *cryptomatte_settings = static_cast<NodeCryptomatte *>(node->storage);
|
|
|
|
|
|
|
|
|
|
CryptomatteOperation *cryptomatte_operation = create_cryptomatte_operation(
|
|
|
|
|
converter, context, *node, cryptomatte_settings);
|
|
|
|
|
converter.addOperation(cryptomatte_operation);
|
|
|
|
|
|
|
|
|
|
NodeOutput *output_matte_socket = this->getOutputSocket(1);
|
|
|
|
|
SeparateChannelOperation *extract_mask_operation = new SeparateChannelOperation;
|
|
|
|
|
extract_mask_operation->setChannel(3);
|
|
|
|
|
converter.addOperation(extract_mask_operation);
|
|
|
|
|
converter.addLink(cryptomatte_operation->getOutputSocket(0),
|
|
|
|
|
extract_mask_operation->getInputSocket(0));
|
|
|
|
|
converter.mapOutputSocket(output_matte_socket, extract_mask_operation->getOutputSocket(0));
|
|
|
|
|
|
|
|
|
|
NodeInput *input_image_socket = this->getInputSocket(0);
|
|
|
|
|
SetAlphaMultiplyOperation *apply_mask_operation = new SetAlphaMultiplyOperation();
|
|
|
|
|
converter.mapInputSocket(input_image_socket, apply_mask_operation->getInputSocket(0));
|
|
|
|
|
converter.addOperation(apply_mask_operation);
|
|
|
|
|
converter.addLink(extract_mask_operation->getOutputSocket(0),
|
|
|
|
|
apply_mask_operation->getInputSocket(1));
|
|
|
|
|
converter.mapOutputSocket(output_image_socket, apply_mask_operation->getOutputSocket(0));
|
|
|
|
|
|
|
|
|
|
NodeOutput *output_pick_socket = this->getOutputSocket(2);
|
|
|
|
|
SetAlphaMultiplyOperation *extract_pick_operation = new SetAlphaMultiplyOperation();
|
|
|
|
|
converter.addOperation(extract_pick_operation);
|
|
|
|
|
converter.addInputValue(extract_pick_operation->getInputSocket(1), 1.0f);
|
|
|
|
|
converter.addLink(cryptomatte_operation->getOutputSocket(0),
|
|
|
|
|
extract_pick_operation->getInputSocket(0));
|
|
|
|
|
converter.mapOutputSocket(output_pick_socket, extract_pick_operation->getOutputSocket(0));
|
2018-07-18 13:03:09 +02:00
|
|
|
}
|
|
|
|
|
|
2021-09-15 10:50:33 +10:00
|
|
|
/** \} */
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
|
2021-09-15 10:50:33 +10:00
|
|
|
/* -------------------------------------------------------------------- */
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
/** \name Cryptomatte V2
|
|
|
|
|
* \{ */
|
2021-09-15 10:50:33 +10:00
|
|
|
|
2021-05-31 14:32:39 +02:00
|
|
|
static std::string prefix_from_node(const CompositorContext &context, const bNode &node)
|
2018-07-18 13:03:09 +02:00
|
|
|
{
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
char prefix[MAX_NAME];
|
2021-05-31 14:32:39 +02:00
|
|
|
ntreeCompositCryptomatteLayerPrefix(context.getScene(), &node, prefix, sizeof(prefix));
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
return std::string(prefix, BLI_strnlen(prefix, sizeof(prefix)));
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
static std::string combined_layer_pass_name(RenderLayer *render_layer, RenderPass *render_pass)
|
|
|
|
|
{
|
|
|
|
|
if (render_layer->name[0] == '\0') {
|
|
|
|
|
return std::string(render_pass->name,
|
|
|
|
|
BLI_strnlen(render_pass->name, sizeof(render_pass->name)));
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
std::string combined_name =
|
|
|
|
|
blender::StringRef(render_layer->name,
|
|
|
|
|
BLI_strnlen(render_layer->name, sizeof(render_layer->name))) +
|
|
|
|
|
"." +
|
|
|
|
|
blender::StringRef(render_pass->name,
|
|
|
|
|
BLI_strnlen(render_pass->name, sizeof(render_pass->name)));
|
|
|
|
|
return combined_name;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void CryptomatteNode::input_operations_from_render_source(
|
|
|
|
|
const CompositorContext &context,
|
|
|
|
|
const bNode &node,
|
2021-04-02 16:13:27 +02:00
|
|
|
Vector<NodeOperation *> &r_input_operations)
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
{
|
|
|
|
|
Scene *scene = (Scene *)node.id;
|
|
|
|
|
if (!scene) {
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
BLI_assert(GS(scene->id.name) == ID_SCE);
|
|
|
|
|
Render *render = RE_GetSceneRender(scene);
|
|
|
|
|
RenderResult *render_result = render ? RE_AcquireResultRead(render) : nullptr;
|
|
|
|
|
|
|
|
|
|
if (!render_result) {
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
2021-06-02 10:05:39 +02:00
|
|
|
short view_layer_id = 0;
|
2021-05-31 14:32:39 +02:00
|
|
|
const std::string prefix = prefix_from_node(context, node);
|
2021-06-02 10:05:39 +02:00
|
|
|
LISTBASE_FOREACH_INDEX (ViewLayer *, view_layer, &scene->view_layers, view_layer_id) {
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
RenderLayer *render_layer = RE_GetRenderLayer(render_result, view_layer->name);
|
|
|
|
|
if (render_layer) {
|
|
|
|
|
LISTBASE_FOREACH (RenderPass *, render_pass, &render_layer->passes) {
|
2021-08-23 15:31:21 +02:00
|
|
|
if (context.has_explicit_view() && !STREQ(render_pass->view, context.getViewName())) {
|
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
const std::string combined_name = combined_layer_pass_name(render_layer, render_pass);
|
|
|
|
|
if (blender::StringRef(combined_name).startswith(prefix)) {
|
|
|
|
|
RenderLayersProg *op = new RenderLayersProg(
|
2021-03-19 14:26:24 +01:00
|
|
|
render_pass->name, DataType::Color, render_pass->channels);
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
op->setScene(scene);
|
2021-06-02 10:05:39 +02:00
|
|
|
op->setLayerId(view_layer_id);
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
op->setRenderData(context.getRenderData());
|
|
|
|
|
op->setViewName(context.getViewName());
|
|
|
|
|
r_input_operations.append(op);
|
|
|
|
|
}
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
}
|
|
|
|
|
}
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
RE_ReleaseResult(render);
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
void CryptomatteNode::input_operations_from_image_source(
|
|
|
|
|
const CompositorContext &context,
|
|
|
|
|
const bNode &node,
|
2021-04-02 16:13:27 +02:00
|
|
|
Vector<NodeOperation *> &r_input_operations)
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
{
|
|
|
|
|
NodeCryptomatte *cryptomatte_settings = (NodeCryptomatte *)node.storage;
|
|
|
|
|
Image *image = (Image *)node.id;
|
|
|
|
|
if (!image) {
|
|
|
|
|
return;
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
BLI_assert(GS(image->id.name) == ID_IM);
|
|
|
|
|
if (image->type != IMA_TYPE_MULTILAYER) {
|
|
|
|
|
return;
|
2018-07-18 13:03:09 +02:00
|
|
|
}
|
2018-07-18 13:34:22 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
ImageUser *iuser = &cryptomatte_settings->iuser;
|
|
|
|
|
BKE_image_user_frame_calc(image, iuser, context.getFramenumber());
|
|
|
|
|
ImBuf *ibuf = BKE_image_acquire_ibuf(image, iuser, nullptr);
|
2018-07-19 16:06:37 +10:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
if (image->rr) {
|
|
|
|
|
int view = 0;
|
|
|
|
|
if (BLI_listbase_count_at_most(&image->rr->views, 2) > 1) {
|
|
|
|
|
if (iuser->view == 0) {
|
|
|
|
|
/* Heuristic to match image name with scene names, check if the view name exists in the
|
|
|
|
|
* image. */
|
|
|
|
|
view = BLI_findstringindex(
|
|
|
|
|
&image->rr->views, context.getViewName(), offsetof(RenderView, name));
|
|
|
|
|
if (view == -1) {
|
|
|
|
|
view = 0;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
else {
|
|
|
|
|
view = iuser->view - 1;
|
|
|
|
|
}
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
2021-05-31 14:32:39 +02:00
|
|
|
const std::string prefix = prefix_from_node(context, node);
|
2021-03-30 11:05:10 +02:00
|
|
|
int layer_index;
|
|
|
|
|
LISTBASE_FOREACH_INDEX (RenderLayer *, render_layer, &image->rr->layers, layer_index) {
|
|
|
|
|
if (!blender::StringRef(prefix).startswith(blender::StringRef(
|
|
|
|
|
render_layer->name, BLI_strnlen(render_layer->name, sizeof(render_layer->name))))) {
|
|
|
|
|
continue;
|
|
|
|
|
}
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
LISTBASE_FOREACH (RenderPass *, render_pass, &render_layer->passes) {
|
|
|
|
|
const std::string combined_name = combined_layer_pass_name(render_layer, render_pass);
|
|
|
|
|
if (blender::StringRef(combined_name).startswith(prefix)) {
|
|
|
|
|
MultilayerColorOperation *op = new MultilayerColorOperation(
|
|
|
|
|
render_layer, render_pass, view);
|
|
|
|
|
op->setImage(image);
|
|
|
|
|
op->setImageUser(iuser);
|
2021-03-30 11:05:10 +02:00
|
|
|
iuser->layer = layer_index;
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
op->setFramenumber(context.getFramenumber());
|
|
|
|
|
r_input_operations.append(op);
|
|
|
|
|
}
|
|
|
|
|
}
|
2021-03-30 11:05:10 +02:00
|
|
|
break;
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
BKE_image_release_ibuf(image, ibuf, nullptr);
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
2021-04-02 16:13:27 +02:00
|
|
|
Vector<NodeOperation *> CryptomatteNode::create_input_operations(const CompositorContext &context,
|
|
|
|
|
const bNode &node)
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
{
|
2021-04-02 16:13:27 +02:00
|
|
|
Vector<NodeOperation *> input_operations;
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
switch (node.custom1) {
|
|
|
|
|
case CMP_CRYPTOMATTE_SRC_RENDER:
|
|
|
|
|
input_operations_from_render_source(context, node, input_operations);
|
|
|
|
|
break;
|
|
|
|
|
case CMP_CRYPTOMATTE_SRC_IMAGE:
|
|
|
|
|
input_operations_from_image_source(context, node, input_operations);
|
|
|
|
|
break;
|
|
|
|
|
}
|
2018-07-18 13:03:09 +02:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
if (input_operations.is_empty()) {
|
|
|
|
|
SetColorOperation *op = new SetColorOperation();
|
|
|
|
|
op->setChannel1(0.0f);
|
|
|
|
|
op->setChannel2(1.0f);
|
|
|
|
|
op->setChannel3(0.0f);
|
|
|
|
|
op->setChannel4(0.0f);
|
|
|
|
|
input_operations.append(op);
|
|
|
|
|
}
|
|
|
|
|
return input_operations;
|
|
|
|
|
}
|
|
|
|
|
CryptomatteOperation *CryptomatteNode::create_cryptomatte_operation(
|
|
|
|
|
NodeConverter &converter,
|
|
|
|
|
const CompositorContext &context,
|
|
|
|
|
const bNode &node,
|
|
|
|
|
const NodeCryptomatte *cryptomatte_settings) const
|
|
|
|
|
{
|
2021-04-02 16:13:27 +02:00
|
|
|
Vector<NodeOperation *> input_operations = create_input_operations(context, node);
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
CryptomatteOperation *operation = new CryptomatteOperation(input_operations.size());
|
|
|
|
|
LISTBASE_FOREACH (CryptomatteEntry *, cryptomatte_entry, &cryptomatte_settings->entries) {
|
|
|
|
|
operation->addObjectIndex(cryptomatte_entry->encoded_hash);
|
|
|
|
|
}
|
|
|
|
|
for (int i = 0; i < input_operations.size(); ++i) {
|
|
|
|
|
converter.addOperation(input_operations[i]);
|
|
|
|
|
converter.addLink(input_operations[i]->getOutputSocket(), operation->getInputSocket(i));
|
|
|
|
|
}
|
|
|
|
|
return operation;
|
|
|
|
|
}
|
|
|
|
|
|
2021-09-15 10:50:33 +10:00
|
|
|
/** \} */
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
|
2021-09-15 10:50:33 +10:00
|
|
|
/* -------------------------------------------------------------------- */
|
|
|
|
|
/** \name Cryptomatte Legacy
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
* \{ */
|
2018-07-19 16:06:37 +10:00
|
|
|
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
CryptomatteOperation *CryptomatteLegacyNode::create_cryptomatte_operation(
|
|
|
|
|
NodeConverter &converter,
|
|
|
|
|
const CompositorContext &UNUSED(context),
|
|
|
|
|
const bNode &UNUSED(node),
|
|
|
|
|
const NodeCryptomatte *cryptomatte_settings) const
|
|
|
|
|
{
|
2021-03-30 12:27:53 +02:00
|
|
|
const int num_inputs = inputs.size() - 1;
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
CryptomatteOperation *operation = new CryptomatteOperation(num_inputs);
|
|
|
|
|
if (cryptomatte_settings) {
|
|
|
|
|
LISTBASE_FOREACH (CryptomatteEntry *, cryptomatte_entry, &cryptomatte_settings->entries) {
|
|
|
|
|
operation->addObjectIndex(cryptomatte_entry->encoded_hash);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
for (int i = 0; i < num_inputs; i++) {
|
|
|
|
|
converter.mapInputSocket(this->getInputSocket(i + 1), operation->getInputSocket(i));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return operation;
|
2018-07-18 13:03:09 +02:00
|
|
|
}
|
Compositor: Redesign Cryptomatte node for better usability
In the current implementation, cryptomatte passes are connected to the node
and elements are picked by using the eyedropper tool on a special pick channel.
This design has two disadvantages - both connecting all passes individually
and always having to switch to the picker channel are tedious.
With the new design, the user selects the RenderLayer or Image from which the
Cryptomatte layers are directly loaded (the type of pass is determined by an
enum). This allows the node to automatically detect all relevant passes.
Then, when using the eyedropper tool, the operator looks up the selected
coordinates from the picked Image, Node backdrop or Clip and reads the picked
object directly from the Renderlayer/Image, therefore allowing to pick in any
context (e.g. by clicking on the Combined pass in the Image Viewer). The
sampled color is looked up in the metadata and the actual name is stored
in the cryptomatte node. This also allows to remove a hash by just removing
the name from the matte id.
Technically there is some loss of flexibility because the Cryptomatte pass
inputs can no longer be connected to other nodes, but since any compositing
done on them is likely to break the Cryptomatte system anyways, this isn't
really a concern in practise.
In the future, this would also allow to automatically translate values to names
by looking up the value in the associated metadata of the input, or to get a
better visualization of overlapping areas in the Pick output since we could
blend colors now that the output doesn't have to contain the exact value.
Idea + Original patch: Lucas Stockner
Reviewed By: Brecht van Lommel
Differential Revision: https://developer.blender.org/D3959
2021-03-16 07:37:30 +01:00
|
|
|
|
2021-09-15 10:50:33 +10:00
|
|
|
/** \} */
|
2021-03-23 17:12:27 +01:00
|
|
|
|
|
|
|
|
} // namespace blender::compositor
|