lightfield_viewport.py

# ##### BEGIN GPL LICENSE BLOCK #####
#
#  Copyright © 2021 Christian Stolze
#
#  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 3 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, see <http://www.gnu.org/licenses/>.
#
# ##### END GPL LICENSE BLOCK #####

# MODULE DESCRIPTION:
# ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
# This includes everything that is related the "live view"

# ------------------ INTERNAL MODULES --------------------
from .globals import *

# ------------------- EXTERNAL MODULES -------------------
import sys, platform
import bpy, bgl
import gpu
import time, timeit
from math import *
from mathutils import *
from gpu_extras.batch import batch_for_shader
from gpu_extras.presets import draw_texture_2d, draw_circle_2d
from bpy_extras.view3d_utils import location_3d_to_region_2d, region_2d_to_origin_3d, region_2d_to_vector_3d
import numpy as np

# append the add-on's path to Blender's python PATH
sys.path.insert(0, LookingGlassAddon.path)
sys.path.insert(0, LookingGlassAddon.libpath)

# TODO: Would be better, if from .lib import pylightio could be called,
#		but for some reason that does not import all modules and throws
#		"AliceLG.lib.pylio has no attribute 'lookingglass"
import pylightio as pylio

# ---------------- GLOBAL ADDON LOGGER -------------------
import logging
LookingGlassAddonLogger = logging.getLogger('Alice/LG')




# ------------ CONTEXT OVERRIDE -------------
# Class for managing a SpaceView3D context override for offscreen rendering
class ContextOverride:

	# ADDON SETTING DATA
	__addon_settings_window_manager = None
	__addon_settings_scene = None

	# CONTEXT DATA
	__context = None
	__override = None

	# CONTEXT DATA BACKUP
	__shading_restore_backup = {}
	__overlay_restore_backup = {}

	# Inititalize the context override
	def __init__(self, context):

		# get the current settings of this scene
		self.__context = context

		# get the current settings of the window manager
		self.__addon_settings_window_manager = context.window_manager.addon_settings

		# get the current settings of this scene
		self.__addon_settings_scene = context.scene.addon_settings

		# create an override context from the invoking context
		self.__override = context.copy()


	# set a new context
	def set_context(self, context):

		# get the current settings of this scene
		self.__context = context


	# set up the camera for each view and the shader of the rendering object
	def setupVirtualCameraForView(self, view, total_views, view_cone, aspect, viewMatrix, projectionMatrix):

		# get focal plane distance depending on synchronization mode
		if self.__addon_settings_scene.toggleFocalSync and self.__addon_settings_scene.lookingglassCamera:
			focalPlane = self.__addon_settings_scene.lookingglassCamera.data.dof.focus_distance
		else:
			focalPlane = self.__addon_settings_scene.focalPlane

		# The field of view set by the camera
		# NOTE 1: - the Looking Glass Factory documentation suggests to use a FOV of 14°. We use the focal length of the Blender camera instead.
		# NOTE 2: - we take the angle directly from the projection matrix
		fov = 2.0 * atan(1 / projectionMatrix[1][1])

		# calculate cameraSize from its distance to the focal plane and the FOV
		# NOTE: - we take an arbitrary distance of 5 m (we could also use the focal distance of the camera, but might be confusing)
		cameraDistance = focalPlane
		cameraSize = cameraDistance * tan(fov / 2)

		# start at viewCone * 0.5 and go up to -viewCone * 0.5
		offsetAngle = (0.5 - view / (total_views - 1)) * radians(view_cone)

		# calculate the offset that the camera should move
		offset = cameraDistance * tan(offsetAngle)

		# translate the view matrix (position) by the calculated offset in x-direction
		viewMatrix = Matrix.Translation((offset, 0, 0)) @ viewMatrix

		# modify the projection matrix, relative to the camera size and aspect ratio
		projectionMatrix[0][2] += offset / (cameraSize * aspect)

		# return the projection matrix
		return viewMatrix, projectionMatrix



	# Save the viewport settings
	def saveViewportSettings(self):

		# SHADING ATTRIBUTES
		# define some exceptions that must not be taken into
		attributeExceptions = ["__doc__", "__module__", "__slots__", "bl_rna", "rna_type", "color_type", "studio_light"]

		# use the "space data" of the selected viewport
		attributeList = dir(self.__override['space_data'].shading)
		for attr in attributeList:

			if not attr in attributeExceptions and hasattr(self.__override['space_data'].shading, attr):
				#print("[SHADING]", attr, " = ", getattr(LookingGlassAddon.BlenderViewport.shading, attr))

				try:
					self.__shading_restore_backup[attr] = getattr(self.__override['space_data'].shading, attr)
				except Exception as e:
					#print(" # ", e)
					pass

		attributeList = dir(self.__override['space_data'].overlay)
		for attr in attributeList:

			if not attr in attributeExceptions and hasattr(self.__override['space_data'].overlay, attr):
				#print("[OVERLAY]", attr, " = ", getattr(self.__override['space_data'].overlay, attr))

				try:
					self.__overlay_restore_backup[attr] = getattr(self.__override['space_data'].overlay, attr)
				except Exception as e:
					#print(" # ", e)
					pass


	# Update the viewport settings
	def updateViewportSettings(self, space_data=None, force_context_data=False):

		# get the space data into the override
		if space_data:
			self.__override['space_data'] = space_data

		# save all shading & overlay settings
		self.saveViewportSettings()


		# APPLY CUSTOM SETTINGS IF REQUIRED
		####################################################################

		# if the custom settings shall be used OR the given space data is invalid
		if (self.__addon_settings_scene.viewportMode == 'CUSTOM' and force_context_data == False) or space_data == None:

			# SHADING ATTRIBUTES
			self.__override['space_data'].shading.type = self.__addon_settings_scene.shadingMode
			self.__override['space_data'].shading.show_xray = bool(self.__addon_settings_scene.viewport_show_xray)
			self.__override['space_data'].shading.xray_alpha = float(self.__addon_settings_scene.viewport_xray_alpha)
			self.__override['space_data'].shading.use_dof = bool(int(self.__addon_settings_scene.viewport_use_dof))

			# OVERLAY ATTRIBUTES: Guides
			self.__override['space_data'].overlay.show_floor = bool(int(self.__addon_settings_scene.viewport_show_floor))
			self.__override['space_data'].overlay.show_axis_x = bool(int(self.__addon_settings_scene.viewport_show_axes[0]))
			self.__override['space_data'].overlay.show_axis_y = bool(int(self.__addon_settings_scene.viewport_show_axes[1]))
			self.__override['space_data'].overlay.show_axis_z = bool(int(self.__addon_settings_scene.viewport_show_axes[2]))
			self.__override['space_data'].overlay.grid_scale = float(self.__addon_settings_scene.viewport_grid_scale)
			# OVERLAY ATTRIBUTES: Objects
			self.__override['space_data'].overlay.show_extras = bool(int(self.__addon_settings_scene.viewport_show_extras))
			self.__override['space_data'].overlay.show_relationship_lines = bool(int(self.__addon_settings_scene.viewport_show_relationship_lines))
			self.__override['space_data'].overlay.show_outline_selected = bool(int(self.__addon_settings_scene.viewport_show_outline_selected))
			self.__override['space_data'].overlay.show_bones = bool(int(self.__addon_settings_scene.viewport_show_bones))
			self.__override['space_data'].overlay.show_motion_paths = bool(int(self.__addon_settings_scene.viewport_show_motion_paths))
			self.__override['space_data'].overlay.show_object_origins = bool(int(self.__addon_settings_scene.viewport_show_origins))
			self.__override['space_data'].overlay.show_object_origins_all = bool(int(self.__addon_settings_scene.viewport_show_origins_all))
			# OVERLAY ATTRIBUTES: Geometry
			self.__override['space_data'].overlay.show_wireframes = bool(int(self.__addon_settings_scene.viewport_show_wireframes))
			self.__override['space_data'].overlay.show_face_orientation = bool(int(self.__addon_settings_scene.viewport_show_face_orientation))

		# if the settings rely on a specific viewport / SpaceView3D
		elif (self.__addon_settings_scene.viewportMode != 'CUSTOM' or force_context_data == True) and space_data != None:

			# if CYCLES is activated in the current viewport
			if space_data.shading.type == 'RENDERED' and self.__context.engine == 'CYCLES':

				# change the shading type to SOLID
				self.__override['space_data'].shading.type = 'SOLID'

				# notify user
				self.report({"WARNING"}, "Render engine (%s) not supported in lightfield previews. Switched to SOLID mode." % self.__context.engine)

		# always disable the hdri preview spheres
		self.__override['space_data'].overlay.show_look_dev = False

	# Restore the viewport settings
	def restoreViewportSettings(self):

		# SHADING ATTRIBUTES
		# define some exceptions that must not be taken into
		attributeExceptions = ["__doc__", "__module__", "__slots__", "bl_rna", "rna_type", "color_type", "studio_light", "type"]

		# use the "space data" of the selected viewport
		attributeList = dir(self.__override['space_data'].shading)
		for attr in attributeList:

			if not attr in attributeExceptions and hasattr(self.__override['space_data'].shading, attr):
				if getattr(self.__override['space_data'].shading, attr) != self.__shading_restore_backup[attr]:
				#print("[SHADING]", attr, " = ", self.__shading_restore_backup[attr])

					try:
						setattr(self.__override['space_data'].shading, attr, self.__shading_restore_backup[attr])
					except Exception as e:
						#print(" # ", e)
						pass

		attributeList = dir(self.__override['space_data'].overlay)
		for attr in attributeList:

			if not attr in attributeExceptions and hasattr(self.__override['space_data'].overlay, attr):
				if getattr(self.__override['space_data'].overlay, attr) != self.__overlay_restore_backup[attr]:
				# print("[OVERLAY]", attr, " = ", self.__overlay_restore_backup[attr])

					try:
						setattr(self.__override['space_data'].overlay, attr, self.__overlay_restore_backup[attr])
					except Exception as e:
						#print(" # ", e)
						pass

	# CLASS PROPERTIES
	# ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
	# (A) context override properties
	@property
	def space_data(self):
		return self.__override['space_data']

	@space_data.setter
	def space_data(self, value):
		pass

	@property
	def region(self):
		return self.__override['region']

	@region.setter
	def region(self, value):
		pass

	@property
	def shading_to_dict(self):
		return self.__shading_restore_backup

	@shading_to_dict.setter
	def shading_to_dict(self, value):
		pass

	@property
	def overlay_to_dict(self):
		return self.__overlay_restore_backup

	@overlay_to_dict.setter
	def overlay_to_dict(self, value):
		pass



# ------------ LIGHTFIELD RENDERING -------------
# Modal operator for controlled redrawing of the lightfield window.
class LOOKINGGLASS_OT_render_viewport(bpy.types.Operator):

	bl_idname = "render.viewport"
	bl_label = "Looking Glass Lightfield Viewport rendering"
	bl_options = {'REGISTER', 'INTERNAL'}


	# PUBLIC CLASS MEMBERS
	# ++++++++++++++++++++++++++++++++++++++++++++++++++

	# ADDON SETTINGS
	settings = None

	# SETTINGS VARIABLES
	preset = 1
	last_preset = 1

	# lightfield
	lightfield_image = None

	# DRAWING OPERATION VARIABLES
	modal_redraw = True
	depsgraph_update_time = 0
	skip_views = 1
	restricted_viewcone_limit = 0

	# DEBUGING VARIABLES
	start_multi_view = 0

	# PROTECTED CLASS MEMBERS
	# ++++++++++++++++++++++++++++++++++++++++++++++++++

	# HANDLER IDENTIFIERS
	_handle_trackDepsgraphUpdates = None
	_handle_trackFrameChanges = None
	_handle_trackActiveWindow = None

	# CONTEXT OVERRIDE
	_override = None

	# SETTINGS BACKUP
	_shading_restore_backup = {}
	_overlay_restore_backup = {}

	# METHODS
	# ++++++++++++++++++++++++++++++++++++++++++++++++++
	# poll method
	@classmethod
	def poll(self, context):

		# check if the context is invalid
		if not context:
			LookingGlassAddonLogger.error("Could not open Lightfield Window. Operator was called from invalid context.")
			return False

		# return True, so the operator is executed
		return True



	# cancel the modal operator
	def cancel(self, context):

		# log info
		LookingGlassAddonLogger.info("Closing lightfield viewport ...")

		# stop timer
		context.window_manager.event_timer_remove(self.timerEvent)

		# remove the app handler that checks for depsgraph updates
		bpy.app.handlers.depsgraph_update_post.remove(self.trackDepsgraphUpdates)
		bpy.app.handlers.frame_change_post.remove(self.trackDepsgraphUpdates)

		# remove the handler for the viewport tracking
		if self._handle_trackActiveWindow: bpy.types.SpaceView3D.draw_handler_remove(self._handle_trackActiveWindow, 'WINDOW')

		# log info
		LookingGlassAddonLogger.info(" [#] Cancelled control handlers.")

		# iterate through all presets
		for i, preset in self.qs.items():

			# loop through all required views
			#for view in range(int((self.qs[self.preset]["total_views"] + 1) / 3), self.qs[self.preset]["total_views"] - int((self.qs[self.preset]["total_views"] + 1) / 3)):
			for view in range(0, len(self.qs[i]["viewOffscreen"])):

				# free the GPUOffscreen for the view rendering
				self.qs[i]["viewOffscreen"][view].free()

			# delete the list of offscreen objects
			self.qs[i]["viewOffscreen"].clear()

		# log info
		LookingGlassAddonLogger.info(" [#] Freed GPUOffscreens of the lightfield views.")

		# set status variables to default state
		#LookingGlassAddon.BlenderWindow = None
		LookingGlassAddon.BlenderViewport = None

		# set the button controls for the lightfield window to False
		if context: context.window_manager.addon_settings.ShowLightfieldWindow = False

		# clear the quilt
		self.device.clear()

		# free the view data of the lightfield image
		if self.lightfield_image: self.lightfield_image.clear_views()

		# delete the current LightfieldImage
		if self.lightfield_image: self.lightfield_image = None

		# log info
		LookingGlassAddonLogger.info(" [#] Done.")

	# invoke the operator
	def invoke(self, context, event):
		start = time.time()

		# log info
		LookingGlassAddonLogger.info("Invoking lightfield viewport ...")

		# get the current settings of this scene
		self.addon_settings_window_manager = context.window_manager.addon_settings
		self.addon_settings_scene = context.scene.addon_settings

		# update the variable for the current Looking Glass device
		if int(self.addon_settings_window_manager.activeDisplay) != -1: self.device = pylio.DeviceManager.get_active()



		# PREPARE THE OFFSCREEN RENDERING
		################################################################

		# set to the currently chosen quality
		self.preset = self.last_preset = int(context.scene.addon_settings.quiltPreset)

		# get all quilt presets from pylio
		self.qs = pylio.LookingGlassQuilt.formats.get()

		# iterate through all presets
		for i, preset in self.qs.items():

			# create a list of offscreen objects for this preset
			self.qs[i]["viewOffscreen"] = []

			# loop through all required views
			#for view in range(int((self.qs[self.preset]["total_views"] + 1) / 3), self.qs[self.preset]["total_views"] - int((self.qs[self.preset]["total_views"] + 1) / 3)):
			for view in range(0, self.qs[i]["total_views"]):

				# create a GPUOffscreen for the views
				self.qs[i]["viewOffscreen"].append(gpu.types.GPUOffScreen(int(self.qs[i]["view_width"]), int(self.qs[i]["view_height"])))

		# log info
		LookingGlassAddonLogger.info(" [#] Prepared GPUOffscreens for view rendering.")


		# PREPARE THE OVERRIDE CONTEXT THAT CONTAINS THE RENDER SETTINGS
		################################################################

		# create an override context from the invoking context
		self._override = ContextOverride(context)

		# update the viewport settings
		self.updateViewportSettings(context)

		# log info
		LookingGlassAddonLogger.info(" [#] Created override context.")


		# REGISTER ALL HANDLERS FOR THE LIGHTFIELD RENDERING
		################################################################

		# HANDLERS FOR CONTROL PURPOSES
		# ++++++++++++++++++++++++++++++
		# we exploit the draw_hanlder of the SpaceView3D to track the SpaceView which is currently modified by the user
		self._handle_trackActiveWindow = bpy.types.SpaceView3D.draw_handler_add(self.trackActiveWindow, (context,), 'WINDOW', 'PRE_VIEW')

		# Register app handlers that check if the LookingGlass shall be updated:
		#  (1) Every time something in the scene changed (for camera movement and scene editing)
		#  (2) Every time, the current frame changed (for animations)
		self._handle_trackDepsgraphUpdates = bpy.app.handlers.depsgraph_update_post.append(self.trackDepsgraphUpdates)
		self._handle_trackFrameChanges = bpy.app.handlers.frame_change_post.append(self.trackDepsgraphUpdates)

		# log info
		LookingGlassAddonLogger.info(" [#] Initialized control handlers.")



		# HANDLERS FOR OPERATOR CONTROL
		# ++++++++++++++++++++++++++++++
		# Create timer event that runs every millisecond to check if the lightfield needs to be updated
		self.timerEvent = context.window_manager.event_timer_add(0.01, window=context.window)

		# add the modal handler
		context.window_manager.modal_handler_add(self)

		# log info
		LookingGlassAddonLogger.info(" [#] Initialized modal operator.")
		LookingGlassAddonLogger.info(" [#] Done.")

		# keep the modal operator running
		return {'RUNNING_MODAL'}



	# modal operator for controlled redrawing of the lightfield
	def modal(self, context, event):

		# update the internal variable for the settings
		self.addon_settings_window_manager = context.window_manager.addon_settings
		self.addon_settings_scene = context.scene.addon_settings

		# if the active scene was changed
		if context.scene.addon_settings != self.addon_settings_scene:

			# update the lightfield window
			# Lightfield Viewport
			if int(self.addon_settings_window_manager.renderMode) == 0:
				context.window_manager.addon_settings.viewport_manual_refresh = True
			# Quilt Viewer
			elif int(self.addon_settings_window_manager.renderMode) == 1:
				LookingGlassAddon.update_lightfield_window(int(self.addon_settings_window_manager.renderMode), LookingGlassAddon.quiltViewerLightfieldImage)

		# cancel the operator, if the lightfield viewport was deactivated
		if not self.addon_settings_window_manager.ShowLightfieldWindow:

			self.cancel(context)
			return {'FINISHED'}

		# update the variable for the current Looking Glass device
		if int(self.addon_settings_window_manager.activeDisplay) != -1: self.device = pylio.DeviceManager.get_active()


		# Control lightfield redrawing in viewport mode
		################################################################

		# if the TIMER event for the lightfield rendering is called AND the automatic render mode is active
		if event.type == 'TIMER' or event.type == 'Z':

			# if something has changed OR the user requested a manual redrawing
			if self.modal_redraw or (not self.modal_redraw and ((self.depsgraph_update_time > 0 and time.time() - self.depsgraph_update_time > LookingGlassAddon.low_resolution_preview_timout) or context.window_manager.addon_settings.viewport_manual_refresh == True)):

				# update the viewport settings
				self.updateViewportSettings(context)

				if (not self.modal_redraw and ((self.depsgraph_update_time > 0 and time.time() - self.depsgraph_update_time > LookingGlassAddon.low_resolution_preview_timout) or context.window_manager.addon_settings.viewport_manual_refresh == True)):

					# reset time of last depsgraph update
					self.depsgraph_update_time = 0

					# reset status variable for manual refreshes
					context.window_manager.addon_settings.viewport_manual_refresh = False

					# set to the currently chosen quality
					self.preset = int(context.scene.addon_settings.quiltPreset)

					# dont skip any views
					self.skip_views = 1
					self.restricted_viewcone_limit = 0

 					# set to redraw
					self.modal_redraw = True

				# render the views
				self.render_view(context)

				# Lightfield Viewport
				if int(self.addon_settings_window_manager.renderMode) == 0 and self.lightfield_image:

					# update the lightfield displayed on the device
					LookingGlassAddon.update_lightfield_window(int(self.addon_settings_window_manager.renderMode), self.lightfield_image)

				# Quilt Viewer
				elif int(self.addon_settings_window_manager.renderMode) == 1 and LookingGlassAddon.quiltViewerLightfieldImage:

					# update the lightfield displayed on the device
					LookingGlassAddon.update_lightfield_window(int(self.addon_settings_window_manager.renderMode), LookingGlassAddon.quiltViewerLightfieldImage)

				else:

					# update the lightfield displayed on the device: show the demo quilt
					LookingGlassAddon.update_lightfield_window(-1, None)

				# running modal
				return {'RUNNING_MODAL'}

		# pass event through
		return {'PASS_THROUGH'}

	# Application handler that continously checks for changes of the depsgraph
	def trackDepsgraphUpdates(self, scene, depsgraph):

		# if no quilt rendering is currently Running
		if not LookingGlassAddon.RenderInvoked:

			# if automatic live view is activated AND something in the scene has changed
			if (int(self.addon_settings_window_manager.renderMode) == 0 and int(self.addon_settings_window_manager.lightfieldMode) == 0) and len(depsgraph.updates.values()) > 0:
				# print("DEPSGRAPH UPDATE: ", depsgraph.updates.values())

				# remember time of last depsgraph update
				self.depsgraph_update_time = time.time()

				# allow an update of the Looking Glass viewport
				self.modal_redraw = True

				# if the "no preview" is activated
				if self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '0':

					# don't allow an update of the Looking Glass viewport
					self.modal_redraw = False

					# we don't redraw, because changes are only updated after the user interaction finished
					pass

				# if the "low resolution preview" is activated
				elif self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '1':

					# activate them
					self.preset = int(list(pylio.LookingGlassQuilt.formats.get().keys())[-1])

				# if the "skip views preview I" is activated
				elif self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '2':

					# skip every second view during rendering
					self.skip_views = 2

				# if the "skip views preview II" is activated
				elif self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '3':

					# skip every third view during rendering
					self.skip_views = 3

				# if the "restricted viewcone preview" is activated
				elif self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '4':

					# only show the center 33% of all views
					self.restricted_viewcone_limit = int(self.qs[self.preset]["total_views"] / 3)

				else:

					# set to the currently chosen quality
					self.preset = int(scene.addon_settings.quiltPreset)
					self.skip_views = 1

			# if quilt viewer is active AND an image is selected
			elif int(self.addon_settings_window_manager.renderMode) == 1 and self.addon_settings_window_manager.quiltImage != None:

				# set status variable
				changed = False

				# set to the currently chosen quality
				self.preset = int(scene.addon_settings.quiltPreset)

				# TODO: Hacky, but this identifies color management changes
				# go through the updates
				for DepsgraphUpdate in depsgraph.updates.values():
					#print(" # ", DepsgraphUpdate.is_updated_geometry, DepsgraphUpdate.is_updated_shading, DepsgraphUpdate.is_updated_transform, DepsgraphUpdate.id.name)

					if DepsgraphUpdate.is_updated_geometry == True and DepsgraphUpdate.is_updated_shading == True and DepsgraphUpdate.is_updated_transform == True:

						# update status variable
						changed = False

						break

				# are there any changes in the image or color management settings?
				if LookingGlassAddon.quiltViewAsRender != self.addon_settings_window_manager.quiltImage.use_view_as_render or LookingGlassAddon.quiltImageColorSpaceSetting.name != self.addon_settings_window_manager.quiltImage.colorspace_settings.name:

					# update status variable
					changed = True

				# update the quilt image, if something had changed
				if changed == True: self.addon_settings_window_manager.quiltImage = self.addon_settings_window_manager.quiltImage


	# this function is called as a draw handler to enable the Looking Glass Addon
	# to keep track of the SpaceView3D which is currently manipulated by the User
	def trackActiveWindow(self, context):

		# if the space data exists AND this is not the active window
		if context.space_data != None and LookingGlassAddon.BlenderWindow != context.window:

			# in any case, we need to track the active window
			# NOTE: this is important for finding the correct "Scene" and "View Layer"
			LookingGlassAddon.BlenderWindow = context.window



	# set up the camera for each view and the shader of the rendering object
	def setupVirtualCameraForView(self, view, viewMatrix, projectionMatrix):

		# use the context override class method
		return self._override.setupVirtualCameraForView(view, self.qs[self.preset]["total_views"], self.device.viewCone, self.device.aspect, viewMatrix, projectionMatrix)


	# Save the viewport settings
	def saveViewportSettings(self):

		# use the context override class method
		return self._override.saveViewportSettings()


	# Update the viewport settings
	def updateViewportSettings(self, context):

		# if the settings shall be taken from a Blender viewport
		if self.addon_settings_scene.viewportMode == 'BLENDER':

			# check if the space still exists
			found = False
			for workspace in bpy.data.workspaces:
				for screen in workspace.screens:
					for area in screen.areas:
						for space in area.spaces:
							if space.type == 'VIEW_3D':
								if LookingGlassAddon.BlenderViewport == space:

									# get the area
									LookingGlassAddon.BlenderViewportArea = area

									found = True
									break

			# if the SpaceView3D still exists
			if found == True:

				# set the context
				self._override.set_context(context)

				# assign the selected viewport
				return self._override.updateViewportSettings(LookingGlassAddon.BlenderViewport)

			else:

				# reset the global variable and fall back to custom settings
				LookingGlassAddon.BlenderViewport = None

				# set the context
				self._override.set_context(context)

				# assign the selected viewport
				return self._override.updateViewportSettings(LookingGlassAddon.BlenderViewport)


	# Restore the viewport settings
	def restoreViewportSettings(self):

		# use the context override class method
		return self._override.restoreViewportSettings()


	@staticmethod
	def from_texture_to_numpy_array(offscreen, array):
		"""copy the current texture to a numpy array"""

		with offscreen.bind():

			# TODO: IN LATER VERSIONS OF ALICE/LG THAT DO NOT SUPPORT 2.93
			#		 ANYMORE, THE bgl.* CALLS SHOULD BE REMOVED
			# for Blender versions earlier than 3.0 (prior to the major BGL changes)
			if bpy.app.version < (3, 0, 0):

				# activate the texture
				bgl.glActiveTexture(bgl.GL_TEXTURE0)
				bgl.glBindTexture(bgl.GL_TEXTURE_2D, offscreen.color_texture)

				# then we pass the numpy array to the bgl.Buffer as template,
				# which causes Blender to write the buffer data into the numpy array directly
				buffer = bgl.Buffer(bgl.GL_BYTE, array.shape, array)

				# set correct colormode
				if array.shape[2] == 3: colormode = bgl.GL_RGB
				if array.shape[2] == 4: colormode = bgl.GL_RGBA

				# write pixel data from texture into the buffer (numpy array)
				bgl.glGetTexImage(bgl.GL_TEXTURE_2D, 0, colormode, bgl.GL_UNSIGNED_BYTE, buffer)
				bgl.glBindTexture(bgl.GL_TEXTURE_2D, 0)

			# for Blender versions later than 3.0 (after the major BGL changes)
			else:

				# then we pass the numpy array to the gpu.types.Buffer as template,
				# which causes Blender to write the buffer data into the numpy array directly
				buffer = gpu.types.Buffer('UBYTE', array.shape, array)

				# get the active framebuffer
				framebuffer = gpu.state.active_framebuffer_get()

				# write pixel data from texture into the buffer (numpy array)
				framebuffer.read_color(0, 0, array.shape[1], array.shape[0], array.shape[2], 0, 'UBYTE', data=buffer)

	# Draw function which copies data from the 3D View
	def render_view(self, context):

		# if the quilt must be redrawn
		if (self.addon_settings_scene.lookingglassCamera or LookingGlassAddon.BlenderViewport):

			# UPDATE QUILT SETTINGS
			# ++++++++++++++++++++++++++++++++++++++++++++++++

			self.start_multi_view = time.time()

			# if the quilt and view settings changed
			if self.last_preset != self.preset or self.lightfield_image == None:

				# update the preset variable
				self.last_preset = self.preset

				# free the view data of the lightfield image
				if self.lightfield_image: self.lightfield_image.clear_views()

				# delete the current LightfieldImage
				if self.lightfield_image: self.lightfield_image = None

				# TODO: Actually we would use "RGB" and a numpy array with 3
				#		color channels, because that would be more efficient.
				#		But we can't read in RGB mode to gpu.types.Buffer
				#	   due to Blender's default OpenGL settings:
				#
				#	   https://developer.blender.org/T91828
				#
				#		If we don't so it that way, it causes crashes:
				#
				#		https://github.com/regcs/AliceLG/issues/59
				#
				#		The Blender behaviour was fixed for v.3.0+. At the
				#		point when Alice/LG does not support 2.93 anymore,
				#		we can change this. (because the Blender fix is not)

				# create a pylio LightfieldImage
				self.lightfield_image = pylio.LightfieldImage.new(pylio.LookingGlassQuilt, id=self.preset, colormode='RGBA')

				# create a new set of LightfieldViews
				self.lightfield_image.set_views([pylio.LightfieldView(np.empty((self.qs[self.preset]["view_height"], self.qs[self.preset]["view_width"], 4), dtype=np.uint8), pylio.LightfieldView.formats.numpyarray) for view in range(0, self.qs[self.preset]["total_views"])], pylio.LightfieldView.formats.numpyarray)

			LookingGlassAddonLogger.debug("Start rendering lightfield views ...")
			LookingGlassAddonLogger.debug(" [#] View dimensions: %i x %i" % (self.qs[self.preset]["view_width"], self.qs[self.preset]["view_height"]))
			LookingGlassAddonLogger.debug(" [#] LightfieldImage views: %i" % len(self.lightfield_image.get_view_data()))
			LookingGlassAddonLogger.debug(" [#] Using quilt preset: %i (%s, %i x %i)" % (self.preset, self.qs[self.preset]['description'], self.lightfield_image.metadata['quilt_width'], self.lightfield_image.metadata['quilt_height']))
			LookingGlassAddonLogger.debug(" [#] Preview mode: %s (selected: %s)" % (self.addon_settings_window_manager.viewport_use_preview_mode, self.addon_settings_window_manager.lightfield_preview_mode))


			# PREPARE VIEW & PROJECTION MATRIX
			# ++++++++++++++++++++++++++++++++++++++++++++++++

			# select camera that belongs to the view
			camera = self.addon_settings_scene.lookingglassCamera

			# PREPARE THE MODELVIEW AND PROJECTION MATRICES
			# if a camera is selected
			if camera != None:

				# get camera's modelview matrix
				view_matrix = camera.matrix_world.copy()

				# correct for the camera scaling
				view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.x, 4, (1, 0, 0))
				view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.y, 4, (0, 1, 0))
				view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.z, 4, (0, 0, 1))

				# calculate the inverted view matrix because this is what the draw_view_3D function requires
				camera_view_matrix = view_matrix.inverted_safe()

				# get the camera's projection matrix
				camera_projection_matrix = camera.calc_matrix_camera(
						depsgraph=context.view_layer.depsgraph,
						x = self.qs[self.preset]["view_width"],
						y = self.qs[self.preset]["view_height"],
						scale_x = 1.0,
						scale_y = (self.qs[self.preset]["rows"] / self.qs[self.preset]["columns"]) / self.device.aspect,
					)

				LookingGlassAddonLogger.debug(" [#] Geting view & projection matrices took %.6f s" % (time.time() - self.start_multi_view))


				# RENDER THE VIEWS
				# ++++++++++++++++++++++++++++++++++++++++++++++++

				# loop through all required views
				for view in range(0, self.qs[self.preset]["total_views"]):

					with self.qs[self.preset]["viewOffscreen"][view].bind():

						start_test = time.time()
						# calculate the offset-projection of the current view
						view_matrix, projection_matrix = self.setupVirtualCameraForView(view, camera_view_matrix.copy(), camera_projection_matrix.copy())

						LookingGlassAddonLogger.debug(" [#] [%i] Setting up view camera took %.3f ms" % (view, (time.time() - start_test) * 1000))
						start_test = time.time()

						# if the "skip views preview" is activated AND this view shall be skipped
						if (self.addon_settings_window_manager.viewport_use_preview_mode and (self.addon_settings_window_manager.lightfield_preview_mode == '2' or self.addon_settings_window_manager.lightfield_preview_mode == '3')) and view % self.skip_views:

							# clear LightfieldView array's color data (so it appears black)
							self.lightfield_image.views[view]['view'].data[:] = 0

							LookingGlassAddonLogger.debug(" [#] [%i] Clearing skipped view's numpy array took %.3f ms" % (view, (time.time() - start_test) * 1000))

						# if the "Restricted viewcone preview" is activated AND this view shall be skipped
						elif (self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '4') and (view < self.restricted_viewcone_limit or view > self.qs[self.preset]["total_views"] - self.restricted_viewcone_limit):

							# clear LightfieldView array's color data (so it appears black)
							self.lightfield_image.views[view]['view'].data[:] = 0

							LookingGlassAddonLogger.debug(" [#] [%i] Clearing skipped view's numpy array took %.3f ms" % (view, (time.time() - start_test) * 1000))

						else:

							# if the lightfield window is not active anymore, stop
							if not (context or context.window_manager.addon_settings.ShowLightfieldWindow):
								continue

							# draw the viewport rendering to the offscreen for the current view
							self.qs[self.preset]["viewOffscreen"][view].draw_view3d(
								# we use the "Scene" and the "View Layer" that is active in the Window
								# the user currently works in
								scene=context.scene,
								view_layer=context.view_layer,
								view3d=self._override.space_data,
								region=self._override.region,
								view_matrix=view_matrix,
								projection_matrix=projection_matrix,
								do_color_management = True)

							LookingGlassAddonLogger.debug(" [#] [%i] Drawing view into offscreen took %.3f ms" % (view, (time.time() - start_test) * 1000))

				# restore all viewport shading and overlay settings
				self.restoreViewportSettings()

				LookingGlassAddonLogger.debug("-----------------------------")
				LookingGlassAddonLogger.debug("Rendering all views took in total %.3f ms" % ((time.time() - self.start_multi_view) * 1000))
				LookingGlassAddonLogger.debug("-----------------------------")


				# COPY THE VIEWS INTO A BUFFER
				# NOTE: We do this in a separate loop, because for an unknown
				#		reason (probably something Blender internal), it is faster.
				# ++++++++++++++++++++++++++++++++++++++++++++++++

				self.start_multi_view = time.time()

				# loop through all required views
				for view in range(0, self.qs[self.preset]["total_views"]):

					# if the "skip views preview" is activated AND this view shall be skipped
					if (self.addon_settings_window_manager.viewport_use_preview_mode and (self.addon_settings_window_manager.lightfield_preview_mode == '2' or self.addon_settings_window_manager.lightfield_preview_mode == '3')) and view % self.skip_views:

						continue
					# if the "Restricted viewcone preview" is activated AND this view shall be skipped
					elif (self.addon_settings_window_manager.viewport_use_preview_mode and self.addon_settings_window_manager.lightfield_preview_mode == '4') and (view < self.restricted_viewcone_limit or view > self.qs[self.preset]["total_views"] - self.restricted_viewcone_limit):

						continue
					else:

						start_test = time.time()

						# copy texture into LightfieldView array
						self.from_texture_to_numpy_array(self.qs[self.preset]["viewOffscreen"][view], self.lightfield_image.views[view]['view'].data[:])

						LookingGlassAddonLogger.debug(" [#] [%i] Copying texture to numpy array took %.3f ms" % (view, (time.time() - start_test) * 1000))

				LookingGlassAddonLogger.debug("-----------------------------")
				LookingGlassAddonLogger.debug("Copying all views took in total %.3f ms" % ((time.time() - self.start_multi_view) * 1000))
				LookingGlassAddonLogger.debug("-----------------------------")

			# reset draw variable:
			# This is here to prevent excessive redrawing
			self.modal_redraw = False



# ------------ CAMERA FRUSTUM RENDERING -------------
# Class for rendering a camera frustum reprsenting the Looking Glass
# in Blenders 3D viewport
class FrustumRenderer:

	# Inititalize the camera frustum drawing
	def __init__(self):

		# Blender draw handler for the frustum
		self.frustum_draw_handler = None

		# variables for the frustum
		self.frustum_indices_lines = None
		self.frustum_indices_faces = None
		self.frustum_indices_focalplane_outline = None
		self.frustum_indices_focalplane_face = None
		self.frustum_shader = None

		# notify addon that frustum is activated
		LookingGlassAddon.FrustumInitialized = True



	# deinititalize the camera frustum drawing
	def __del__(self):

		# remove the draw handler for the frustum drawing
		if self.frustum_draw_handler:
			bpy.types.SpaceView3D.draw_handler_remove(self.frustum_draw_handler, 'WINDOW')
			self.frustum_draw_handler = None

		# notify addon that frustum is deactivated
		LookingGlassAddon.FrustumInitialized = False



	# cancel frustum drawing
	def stop(self):

		# remove the draw handler for the frustum drawing
		if self.frustum_draw_handler:
			bpy.types.SpaceView3D.draw_handler_remove(self.frustum_draw_handler, 'WINDOW')
			self.frustum_draw_handler = None

		LookingGlassAddon.FrustumInitialized = False

		# log info
		LookingGlassAddonLogger.info("Camera frustum drawing handler stopped.")

		# return None since this is expected by the operator
		return None



	# start the frustum drawing
	def start(self, context):

		# setup the camera frustum & shader
		self.setupCameraFrustumShader()

		# add draw handler to display the frustum of the Looking Glass camera
		# after everything else has been drawn in the view
		self.frustum_draw_handler = bpy.types.SpaceView3D.draw_handler_add(self.drawCameraFrustum, (context,), 'WINDOW', 'POST_VIEW')

		# log info
		LookingGlassAddonLogger.info("Camera frustum drawing handler started.")

		# keep the modal operator running
		return self.frustum_draw_handler



	# setup the camera frustum shader
	def setupCameraFrustumShader(self):

		# we predefine the indices because these never change
		self.frustum_indices_lines = (
			(0, 1), (0, 3), (1, 2), (3, 2),
			(4, 5), (4, 7), (5, 6), (7, 6),
			(0, 4), (1, 5), (3, 7), (2, 6))

		self.frustum_indices_faces = (
				# front
				(0, 1, 2),
				(2, 3, 0),
				# right
				(1, 5, 6),
				(6, 2, 1),
				# back
				(7, 6, 5),
				(5, 4, 7),
				# left
				(4, 0, 3),
				(3, 7, 4),
				# bottom
				(4, 5, 1),
				(1, 0, 4),
				# top
				(3, 2, 6),
				(6, 7, 3)
			)

		self.frustum_indices_focalplane_outline = (
			(8, 9), (9, 10), (10, 11), (11, 8))

		self.frustum_indices_focalplane_face = (
				# focal plane
				(8, 9, 10),
				(10, 11, 8)
			)

		# compile the shader that will be used for drawing
		if sys.version_info.major >= 4 or (sys.version_info.major >= 0 and sys.version_info.minor >= 0):
			self.frustum_shader = gpu.shader.from_builtin('UNIFORM_COLOR')
		else:
			self.frustum_shader = gpu.shader.from_builtin('3D_UNIFORM_COLOR')



	# drawing function, which draws the camera frustum in the active SpaceView3D
	def drawCameraFrustum(self, context):

		# if a camera is selected AND the space is not in camera mode
		if self and context:
			if hasattr(context.scene, "addon_settings") and context.scene.addon_settings.lookingglassCamera in [obj for obj in context.view_layer.objects]:
				if (context.space_data != None and context.space_data.region_3d != None) and context.space_data.region_3d.view_perspective != 'CAMERA':

					# currently selected camera
					camera = context.scene.addon_settings.lookingglassCamera

					# if the camera is visible
					if camera.hide_get() == False:

						# get modelview matrix
						view_matrix = camera.matrix_world.copy()

						# correct for the camera scaling
						view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.x, 4, (1, 0, 0))
						view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.y, 4, (0, 1, 0))
						view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.z, 4, (0, 0, 1))

						# we obtain the viewframe of the camera to calculate the focal and clipping world_clip_planes_calc_clip_distance
						# based on the intercept theorems
						view_frame = camera.data.view_frame(scene=context.scene)
						view_frame_upper_right = view_frame[0]
						view_frame_lower_right = view_frame[1]
						view_frame_lower_left = view_frame[2]
						view_frame_upper_left = view_frame[3]
						view_frame_distance = abs(view_frame_upper_right[2])

						# get the clipping settings
						clipStart = camera.data.clip_start
						clipEnd = camera.data.clip_end
						focalPlane = context.scene.addon_settings.focalPlane

						# TODO: Find a way to predefine the vertex buffers and batches so that these don't need to be created in every frame
						# define the vertices of the camera frustum in camera coordinates
						# NOTE: - the z-value is negative, because the Blender camera always looks into negative z-direction
						coords_local = [
										# near clipping plane
										(view_frame_lower_right[0] / view_frame_distance * clipStart, view_frame_lower_right[1] / view_frame_distance * clipStart, -clipStart), (view_frame_lower_left[0] / view_frame_distance * clipStart, view_frame_lower_left[1] / view_frame_distance * clipStart, -clipStart),
										(view_frame_upper_left[0] / view_frame_distance * clipStart, view_frame_upper_left[1] / view_frame_distance * clipStart, -clipStart), (view_frame_upper_right[0] / view_frame_distance * clipStart, view_frame_upper_right[1] / view_frame_distance * clipStart, -clipStart),
										# far clipping plane
										(view_frame_lower_right[0] / view_frame_distance * clipEnd, view_frame_lower_right[1] / view_frame_distance * clipEnd, -clipEnd), (view_frame_lower_left[0] / view_frame_distance * clipEnd, view_frame_lower_left[1] / view_frame_distance * clipEnd, -clipEnd),
										(view_frame_upper_left[0] / view_frame_distance * clipEnd, view_frame_upper_left[1] / view_frame_distance * clipEnd, -clipEnd), (view_frame_upper_right[0] / view_frame_distance * clipEnd, view_frame_upper_right[1] / view_frame_distance * clipEnd, -clipEnd),
										# focal plane
										(view_frame_lower_right[0] / view_frame_distance * focalPlane, view_frame_lower_right[1] / view_frame_distance * focalPlane, -focalPlane), (view_frame_lower_left[0] / view_frame_distance * focalPlane, view_frame_lower_left[1] / view_frame_distance * focalPlane, -focalPlane),
										(view_frame_upper_left[0] / view_frame_distance * focalPlane, view_frame_upper_left[1] / view_frame_distance * focalPlane, -focalPlane), (view_frame_upper_right[0] / view_frame_distance * focalPlane, view_frame_upper_right[1] / view_frame_distance * focalPlane, -focalPlane),
										]

						# if the camera fustum shall be drawn
						if context.scene.addon_settings.showFrustum == True:
							batch_lines = batch_for_shader(self.frustum_shader, 'LINES', {"pos": coords_local}, indices=self.frustum_indices_lines)
							batch_faces = batch_for_shader(self.frustum_shader, 'TRIS', {"pos": coords_local}, indices=self.frustum_indices_faces)

						# if the focal plane shall be drawn
						if context.scene.addon_settings.showFocalPlane == True:
							batch_focalplane_outline = batch_for_shader(self.frustum_shader, 'LINES', {"pos": coords_local}, indices=self.frustum_indices_focalplane_outline)
							batch_focalplane_face = batch_for_shader(self.frustum_shader, 'TRIS', {"pos": coords_local}, indices=self.frustum_indices_focalplane_face)

						# draw everything
						self.frustum_shader.bind()

						# get the current projection matrix
						viewMatrix = gpu.matrix.get_model_view_matrix()
						projectionMatrix = gpu.matrix.get_projection_matrix()

						# load the model view matrix, which transforms the local camera coordinates to world coordinates.
						# this makes sure that the frustum is always drawn relative to the camera location
						gpu.matrix.reset()
						gpu.matrix.load_matrix(viewMatrix @ view_matrix)
						gpu.matrix.load_projection_matrix(projectionMatrix)

						gpu.state.depth_test_set('LESS_EQUAL')
						gpu.state.depth_mask_set(True)

						# if the camera fustum shall be drawn
						if context.scene.addon_settings.showFrustum == True:
							# draw outline
							self.frustum_shader.uniform_float("color", (0.3, 0, 0, 1))
							batch_lines.draw(self.frustum_shader)

						# if the focal plane shall be drawn
						if context.scene.addon_settings.showFocalPlane == True:
							# draw focal plane outline
							self.frustum_shader.uniform_float("color", (1, 1, 1, 1))
							batch_focalplane_outline.draw(self.frustum_shader)

						gpu.state.depth_mask_set(False)
						gpu.state.blend_set('ALPHA')

						# if the camera fustum shall be drawn
						if context.scene.addon_settings.showFrustum == True:
							# fill faces
							self.frustum_shader.uniform_float("color", (0.5, 0.5, 0.5, 0.05))
							batch_faces.draw(self.frustum_shader)

						# if the focal plane shall be drawn
						if context.scene.addon_settings.showFocalPlane == True:
							# draw focal plane face
							self.frustum_shader.uniform_float("color", (0.1, 0.1, 0.1, 0.25))
							batch_focalplane_face.draw(self.frustum_shader)

						gpu.state.depth_test_set('NONE')
						gpu.state.blend_set('NONE')

						# reset the matrices to their original state
						gpu.matrix.reset()
						gpu.matrix.load_matrix(viewMatrix)
						gpu.matrix.load_projection_matrix(projectionMatrix)



# ------------ BLOCKS PREVIEW RENDERING -------------
# Class for blocks. Instances store all the data of a Block.
class Block:

	# Inititalize the block
	def __init__(self, x, y, width, height):

		# the texture of an image in the image editor
		self.image_texture = None

		# quilt presets
		self.qs = pylio.LookingGlassQuilt.formats.get()
		self.preset = None
		self.device = None

		# block parameters
		self.active = False
		self.x = x
		self.y = y
		self.width = int(width * bpy.context.scene.addon_settings.viewport_block_scaling_factor)
		self.height = int(height * bpy.context.scene.addon_settings.viewport_block_scaling_factor)
		self.border_width = 0.000
		self.border_color = (0.0086, 0.0086, 0.0086, 1.0)
		self.aspect = self.width / self.height
		self.angle = 0
		self.view = 0
		self.view_cone = 40
		self.shading_type = 'SOLID'

		# GPU and rendering
		self.changed = True
		self.camera = None
		self.view_matrix = None
		self.projection_matrix = None
		self.offscreen_canvas = gpu.types.GPUOffScreen(self.width, self.height)
		self.offscreen_view = None

		# create shaders
		self.__update_shaders()



	# create a view matrix from a virtual camera position, target vector, and up
	# vector
	def __update_shaders(self):
		
		# in Blender versions prior to BGL deprecation:
		if bpy.app.version < (3, 4, 0):

			# define shaders
			self.vertex_shader = '''
				uniform mat4 modelMatrix;
				uniform mat4 viewProjectionMatrix;

				in vec2 position;
				in vec2 uv;

				out vec2 texCoord;

				void main()
				{
					texCoord = uv;
					gl_Position = viewProjectionMatrix * modelMatrix * vec4(position, 0.0, 1.0);
				}
			'''

			self.fragment_shader = '''
				uniform sampler2D view_texture;
				uniform vec2 dimensions;
				uniform float alpha;

				in vec2 texCoord;
				out vec4 FragColor;

				void main()
				{
					// Input info
					vec2 boxPos; // The position of the center of the box (in normalized coordinates)
					vec2 boxBnd; // The half-bounds (radii) of the box (in normalzied coordinates)
					float radius;// Radius


					boxPos = vec2(0.5, 0.5);	// center of the screen
					boxBnd = vec2(0.5, 0.5);  // half of the area
					radius = 0.04;

					// Normalize the pixel coordinates (this is "passTexCoords" in your case)
					vec2 uv = texCoord;
					vec2 aspectRatio = vec2(dimensions.x/dimensions.y, 1.0);

					// In order to make sure visual distances are preserved, we multiply everything by aspectRatio
					uv *= aspectRatio;
					boxPos *= aspectRatio;
					boxBnd *= aspectRatio;

					// Output to screen
					float a = length(max(abs(uv - boxPos) - boxBnd + radius, 0.0)) - radius;

					// Shadertoy doesn't have an alpha in this case
					if(a <= 0.0){
						FragColor = texture(view_texture, texCoord);
						FragColor.a = alpha;
					}else{
						FragColor = vec4(0.0, 0.0, 0.0, 0.0);
					}

				}
			'''

			# compile hologram preview / blocks shader
			self.shader = gpu.types.GPUShader(self.vertex_shader, self.fragment_shader)

		# from Blender 3.4 on we need to use the GPU module to create shaders,
		# which is the new way to compile cross-platform shaders
		else:

			# The shader info stores all th
			hologram_preview_shader_info = gpu.types.GPUShaderCreateInfo()

			# Defines vertex shader inputs and uniforms that are now called constants
			hologram_preview_shader_info.vertex_in(0, 'VEC2', "position")
			hologram_preview_shader_info.vertex_in(1, 'VEC2', "uv")
			hologram_preview_shader_info.push_constant('MAT4', "modelMatrix")
			hologram_preview_shader_info.push_constant('MAT4', "viewProjectionMatrix")
			
			# Defines fragment shader inputs and uniforms that are now called constants
			hologram_preview_shader_info.sampler(0, 'FLOAT_2D', "view_texture")
			hologram_preview_shader_info.push_constant('VEC2', "dimensions")
			hologram_preview_shader_info.push_constant('FLOAT', "alpha")

			# Define as Interface the attributes that will be transferred from the vertex shader to the fragment shader. 
			# Before they would be both a vertex shader output and fragment shader input.
			# An interface can be flat(), no_perspective() or smooth()
			# Warning: You need to give a string to the GPUStageInterfaceInfo() or the shader will not work. Any string will work.
			hologram_preview_shader_interface = gpu.types.GPUStageInterfaceInfo("shader_hologram_preview")	
			hologram_preview_shader_interface.smooth('VEC2', "texCoord")
			hologram_preview_shader_info.vertex_out(hologram_preview_shader_interface)

			# fragment shader output
			hologram_preview_shader_info.fragment_out(0, 'VEC4', 'FragColor')

			# define shaders
			hologram_preview_shader_info.vertex_source(
				'''
				void main()
				{
					texCoord = uv;
					gl_Position = viewProjectionMatrix * modelMatrix * vec4(position, 0.0, 1.0);
				}
				'''
			)

			hologram_preview_shader_info.fragment_source(
				'''
				void main()
				{
					// Input info
					vec2 boxPos; // The position of the center of the box (in normalized coordinates)
					vec2 boxBnd; // The half-bounds (radii) of the box (in normalzied coordinates)
					float radius;// Radius

					boxPos = vec2(0.5, 0.5);	// center of the screen
					boxBnd = vec2(0.5, 0.5);  // half of the area
					radius = 0.04;

					// Normalize the pixel coordinates (this is "passTexCoords" in your case)
					vec2 uv = texCoord;
					vec2 aspectRatio = vec2(dimensions.x/dimensions.y, 1.0);

					// In order to make sure visual distances are preserved, we multiply everything by aspectRatio
					uv *= aspectRatio;
					boxPos *= aspectRatio;
					boxBnd *= aspectRatio;

					// Output to screen
					float a = length(max(abs(uv - boxPos) - boxBnd + radius, 0.0)) - radius;

					// Shadertoy doesn't have an alpha in this case
					if(a <= 0.0){
						FragColor = texture(view_texture, texCoord);
						FragColor.a = alpha;
					}else{
						FragColor = vec4(0.0, 0.0, 0.0, 0.0);
					}

				}
				'''
			)

			# compile hologram preview / blocks shader
			self.shader = gpu.shader.create_from_info(hologram_preview_shader_info)
			del hologram_preview_shader_info
			del hologram_preview_shader_interface

		# for portrait orientations
		if self.aspect < 1:
			position_vertices = ((-self.aspect, -1), (self.aspect, -1), (self.aspect, 1), (-self.aspect, 1))

		# for portrait orientations
		elif self.aspect >= 1:
			position_vertices = ((-1, -1/self.aspect), (1, -1/self.aspect), (1, 1/self.aspect), (-1, 1/self.aspect))

		# compile block shader
		self.batch = batch_for_shader(
			self.shader, 'TRI_FAN',
			{
				"position": position_vertices,
				"uv": ((0, 0), (1, 0), (1, 1), (0, 1)),
			},
		)

	# this is widely based on the following stackexchange thread:
	# https://blender.stackexchange.com/questions/16472/how-can-i-get-the-cameras-projection-matrix/86570#86570
	# which is a python translation of Blenders internal projection matrix calculations
	def __projection_matrix(self, x, y, scale_x, scale_y):

		# set camera parameters that we require for the block rendering
		camera_lens = 50.0
		camera_sensor_width = 36.0
		camera_sensor_height = 24.0
		camera_shift_x = 0
		camera_shift_y = 0
		camera_clip_start = 0.001
		camera_clip_end = 100

		def view_plane(winx, winy, xasp, yasp):

			#/* fields rendering */
			ycor = yasp / xasp

			#/* perspective camera */
			pixsize = (camera_sensor_width * 0.001) / camera_lens

			# for landscape devices
			if (self.aspect >= 1):
			  viewfac = winx

			# for portrait devices
			else:
			  viewfac = ycor * winy

			pixsize /= viewfac

			#/* extra zoom factor */
			pixsize *= 1

			#/* compute view plane:
			# * fully centered, zbuffer fills in jittered between -.5 and +.5 */
			xmin = -0.5 * winx
			ymin = -0.5 * ycor * winy
			xmax =  0.5 * winx
			ymax =  0.5 * ycor * winy

			#/* lens shift and offset */
			dx = camera_shift_x * viewfac
			dy = camera_shift_y * viewfac

			xmin += dx
			ymin += dy
			xmax += dx
			ymax += dy

			#/* the window matrix is used for clipping, and not changed during OSA steps */
			#/* using an offset of +0.5 here would give clip errors on edges */
			xmin *= pixsize
			xmax *= pixsize
			ymin *= pixsize
			ymax *= pixsize

			return xmin, xmax, ymin, ymax

		left, right, bottom, top = view_plane(x, y, scale_x, scale_y)

		farClip, nearClip = camera_clip_end, camera_clip_start

		Xdelta = right - left
		Ydelta = top - bottom
		Zdelta = farClip - nearClip

		mat = [[0]*4 for i in range(4)]

		mat[0][0] = nearClip * 2 / Xdelta
		mat[1][1] = nearClip * 2 / Ydelta
		mat[2][0] = (right + left) / Xdelta #/* note: negate Z  */
		mat[2][1] = (top + bottom) / Ydelta
		mat[2][2] = -(farClip + nearClip) / Zdelta
		mat[2][3] = -1
		mat[3][2] = (-2 * nearClip * farClip) / Zdelta

		return sum([c for c in mat], [])

	# create a view matrix from a virtual camera position, target vector, and up
	# vector
	def __matrix_look_at(self, camera_location, camera_target, camera_up):

		# calculate coordinate vectors
		camera_direction = (camera_location - camera_target).normalized()
		camera_right = camera_up.cross(camera_direction).normalized()
		camera_up = camera_direction.cross(camera_right).normalized()

		# rotation matrix
		matrix_rotation = Matrix()
		matrix_rotation[0][:3] = camera_right[:]
		matrix_rotation[1][:3] = camera_up[:]
		matrix_rotation[2][:3] = camera_direction[:]

		# translation matrix
		matrix_translation = Matrix.Translation(-camera_location)

		# calculate and return view matrix ("lookat" matrix)
		return matrix_rotation @ matrix_translation

	# update the view and perspective matrices
	def __update_matrices(self, context):

		if context.space_data is None:
			return None

		# calculate the inverted view matrix because this is what the draw_view_3D function requires
		self.view_matrix = self.__matrix_look_at(Vector((sin(radians(self.angle)) * 3.1, 0, cos(radians(self.angle)) * 3.1)), Vector((0, 0, 0)), Vector((0, 1, 0)))

		# get the camera's projection matrix
		self.projection_matrix = self.__projection_matrix(
				x = self.qs[self.preset]['view_width'],
				y = self.qs[self.preset]['view_height'],
				scale_x = 1.0,
				scale_y = (self.qs[self.preset]['rows'] / self.qs[self.preset]['columns']) / self.aspect,
			)

	# test if mouse is hovering over the block
	def is_mouse_over(self, context):
		if context is None or context.space_data is None:
			return False

		# for 3D viewports
		if context.space_data.type == 'VIEW_3D':
			# for left positioned blocks
			if context.scene.addon_settings.viewport_block_alignment == 'left':
				if (LookingGlassAddon.mouse_region_x > self.x and LookingGlassAddon.mouse_region_x < self.x + self.width) and (LookingGlassAddon.mouse_region_y > self.y and LookingGlassAddon.mouse_region_y < self.y + self.height):
					return True

			# for right positioned blocks
			if context.scene.addon_settings.viewport_block_alignment == 'right':
				if (LookingGlassAddon.mouse_region_x > context.area.width - self.width - self.x - 50 and LookingGlassAddon.mouse_region_x < context.area.width - self.width - self.x - 50 + self.width) and (LookingGlassAddon.mouse_region_y > self.y and LookingGlassAddon.mouse_region_y < self.y + self.height):
					return True

		# for image editors
		elif context.space_data.type == 'IMAGE_EDITOR':

			if (LookingGlassAddon.mouse_region_x > self.x and LookingGlassAddon.mouse_region_x < self.x + self.width) and (LookingGlassAddon.mouse_region_y > self.y and LookingGlassAddon.mouse_region_y < self.y + self.height):
				return True

	# check if block has changed
	def has_changed(self, context):
		if self.changed or self.shading_type != context.space_data.shading.type:
			return True

	# store lightfield image in the block data
	def set_lightfield_image(self, lightfield_image):
		if  not (lightfield_image is None or self.lightfield_image == lightfield_image):
			self.lightfield_image = lightfield_image

			# block has changed
			self.changed = True

	# store dimensions in the block data
	def set_dimensions(self, width, height):

		if (not width is None and width != self.width) or (not height is None and height != self.height):

			if not width is None:
				self.width = int(width)

			if not height is None:
				self.height = int(height)

			# free the offscreen object
			if self.offscreen_canvas: self.offscreen_canvas.free()

			# create a new offscreen block is drawn in
			self.offscreen_canvas = gpu.types.GPUOffScreen(self.width, self.height)

			# block has changed
			self.changed = True

	# set the device type for this Block
	def set_device(self, device):

		if not (device is None or self.device == device):
			self.device = device

			# set the aspect and view cone
			self.aspect = device.aspect
			self.view_cone = device.viewCone

			# update shaders to include the new aspect ratio
			self.__update_shaders()

			# block has changed
			self.changed = True

	# set the quilt preset for this Block
	def set_preset(self, preset):

		if not (preset is None or self.preset == preset):
			self.preset = preset

			# get list of formats
			self.qs = pylio.LookingGlassQuilt.formats.get()

			# free offscreen
			if self.offscreen_view: self.offscreen_view.free()

			# create a new offscreen the view is drawn to
			self.offscreen_view = gpu.types.GPUOffScreen(self.qs[self.preset]['view_width'], self.qs[self.preset]['view_height'])

			# update shaders to include the new aspect ratio
			self.__update_shaders()

			# block has changed
			self.changed = True

	# store view in the block data
	def set_view(self, view):
		if not (view is None or self.view == view):
			self.view = view

			# block has changed
			self.changed = True

	# store view cone in the block data
	def set_view_cone(self, view_cone):
		if not (view_cone is None or self.view_cone == view_cone):
			self.view_cone = view_cone

			# block has changed
			self.changed = True

	# store angle in the block data
	def set_angle(self, angle):
		if not (angle is None or self.angle == angle):
			self.angle = angle

			# block has changed
			self.changed = True

	# store activity value in the block data
	def set_active(self, active):
		if not (active is None or self.active == active):
			self.active = active

			# block has changed
			self.changed = True

	# store aspect ratio of the views in the block data
	def set_aspect(self, aspect):
		if not (aspect is None or self.aspect == aspect):
			self.aspect = aspect

			# update shaders to include the new aspect ratio
			self.__update_shaders()

			# block has changed
			self.changed = True

	# store border_width value in the block data
	def set_border_width(self, border_width):
		if not border_width is None:
			self.border_width = border_width

	# store border_color value in the block data
	def set_border_color(self, border_color):
		if not border_color is None:
			self.border_color = border_color

	# update the block by rendering it in its own offscreen
	def update(self, context):

		# if the block has a view texture
		if self.offscreen_view:

			# update matrices
			self.__update_matrices(context)

			# update shading type variable
			if hasattr(context.space_data, 'shading'): self.shading_type = context.space_data.shading.type

			# start drawing into the preview offscreen
			with self.offscreen_canvas.bind():

				# get active frame buffer
				framebuffer = gpu.state.active_framebuffer_get()

				# clear the framebuffer
				framebuffer.clear(color=(0.0, 0.0, 0.0, 0.0))

				# update shader uniforms
				from struct import pack
				self.shader.bind()
				self.shader.uniform_float("modelMatrix", self.view_matrix)
				self.shader.uniform_float("viewProjectionMatrix", self.projection_matrix)
				#self.shader.uniform_float("aspect", self.aspect)
				self.shader.uniform_vector_float(self.shader.uniform_from_name("dimensions"), pack("2f", self.width, self.height), 2)
				#self.shader.uniform_float("border_width", self.border_width)
				#self.shader.uniform_float("border_color", self.border_color)
				if context.space_data.type == 'VIEW_3D':
					if self.active: self.shader.uniform_float("alpha", 1.0)
					if not self.active: self.shader.uniform_float("alpha", context.scene.addon_settings.viewport_block_alpha)
				if context.space_data.type == 'IMAGE_EDITOR':
					self.shader.uniform_float("alpha", 1)
				self.shader.uniform_sampler("view_texture", self.offscreen_view.texture_color)

				# draw the block image
				self.batch.draw(self.shader)

	# free the block and all its resources
	def free(self):

		# make sure the block is not drawn
		self.changed = False

		# free shader and batch
		if hasattr(self, "shader"): del self.shader
		if hasattr(self, "batch"): del self.batch

		# free offscreens
		try: 
			self.offscreen_view.free()
			self.offscreen_canvas.free()
		finally:
			return

# Class for rendering a Looking Glass Block in Blenders 3D viewport for live preview
class BlockRenderer:

	# ------------ KEYMAP OPERATOR FOR BLOCK RENDERER UPDATES -------------
	class LOOKINGGLASS_OT_update_block_renderer(bpy.types.Operator):
		""" This operator updates the block renderer"""
		bl_idname = "wm.update_block_renderer"
		bl_label = "Alice/LG: Block Renderer Update"

		def invoke(self, context, event):

			if (context is None) or (event is None) or (not context is None and context.area is None) or not hasattr(context.scene, "addon_settings"):
				return {'PASS_THROUGH'}

			# if the hologram preview is not active
			if context.space_data.type == "VIEW_3D" and not context.scene.addon_settings.viewport_block_show:
				return {'PASS_THROUGH'}

			# if the hologram preview is not active
			if context.space_data.type == 'IMAGE_EDITOR' and not context.scene.addon_settings.imageeditor_block_show:
				return {'PASS_THROUGH'}

			# mouse position in window
			LookingGlassAddon.mouse_window_x = event.mouse_x
			LookingGlassAddon.mouse_window_y = event.mouse_y

			# if the mouse is not in this area
			if not ((LookingGlassAddon.mouse_window_x >= context.area.x and LookingGlassAddon.mouse_window_x <= context.area.x + context.area.width) and (LookingGlassAddon.mouse_window_y >= context.area.y and LookingGlassAddon.mouse_window_y <= context.area.y + context.area.height)):
				return {'PASS_THROUGH'}

			# mouse position in current context region
			LookingGlassAddon.mouse_region_x = event.mouse_x - context.region.x
			LookingGlassAddon.mouse_region_y = event.mouse_y - context.region.y

			# for 3D viewports
			if context.space_data.type == "VIEW_3D":

				# if the settings are to be taken from device selection AND a device is active
				if context.scene.addon_settings.render_use_device == True and pylio.DeviceManager.get_active() is not None:

					# currently selected device
					device = pylio.DeviceManager.get_active()

				else:

					# make the emulated device the active device, if one was found
					device = pylio.DeviceManager.get_device(key='index', value=int(context.scene.addon_settings.render_device_type))

				# get the viewport block
				block = LookingGlassAddon.ViewportBlockRenderer.get_viewport_block()
				if block and device:

					# infer the current view
					if context.region.type != 'HEADER' and block.is_mouse_over(context):

						# for left positioned blocks
						if context.scene.addon_settings.viewport_block_alignment == 'left':
							# calculate the view number and angle
							view = floor(block.qs[block.preset]["total_views"] * (1 - (LookingGlassAddon.mouse_region_x - block.x) / block.width))
							angle = -floor(device.viewCone * ((LookingGlassAddon.mouse_region_x - block.x) / block.width - 0.5))

						# for right positioned blocks
						elif context.scene.addon_settings.viewport_block_alignment == 'right':

							# calculate the view number and angle
							view = floor(block.qs[block.preset]["total_views"] * (1 - (LookingGlassAddon.mouse_region_x - (context.area.width - block.width - block.x)) / block.width))
							angle = -floor(device.viewCone * ((LookingGlassAddon.mouse_region_x - (context.area.width - block.width - block.x - 50)) / block.width - 0.5))

						# update state variables
						block.set_active(True)
						block.set_device(device)
						block.set_view(view)
						block.set_angle(angle)

						# redraw region
						context.area.tag_redraw()

						return {'PASS_THROUGH'}

			# for Image Editors
			elif context.space_data.type == "IMAGE_EDITOR":

				# get the selected device
				device = pylio.DeviceManager.get_device(key='index', value=int(context.scene.addon_settings.imageeditor_block_device_type))

				# get the image
				image = context.space_data.image

				# get the image editor block
				block = LookingGlassAddon.ImageBlockRenderer.get_imageeditor_block()
				if block and device and image:

					# get image name
					image_name = image.name
					preset, quilt_format = LookingGlassAddon.ImageBlockRenderer.detect_from_quilt_suffix(context, image_name)
					if preset:

						# set the preset
						block.set_preset(preset)

						# update dimensions
						block.set_dimensions(int(device.width * context.space_data.zoom[0]), int(device.height * context.space_data.zoom[1]))

						# infer the current view
						if context.region.type != 'HEADER' and block.is_mouse_over(context):

							# calculate the view number and angle
							view = floor(block.qs[preset]["total_views"] * (1 - (LookingGlassAddon.mouse_region_x - block.x) / block.width))
							angle = -floor(device.viewCone * ((LookingGlassAddon.mouse_region_x - block.x) / block.width - 0.5))

							# update state variables
							block.set_device(device)
							block.set_view(view)
							block.set_angle(angle)

							# redraw region
							context.area.tag_redraw()

							return {'PASS_THROUGH'}

			# otherwise show centered view as default
			if block and device and block.preset:

				# if the mouse is outside the area use the center view
				view = floor(block.qs[block.preset]["total_views"] / 2)
				angle = floor(0)

				# update state variables
				block.set_active(False)
				block.set_device(device)
				block.set_view(view)
				block.set_angle(angle)

				# redraw region
				context.area.tag_redraw()

			return {'PASS_THROUGH'}



	# ------------ BLOCK RENDERER -------------
	# Inititalize the block renderer
	def __init__(self):

		# renderer type and state
		self.__type = None
		self.__is_running = False

		# drawing handlers
		self.__block_draw_view_view3d_handler = None
		self.__block_draw_block_view3d_handler = None
		self.__block_depsgraph_handler = None
		self.__block_frame_change_handler = None
		self.__block_draw_view_imageeditor_handler = None
		self.__block_draw_block_imageeditor_handler = None
		self.__block_imageeditor_autodetected = False

		# private class properties
		self.__blocks = {}
		self.__last_id = 0

		# active blocks
		self.__viewport3d_block = None
		self.__imageeditor_block = None

		# active presets
		self.__viewport3d_preset = 0
		self.__imageeditor_preset = 0

		# context override
		self.__override = None



	# deinititalize the block renderer
	def __del__(self):

		# if this is for the viewport
		if self.__type == 'VIEW_3D':

			# remove the draw handlers
			if self.__block_draw_view_view3d_handler:
				bpy.types.SpaceView3D.draw_handler_remove(self.__block_draw_view_view3d_handler, 'WINDOW')
				self.__block_draw_view_view3d_handler = None

			if self.__block_draw_block_view3d_handler:
				bpy.types.SpaceView3D.draw_handler_remove(self.__block_draw_block_view3d_handler, 'WINDOW')
				self.__block_draw_block_view3d_handler = None

		# if this is for the viewport
		if self.__type == 'IMAGE_EDITOR':

			# remove the draw handlers
			if self.__block_draw_view_imageeditor_handler:
				bpy.types.SpaceImageEditor.draw_handler_remove(self.__block_draw_view_imageeditor_handler, 'WINDOW')
				self.__block_draw_view_imageeditor_handler = None

			if self.__block_draw_block_imageeditor_handler:
				bpy.types.SpaceImageEditor.draw_handler_remove(self.__block_draw_block_imageeditor_handler, 'WINDOW')
				self.__block_draw_block_imageeditor_handler = None

		# free all blocks
		for id, block in self.__blocks.items():
			block.free()



	# status of the block renderer
	def is_running(self):
		return self.__is_running



	# start the block renderer
	def start(self, context):

		# if the renderer is NOT running
		if not self.is_running():

			# renderer type
			self.__type = context.space_data.type

			# create a context override
			self.__override = ContextOverride(context)

			# if this is for the viewport
			if self.__type == 'VIEW_3D':

				# setup the viewport block
				self.add_block(0, 10, 10, 420, 560)
				self.set_viewport_block(0)

				# add depsgraph update handler to react to scene changes
				self.__block_depsgraph_handler = bpy.app.handlers.depsgraph_update_post.append(self.__depsgraph_changes)

				# add update handler to react to frame changes
				self.__block_frame_change_handler = bpy.app.handlers.frame_change_post.append(self.__depsgraph_changes)

				# add draw handler to display the frustum of the Looking Glass camera
				# after everything else has been drawn in the view
				self.__block_draw_view_view3d_handler = bpy.types.SpaceView3D.draw_handler_add(self.__viewport_render_view, (context,), 'WINDOW', 'POST_PIXEL')
				self.__block_draw_block_view3d_handler = bpy.types.SpaceView3D.draw_handler_add(self.__viewport_render_block, (context,), 'WINDOW', 'POST_PIXEL')

			# if this is for the viewport
			if self.__type == 'IMAGE_EDITOR':

				# setup the image editor block
				self.add_block(0, 0, 0, 420, 560)
				self.set_imageeditor_block(0)

				# add draw handler to display the frustum of the Looking Glass camera
				# after everything else has been drawn in the view
				self.__block_draw_view_imageeditor_handler = bpy.types.SpaceImageEditor.draw_handler_add(self.__imageeditor_render_view, (context,), 'WINDOW', 'PRE_VIEW')
				self.__block_draw_block_imageeditor_handler = bpy.types.SpaceImageEditor.draw_handler_add(self.__imageeditor_render_block, (context,), 'WINDOW', 'POST_PIXEL')

			# update status
			self.__is_running = True

			# log info
			LookingGlassAddonLogger.info("Hologram preview started.")

		# keep the modal operator running
		return (self.__block_draw_block_view3d_handler, self.__block_draw_view_imageeditor_handler, self.__block_draw_block_imageeditor_handler)

	# stop the block renderer
	def stop(self):

		# if the renderer is running
		if self.is_running():

			# free all blocks
			for id, block in self.__blocks.items():
				block.free()

			# if this is for the viewport
			if self.__type == 'VIEW_3D':

				# remove the draw handlers
				if self.__block_draw_view_view3d_handler:
					bpy.types.SpaceView3D.draw_handler_remove(self.__block_draw_view_view3d_handler, 'WINDOW')
					self.__block_draw_view_view3d_handler = None

				# remove the draw handlers
				if self.__block_draw_block_view3d_handler:
					bpy.types.SpaceView3D.draw_handler_remove(self.__block_draw_block_view3d_handler, 'WINDOW')
					self.__block_draw_block_view3d_handler = None

			# if this is for the viewport
			if self.__type == 'IMAGE_EDITOR':

				if self.__block_draw_view_imageeditor_handler:
					bpy.types.SpaceImageEditor.draw_handler_remove(self.__block_draw_view_imageeditor_handler, 'WINDOW')
					self.__block_draw_view_imageeditor_handler = None

				if self.__block_draw_block_imageeditor_handler:
					bpy.types.SpaceImageEditor.draw_handler_remove(self.__block_draw_block_imageeditor_handler, 'WINDOW')
					self.__block_draw_block_imageeditor_handler = None

			# update status
			self.__is_running = False

			# log info
			LookingGlassAddonLogger.info("Hologram preview stopped.")

		# return None since this is expected by the operator
		return None

	# add a block to the block renderer
	def add_block(self, id, x, y, width, height):

		# add the block to the dictionary
		self.__blocks[id] = Block(x, y, width, height)

	# set the block to be rendered in the Blender viewport
	def set_viewport_block(self, id):

		# set the active VIEW3D block
		self.__viewport3d_block = id

		# get the block
		block = self.get_viewport_block()

		# if the settings are to be taken from device selection AND a device is active
		if bpy.context.scene.addon_settings.render_use_device == True and pylio.DeviceManager.get_active() is not None:

			# currently selected device
			block.set_device(pylio.DeviceManager.get_active())

			# set selected preset for this block
			block.set_preset(int(bpy.context.scene.addon_settings.quiltPreset))

		else:

			# set the selected emulated device as device for this block
			block.set_device(pylio.DeviceManager.get_device(key='index', value=int(bpy.context.scene.addon_settings.render_device_type)))

			# set selected preset for this block
			block.set_preset(int(bpy.context.scene.addon_settings.render_quilt_preset))

	# get the block to be rendered in the Blender viewport
	def get_viewport_block(self):
		if not self.__viewport3d_block is None:
			return self.__blocks[self.__viewport3d_block]

	# set the block to be rendered in the image editor
	def set_imageeditor_block(self, id):

		# set the active IMAGE_EDITOR block
		self.__imageeditor_block = id

		# if the settings are to be taken from device selection AND a device is active
		if bpy.context.scene.addon_settings.render_use_device == True and pylio.DeviceManager.get_active() is not None:

			# currently selected device
			self.__blocks[self.__imageeditor_block].set_device(pylio.DeviceManager.get_active())

		else:

			# set the selected emulated device as device for this block
			self.__blocks[self.__imageeditor_block].set_device(pylio.DeviceManager.get_device(key='index', value=int(bpy.context.scene.addon_settings.render_device_type)))

	# get the block to be rendered in the image editor
	def get_imageeditor_block(self):
		if not self.__imageeditor_block is None:
			return self.__blocks[self.__imageeditor_block]

	# check if the quilt and device format for the image editor have been
	# automatically detected
	def is_imageeditor_detected(self):
		if not self.__imageeditor_block is None:
			return self.__block_imageeditor_autodetected


	# Application handler that continously checks for changes of the depsgraph
	def __depsgraph_changes(self, scene, depsgraph):

		# if the block renderer is actively rendering
		if scene.addon_settings.viewport_block_show:

			# if something in the scene has changed
			if len(depsgraph.updates.values()) > 0:

				# select current viewport block
				block = self.get_viewport_block()
				if block is None or block.preset is None or block.offscreen_view is None:
					return

				# update the blocks status variable
				block.changed = True


	# render the view for the block
	def __viewport_render_view(self, context):

		# if a camera is selected AND the space is not in camera mode AND
		# the block viewport preview shall be drawn
		if self and context:
			if hasattr(context.scene, "addon_settings") and context.scene.addon_settings.lookingglassCamera in [obj for obj in context.view_layer.objects] and context.scene.addon_settings.viewport_block_show:
				if (context.space_data != None):

					# if the cycles render engine is active in this viewport
					if context.space_data.shading.type == 'RENDERED' and context.engine == 'CYCLES':
						return

					# select correct block
					block = self.get_viewport_block()
					if block is None or block.preset is None or block.offscreen_view is None:
						return

					# currently selected device
					device = block.device

					# if the device exists
					if device:

						# update dimensions
						if device.aspect < 1: block.set_dimensions(int(sqrt(context.area.width * context.area.height) * context.scene.addon_settings.viewport_block_scaling_factor * device.aspect), int(sqrt(context.area.width * context.area.height) * context.scene.addon_settings.viewport_block_scaling_factor))
						if device.aspect >= 1: block.set_dimensions(int(sqrt(context.area.width * context.area.height) * context.scene.addon_settings.viewport_block_scaling_factor), int(sqrt(context.area.width * context.area.height) * context.scene.addon_settings.viewport_block_scaling_factor / device.aspect))

					# if the block changed
					if block.has_changed(context):

						# select camera that belongs to the view
						camera = context.scene.addon_settings.lookingglassCamera

						# PREPARE THE MODELVIEW AND PROJECTION MATRICES
						# if a camera is selected
						if camera != None:

							# copy context
							context_copy = context.copy()

							# set the context for the override
							self.__override.set_context(context)

							# update all viewport shading and overlay settings
							self.__override.updateViewportSettings(context_copy['space_data'], force_context_data=True)

							# get camera's modelview matrix
							view_matrix = camera.matrix_world.copy()

							# correct for the camera scaling
							view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.x, 4, (1, 0, 0))
							view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.y, 4, (0, 1, 0))
							view_matrix = view_matrix @ Matrix.Scale(1/camera.scale.z, 4, (0, 0, 1))

							# calculate the inverted view matrix because this is what the draw_view_3D function requires
							camera_view_matrix = view_matrix.inverted_safe()

							# get the camera's projection matrix
							camera_projection_matrix = camera.calc_matrix_camera(
									depsgraph=context.view_layer.depsgraph,
									x = block.qs[block.preset]["view_width"],
									y = block.qs[block.preset]["view_height"],
									scale_x = 1.0,
									scale_y = (block.qs[block.preset]["rows"] / block.qs[block.preset]["columns"]) / block.aspect,
								)


							# RENDER THE VIEW
							# ++++++++++++++++++++++++++++++++++++++++++++++++
							with block.offscreen_view.bind():

								# calculate the offset-projection of the current view
								view_matrix, projection_matrix = self.__override.setupVirtualCameraForView(block.view, block.qs[block.preset]["total_views"], block.view_cone, block.aspect, camera_view_matrix.copy(), camera_projection_matrix.copy())

								# draw the viewport rendering to the offscreen for the current view
								block.offscreen_view.draw_view3d(
									# we use the "Scene" and the "View Layer" that is active in the Window
									# the user currently works in
									scene=context.scene,
									view_layer=context.view_layer,
									view3d=context.space_data,
									region=context.region,
									view_matrix=view_matrix,
									projection_matrix=projection_matrix,
									do_color_management = False)

						# restore all viewport shading and overlay settings
						self.__override.restoreViewportSettings()

						# reset status variable
						block.changed = False



	# render the block into the viewport
	def __viewport_render_block(self, context):

		# if a camera is selected AND the space is not in camera mode AND
		# the block viewport preview shall be drawn
		if self and context:
			if hasattr(context.scene, "addon_settings") and context.scene.addon_settings.lookingglassCamera in [obj for obj in context.view_layer.objects] and context.scene.addon_settings.viewport_block_show:
				if (context.space_data != None):

					# if the cycles render engine is active in this viewport
					if context.space_data.shading.type == 'RENDERED' and context.engine == 'CYCLES':
						return

					# select correct block
					block = self.get_viewport_block()
					if block is None or block.preset is None or block.offscreen_view is None:
						return

					# update the block
					block.update(context)

					#gpu.state.depth_test_set('LESS_EQUAL')
					gpu.state.depth_mask_set(True)
					gpu.state.blend_set('ALPHA')

					# if the block is drawn in a SpaceView3D
					if context.space_data.type == 'VIEW_3D' and context.space_data.region_3d != None:

						if context.scene.addon_settings.viewport_block_alignment == 'left': draw_texture_2d(block.offscreen_canvas.texture_color, (block.x, block.y), block.width, block.height)
						if context.scene.addon_settings.viewport_block_alignment == 'right': draw_texture_2d(block.offscreen_canvas.texture_color, (context.area.width - block.width - block.x - 50, block.y), block.width, block.height)

					gpu.state.depth_mask_set(False)
					gpu.state.blend_set('NONE')
					#print("blend_set")


	def detect_from_quilt_suffix(self, context, quilt_name):
		import re

		# select correct block
		block = self.get_imageeditor_block()

		# values from the metadata
		columns = None
		rows = None
		aspect = None

		# if a quilt name was given
		if quilt_name:

			# try to extract some metadata information from the quiltname
			try:

				columns = int(re.search('_qs(\d+)x(\d+)a(\d+.?\d*)', quilt_name).group(1))
				rows = int(re.search('_qs(\d+)x(\d+)a(\d+.?\d*)', quilt_name).group(2))
				aspect = float(re.search('_qs(\d+)x(\d+)a(\d+.?\d*)', quilt_name).group(3))

			except AttributeError:

				try:

					columns = int(re.search('_qs(\d+)x(\d+).', quilt_name).group(1))
					rows = int(re.search('_qs(\d+)x(\d+).', quilt_name).group(2))

				except AttributeError:

					pass

			# for each supported quilt format
			for preset, qf in pylio.LookingGlassQuilt.formats.get().items():

				# if the quilt
				if not (columns is None or rows is None or aspect is None) and columns == qf['columns'] and rows == qf['rows']:

					# update status variable
					self.__block_imageeditor_autodetected = True

					# update UI
					if context.scene.addon_settings.imageeditor_block_quilt_preset != str(preset):
						context.scene.addon_settings.imageeditor_block_quilt_preset = str(preset)

					return preset, qf

		# return preset and format from the settings
		self.__block_imageeditor_autodetected = False
		return int(context.scene.addon_settings.imageeditor_block_quilt_preset), block.qs[int(context.scene.addon_settings.imageeditor_block_quilt_preset)]


	# render the view into the offscreen
	def __imageeditor_render_view(self, context):

		# if a camera is selected AND the space is not in camera mode
		if self and context:
			if hasattr(context.scene, "addon_settings"):
				if (context.space_data != None and context.space_data.type == 'IMAGE_EDITOR'):

					# get the image loaded in the image editor
					image = context.space_data.image
					if not image is None and image.type != "RENDER_RESULT":

						# select correct block
						block = self.get_imageeditor_block()

						# get the selected device
						device = pylio.DeviceManager.get_device(key='index', value=int(context.scene.addon_settings.imageeditor_block_device_type))

						# get quilt preset and format parameters
						preset, quilt_format = self.detect_from_quilt_suffix(context, image.name)
						if quilt_format:

							# update dimensions
							block.set_dimensions(int(device.width * context.space_data.zoom[0]), int(device.height * context.space_data.zoom[1]))

						# if the block mode is activated
						if context.scene.addon_settings.imageeditor_block_show:

							# if the block changed
							if block.changed and block.offscreen_view:

								# create a texture from the image
								#if not block.image_texture:
								block.image_texture = gpu.texture.from_image(image)

								# if the texture was created
								if block.image_texture:

									# calculate view grid indices in the quilt
									view_ix = block.view % block.qs[block.preset]['columns']
									view_iy = floor(block.view / block.qs[block.preset]['columns'])

									# RENDER THE VIEW
									# ++++++++++++++++++++++++++++++++++++++++++++++++
									with block.offscreen_view.bind():

										# get the current projection matrix
										viewMatrix = gpu.matrix.get_model_view_matrix()
										projectionMatrix = gpu.matrix.get_projection_matrix()

										with gpu.matrix.push_pop():

											# reset matrices -> use normalized device coordinates [-1, 1]
											gpu.matrix.load_matrix(Matrix.Identity(4))
											gpu.matrix.load_projection_matrix(Matrix.Identity(4))

											gpu.state.depth_test_set('GREATER_EQUAL')
											gpu.state.depth_mask_set(True)
											gpu.state.blend_set('ALPHA')

											# draw the viewport rendering to the offscreen for the current view
											draw_texture_2d(block.image_texture, (-1 - 2 * view_ix, -1 - 2 * view_iy), 2 * block.qs[block.preset]['columns'], 2 * block.qs[block.preset]['rows'])

											gpu.state.blend_set('NONE')
											gpu.state.depth_mask_set(False)
											gpu.state.depth_test_set('NONE')

											# reload original matrices
											gpu.matrix.load_matrix(viewMatrix)
											gpu.matrix.load_projection_matrix(projectionMatrix)

								# reset status variable
								block.changed = False

								# update the block
								block.update(context)

							# convert view2d coordinates of the image center to
							# region corrdinates
							view_2d_x, view_2d_y = context.region.view2d.view_to_region(0.5, 0.5)

							# center the block
							block.x = view_2d_x - block.width / 2
							block.y = view_2d_y - block.height / 2

	# render the block into the image editor
	def __imageeditor_render_block(self, context):

		# if a camera is selected AND the space is not in camera mode
		if self and context:
			if hasattr(context.scene, "addon_settings") and context.scene.addon_settings.imageeditor_block_show:
				if not context.space_data is None:

					# get the image loaded in the image editor
					image = context.space_data.image
					if not image is None and image.type != "RENDER_RESULT":

						# select correct block
						block = self.get_imageeditor_block()
						if block:

							# get active frame buffer
							framebuffer = gpu.state.active_framebuffer_get()
							with framebuffer.bind():

								# clear the framebuffer
								c = context.preferences.themes[0].image_editor.grid
								framebuffer.clear(color=(c[0] * 0.15, c[1] * 0.15, c[2] * 0.15, c[3]))

								gpu.state.depth_mask_set(True)
								gpu.state.blend_set('ALPHA')

								# for Blender versions prior to BGL deprecation
								if bpy.app.version < (3, 4, 0):
								    gpu.state.depth_test_set('GREATER_EQUAL')

								# draw the block
								draw_texture_2d(block.offscreen_canvas.texture_color, (block.x, block.y), block.width, block.height)

								gpu.state.blend_set('NONE')
								gpu.state.depth_mask_set(False)

								# for Blender versions prior to BGL deprecation
								if bpy.app.version < (3, 4, 0):
									gpu.state.depth_test_set('NONE')