Area emitter (default)

Overview

This tutorial sets up a very simple scene with an area light and shows the resulting image/video for two different angles/points of view.

🚀 You will learn how to:

• Use the angulararea plugin

The two tutorials render_angular_1light.ipynb and render_area_1light.ipynb show the same scene. The first one uses an angulararea emitter (our new plugin), and the second one uses the area emitter (Mitsuba 3’s default).

You should use these tutorials to compare the effect of one light vs. the other and decide which one suits best your use case.

# If you have compiled Mitsuba 3 yourself, you will need to specify the path
# to the compilation folder
# import sys
# sys.path.insert(0, '<mitsuba-path>/mitsuba3/build/python')
import mitsuba as mi
# To set a variant, you need to have set it in the mitsuba.conf file
# https://mitsuba.readthedocs.io/en/latest/src/key_topics/variants.html
mi.set_variant('llvm_ad_rgb')

import mitransient as mitr

print('Using mitsuba version:', mi.__version__)
print('Using mitransient version:', mitr.__version__)

Using mitsuba version: 3.6.4
Using mitransient version: 1.2.0

# Load XML file
import os
# You can also use mi.load_dict and pass a Python dict object
# but it is probably much easier for your work to use XML files
scene = mi.load_file(os.path.abspath('scenes/area_1light.xml'))


# Render the scene and develop the data
# Because mitsuba works with Dr.JIT, this cell will execute
# very fast. But when you actually read the variables
# it will actually start computing everyting (and take a while).
data_steady, data_transient = mi.render(scene, spp=256)

import matplotlib.pyplot as plt
import numpy as np

plt.imshow((data_steady / np.max(data_steady)) ** (1 / 4.0))
plt.show()

mitr.vis.show_video(
    (data_transient / np.max(data_transient)) ** (1 / 4.0),
    axis_video=2,
)

scene = mi.load_file(os.path.abspath('scenes/area_1light.xml'))
params = mi.traverse(scene)

params['sensor.to_world'] = params['sensor.to_world'].look_at(
    mi.ScalarPoint3f(0, 50, 10),  # origin
    mi.ScalarPoint3f(0, 0, 30),      # target
    mi.ScalarPoint3f(0, 0, 1)    # up
)
params.update()

# Render the scene and develop the data
# Because mitsuba works with Dr.JIT, this cell will execute
# very fast. But when you actually read the variables
# it will actually start computing everyting (and take a while).
data_steady, data_transient = mi.render(scene, spp=64)

import matplotlib.pyplot as plt
import numpy as np

plt.imshow((data_steady / np.max(data_steady)) ** (1 / 4.0))
plt.show()

mitr.vis.show_video(
    (data_transient / np.max(data_transient)) ** (1 / 4.0),
    axis_video=2,
)