@@ -17,3 +17,1 @@
-Unlike Cycles, EEVEE is not a raytrace render engine.
-Instead of computing each ray of light, EEVEE uses a process called rasterization.
-Rasterization estimates the way light interacts with objects and materials using numerous algorithms.
+EEVEE is a based on rasterization and is not a path tracer.

@@ -18,2 +17,2 @@
-Instead of computing each ray of light, EEVEE uses a process called rasterization.
-Rasterization estimates the way light interacts with objects and materials using numerous algorithms.
+EEVEE is a based on rasterization and is not a path tracer.
+Instead of computing each ray of light, EEVEE uses a process called rasterization to determine what

@@ -20,3 +21,3 @@
 While EEVEE is designed to use :abbr:`PBR (Physically Based Rendering)` principles,
 it is not perfect and Cycles will always provide more physically accurate renders.
-Because EEVEE uses rasterization it has a large set of :doc:`limitations </render/eevee/limitations>`.
+Because EEVEE uses a lot of different algorithms it has a set of

@@ -8,2 +7,3 @@
 
-There are three different probe types. One for diffuse lighting, two for specular lighting.
+There are three different types of light probes.
+Each type of light probe record the lighting at different resolution and frequency.

@@ -9,1 +8,3 @@
-There are three different probe types. One for diffuse lighting, two for specular lighting.
+There are three different types of light probes.
+Each type of light probe record the lighting at different resolution and frequency.
+They are used together to recover incoming light information when using ray tracing is not possible (either for performance or for technical limitations).

@@ -0,0 +8,4 @@
+
+This type of light probe is suited to smooth planar surfaces.
+
+Each visible planar light probes increases the render time as the scene needs to be rendered for

@@ -0,0 +11,4 @@
+Each visible planar light probes increases the render time as the scene needs to be rendered for
+each of them.
+
+Light probe planes only work when ray tracing method is set to `Screen-Trace`. When enabled, they

@@ -0,0 +16,4 @@
+
+.. note::
+
+   Reflections and Volumetrics are not supported inside Light probe planes.

@@ -0,0 +34,4 @@
+
+Distance
+   A probe object only influences the lighting of surfaces inside its influence zone.
+   This influence zone is defined by the Distance parameter and object scaling.

@@ -0,0 +36,4 @@
+   A probe object only influences the lighting of surfaces inside its influence zone.
+   This influence zone is defined by the Distance parameter and object scaling.
+
+   For light probe planes the influence distance is the distance from the plane.

@@ -0,0 +44,4 @@
+=======
+
+Clipping Offset
+   Define how much below the plane the near clip is when capturing the scene.

@@ -0,0 +6,4 @@
+A light probe sphere records the light incomming from many directions at a single location.
+
+They are used for smooth and semi-rough reflections.
+They then smoothly blend to light probe volume lighting for completely diffuse reflections.

@@ -0,0 +8,4 @@
+They are used for smooth and semi-rough reflections.
+They then smoothly blend to light probe volume lighting for completely diffuse reflections.
+
+If *Raytracing* is turned on, they are used as a fall back if a ray misses.

@@ -0,0 +12,4 @@
+
+.. note::
+
+   In both usage, the light probe spheres are shadowed by light probe volume. This is to

@@ -0,0 +20,4 @@
+
+   :doc:`Scene </render/eevee/scene>`.
+
+The world also have an internal light probe sphere which resolution can be adjusted in the *World* data panel.

@@ -0,0 +31,4 @@
+   :Panel:     :menuselection:`Object Data --> Probe`
+
+Type
+   Select the shape of the influence volume. Can be set to Sphere or Box.

@@ -0,0 +38,4 @@
+   This influence zone is defined by the size parameter and object scaling.
+
+Falloff
+   Percentage of the influence distance during which the influence of a probe fades linearly.

@@ -0,0 +59,4 @@
+   :Panel:     :menuselection:`Object Data --> Custom Parallax`
+
+By default, the influence volume is also the parallax volume.
+The parallax volume is a volume on which the recorded light is projected.

@@ -0,0 +62,4 @@
+The parallax volume is a volume on which the recorded light is projected.
+It should roughly fit it surrounding area. In some cases it may be better to
+adjust the parallax volume without touching the influence parameters.
+In this case, just enable the *Custom Parallax* and

@@ -0,0 +3,4 @@
+Light Probe Volume
+******************
+
+A volume probe records the light incomming from all directions at a many locations inside a volume.

@@ -0,0 +41,4 @@
+   Can lead to view dependant result if set too high. Prefer this if camera is static.
+
+Facing Bias
+   When set to zero, avoids capture points behind the shaded surface to bleed light onto

@@ -0,0 +42,4 @@
+
+Facing Bias
+   When set to zero, avoids capture points behind the shaded surface to bleed light onto
+   the shaded surface, but this produces non-smooth interpolation when the capture resolution is high.

@@ -0,0 +50,4 @@
+===================
+
+During the baking process, a validity score is assigned to each capture point.
+This score is based on the number of back-face was hit when capturing the incoming lighting.

@@ -0,0 +75,4 @@
+   Baking uses the render visibility of the objects in the scene.
+
+During baking, the scene is converted into a different representation to accelerate light transport.
+This representation can be very memory intensive and prevent baking if it cannot fit inside the GPU memory.

@@ -0,0 +77,4 @@
+During baking, the scene is converted into a different representation to accelerate light transport.
+This representation can be very memory intensive and prevent baking if it cannot fit inside the GPU memory.
+There are a few way to deal with this issue:
+  - Larger scene should be divided into smaller sections or using different level of details.

@@ -0,0 +88,4 @@
+Resolution
+   Spatial resolution for volumetric light probes is determined per probe.
+   The local volume is divided into a regular grid of the specified dimensions.
+   Irradiance light sample will be computed for each cell in this grid.

@@ -0,0 +96,4 @@
+
+Surfel Resolution
+   Number of surfels to spawn in one local unit distance.
+   Higher values increase quality. Have a huge impact on memory usage.

@@ -0,0 +123,4 @@
+
+Clamp Direct
+   Clamp incoming direct light. 0.0 disables direct light clamping.
+   Here direct light refers to the light that bounces only once (from light object)

@@ -0,0 +124,4 @@
+Clamp Direct
+   Clamp incoming direct light. 0.0 disables direct light clamping.
+   Here direct light refers to the light that bounces only once (from light object)
+   or coming from emissive materials.

@@ -0,0 +139,4 @@
+Offset
+======
+
+In order to reduce artifacts caused by bad capture points positioning, the bake process will adjust their location.

@@ -0,0 +140,4 @@
+======
+
+In order to reduce artifacts caused by bad capture points positioning, the bake process will adjust their location.
+First by moving them slightly away from surrounding surfaces, avoiding banding artifacts.

@@ -0,0 +141,4 @@
+
+In order to reduce artifacts caused by bad capture points positioning, the bake process will adjust their location.
+First by moving them slightly away from surrounding surfaces, avoiding banding artifacts.
+It also tries to move them out of objects if they are not so far bellow the surface.

@@ -0,0 +8,4 @@
+   :Panel:     :menuselection:`Properties --> Light`
+               :menuselection:`Shader Editor --> Sidebar --> Options`
+
+Next to lighting from the background and any object with an emission shader,

@@ -0,0 +24,4 @@
+only where it is needed. It also includes a very efficient caching mechanism.
+
+.. tip::
+   - The error message *Shadow buffer full* means that the system cannot allocate.

@@ -0,0 +45,4 @@
+.. _bpy.types.Light.shadow_filter_radius:
+
+Filter
+   Blur shadow aliasing using Percentage Closer Filtering with a circular kernel.

@@ -0,0 +52,4 @@
+
+Resolution Limit
+   Minimum size of a shadow map pixel. Higher values use less memory at the cost of shadow quality.
+   Higher values also speed-up rendering of heavy scene.

@@ -0,0 +25,4 @@
+Lights
+======
+
+- For now lights can only have one color and do not support light node trees.

@@ -0,0 +57,4 @@
+      This only modifies the shading normal of the object. Vertex position is not affected.
+
+   :Displacement Only:
+      This mode is not supported and fallback to *Displacement and Bump*.

@@ -0,0 +117,4 @@
+---------------
+
+When using *Blended* render method, the order in which the color blending happens is important as it
+can change the final output color. As of now EEVEE does not support per-fragment (pixel) sorting or per-triangle sorting.

@@ -0,0 +122,4 @@
+Opaque surfaces (i.e. that have no transparency) will still have correct sorting regardless of the render method.
+
+.. tip::
+   Face order can be adjusted in edit mode by using :doc:`sort element </modeling/meshes/editing/mesh/sort_elements>`.

@@ -0,0 +127,4 @@
+.. note::
+
+   Per-object sorting has a performance cost and having thousands of
+   these objects in a scene will greatly degrade performance.

@@ -0,0 +144,4 @@
+This feature is used to approximate the inner geometry structure of the object without heavy computation.
+This is currently used for Subsurface, Translucent BSDF, Refraction BSDF and the nodes containing them.
+
+If no value is plugged into the output node, then a default thickness based on the smallest dimension of the object is computed.

@@ -0,0 +150,4 @@
+
+.. note::
+   The thickness is used to skip the inner part of the object. This means that:
+   - refraction will skip objects inside this thickness distance.

@@ -0,0 +22,4 @@
+
+
+Indirect Light
+   Similar to **Direct Light* but limits the maximum light intensity reflected using ray-tracing and light-probes.

@@ -0,0 +26,4 @@
+
+.. note::
+
+    These options provides a way to limit :term:`Fireflies` and :term:`Aliasing` of highly reflective surfaces and dense volumes.

@@ -15,0 +26,4 @@
+.. _bpy.types.SceneEEVEE.gi_irradiance_pool_size:
+
+Light Probes Volume Pool Size
+   A bigger pool size allows for more irradiance grid in the scene but might not fit into GPU memory and decreases performance.

@@ -0,0 +22,4 @@
+
+   :Light Probe:
+      Use light-probe spheres and planes to find scene intersection.
+      This option is has the lowest tracing cost but rely on manually placed light-probes.

@@ -0,0 +44,4 @@
+They are only visible if *Screen-Trace* is the chosen tracing *Method*.
+
+Precision
+   Higher values increase precision of the screen space ray-tracing but lowers the maximum trace distance.

@@ -0,0 +55,4 @@
+Denoising
+_________
+
+Denoising can be enabled to reduce the noise amount from the raw ray-traced output.

@@ -0,0 +63,4 @@
+   Can introduce some light leaks across surfaces.
+
+Temporal Accumulation
+   Accumulate samples by re-projecting last ray tracing results.

@@ -0,0 +74,4 @@
+Fast GI Approximation
+_____________________
+
+This Fast GI Approximation is a fallback to the ray-tracing pipeline for

@@ -0,0 +76,4 @@
+
+This Fast GI Approximation is a fallback to the ray-tracing pipeline for
+:abbr:`BSDF (Bidirectional Scattering Distribution Function)` with high roughness.
+It produces less noisy output and capture bounce lighting more efficiently than individually traced rays.

@@ -0,0 +78,4 @@
+:abbr:`BSDF (Bidirectional Scattering Distribution Function)` with high roughness.
+It produces less noisy output and capture bounce lighting more efficiently than individually traced rays.
+
+This is currently implemented as a screen space effect and will inherit.

@@ -0,0 +83,4 @@
+.. seealso:: :ref:`Limitations <eevee-limitations-screenspace>`.
+
+Method
+   Determine the methods used to compute the fast GI approximation.

@@ -0,0 +97,4 @@
+   Lower options will be faster and use less memory but will produce blurrier results.
+
+Ray Count
+   Amount of GI ray per pixels at the specified *Resolution*.

@@ -0,0 +101,4 @@
+   Higher values will reduce noise.
+
+Step Count
+   Amount of screen sample per GI ray.

@@ -0,0 +106,4 @@
+
+   .. note::
+
+      With higher step count, there is less chance to miss other surfaces that could reflect of block the light.

@@ -25,0 +41,4 @@
+=======
+
+Shadow Rays Count
+   Amount of shadow rays to trace for each light.

@@ -0,0 +23,4 @@
+   These samples are distributed along the view depth (view Z axis).
+
+Distribution
+   Blend between linear and exponential sample distribution. Higher values puts more samples near the camera.

@@ -0,0 +33,4 @@
+Custom Range
+============
+
+When working with volumes objects, EEVEE automatically compute the best depth range where to compute

@@ -0,0 +35,4 @@
+
+When working with volumes objects, EEVEE automatically compute the best depth range where to compute
+the volume sampling and lighting.
+In certain situation, this isn't enough and produces sub-optimal sampling which increases noise.

@@ -0,0 +36,4 @@
+When working with volumes objects, EEVEE automatically compute the best depth range where to compute
+the volume sampling and lighting.
+In certain situation, this isn't enough and produces sub-optimal sampling which increases noise.
+This is particularly the case when using a volume shader inside the *World* or when working with large of volume objects.

@@ -0,0 +6,4 @@
+The world environment can emit light, ranging from a single solid color
+to arbitrary textures.
+
+In EEVEE, the world lighting contribution is stored into internal :ref:`Light Probe`.

@@ -0,0 +83,4 @@
+Sun
+---
+
+EEVEE can separate the light from intense light sources (e.g. a sun from an outdoor HDRI) and

@@ -0,0 +84,4 @@
+---
+
+EEVEE can separate the light from intense light sources (e.g. a sun from an outdoor HDRI) and
+replace them by a sun light. This increases the quality of the lighting as the internal light probes

@@ -0,0 +95,4 @@
+
+Angle
+   Angle of the extracted sun light.
+   Angular diameter of the Sun as seen from the Earth