Reviewed changes to path tracing documentation #5823
Conversation
|
Hi! This comment will help you figure out which jobs to run before merging your PR. The suggestions are dynamic based on what files you have changed. HDRP Depending on the scope of your PR, you may need to run more jobs than what has been suggested. Please speak to your lead or a Graphics SDET (#devs-graphics-automation) if you are unsure. |
|
It appears that you made a non-draft PR! |
|
@eturquin I changed your initial layout and notes so please check if I preserved your meaning. |
|  | ||
|
|
||
| Single-sided or no Transmission | ||
| Single-sided with no Transmission. |
There was a problem hiding this comment.
Should be "Single-sided, or no Transmission"
| | ---------------- | ------------------------------------------------------------ | ------------------------------------------------------------ | | ||
| | **Box** | A box-shaped model where incident light enters through a flat surface and leaves through a flat surface.<br/><br/>Suitable for thick objects like a paperweight or a crystal ball. | A double-sided Material that has its **Normal mode** set to **None**. | | ||
| | **Sphere** | A sphere-shaped model that produces a magnifying glass-like effect to refraction. <br/><br/>Suitable for thick objects like a paperweight or a crystal ball. | A double-sided Material that has its **Normal mode** set to **None**. | | ||
| | **Thin** | A thin box surface type, similar to **Box** with a fixed thickness of 5cm. Select this for thin window-like surfaces. <br/>The behaviour of the **Thin** refraction model does not change with path tracing enabled .<br/><br/>Suitable for thin objects like a soap bubble or a window. | A double-sided Material that has its **Normal mode** set to **Flip** or **Mirror**. | |
There was a problem hiding this comment.
| Thin | A thin surface type, characterized by infinitesimal thickness. Select this for thin window-like surfaces.
The behaviour of the Thin refraction model does not change with path tracing enabled .
Suitable for thin objects like a soap bubble or a glass window. | A double-sided Material that has its Normal mode set to Flip or Mirror. |
| | Refraction model | Path tracing behavior | Compatible Surface sides | | ||
| | ---------------- | ------------------------------------------------------------ | ------------------------------------------------------------ | | ||
| | **Box** | A box-shaped model where incident light enters through a flat surface and leaves through a flat surface.<br/><br/>Suitable for thick objects like a paperweight or a crystal ball. | A double-sided Material that has its **Normal mode** set to **None**. | | ||
| | **Sphere** | A sphere-shaped model that produces a magnifying glass-like effect to refraction. <br/><br/>Suitable for thick objects like a paperweight or a crystal ball. | A double-sided Material that has its **Normal mode** set to **None**. | |
There was a problem hiding this comment.
Box and Sphere are exactly the same in path tracing, they correspond to a thick mode. It's a bit confusing to have them separate in this table, as if they have different effects.
There was a problem hiding this comment.
I'll double-check with the editor about this. I know it seems obvious, but we don't want to confuse the user into thinking there is one Refraction model called "Box and Sphere"!
| | **Box** | A box-shaped model where incident light enters through a flat surface and leaves through a flat surface.<br/><br/>Suitable for thick objects like a paperweight or a crystal ball. | A double-sided Material that has its **Normal mode** set to **None**. | | ||
| | **Sphere** | A sphere-shaped model that produces a magnifying glass-like effect to refraction. <br/><br/>Suitable for thick objects like a paperweight or a crystal ball. | A double-sided Material that has its **Normal mode** set to **None**. | | ||
| | **Thin** | A thin box surface type, similar to **Box** with a fixed thickness of 5cm. Select this for thin window-like surfaces. <br/>The behaviour of the **Thin** refraction model does not change with path tracing enabled .<br/><br/>Suitable for thin objects like a soap bubble or a window. | A double-sided Material that has its **Normal mode** set to **Flip** or **Mirror**. | | ||
| | **None** | A thin refractive surface, hardcoded to be fully smooth to simulate alpha blending.<br/>The behaviour of the **None **refraction model does not change with path tracing enabled .<br/><br/>Suitable for thin objects like a soap bubble or a window. | A double-sided Material that has its **Normal mode** set to **Flip** or **Mirror**. | |
There was a problem hiding this comment.
I wouldn't say that this mode is suitable for anything really, it's just here to keep compatibility with raster mode when you switch to path tracing, but other than that thin should be prefered in every single situation.
| ## Materials parameterization | ||
|
|
||
| Some phenomena like refraction, absorption in transmissive objects, or subsurface scattering, can be expressed more naturally in a path tracing setting than in its rasterized counterpart, and as such require less material parameters (e.g. additional thickness information, expected shape of the refractive object, ...). For that reason, some parameters have no effect in path tracing, while others bear a slightly different meaning. | ||
| HDRP can perform effects like light refraction and absorption in transmissive objects, or subsurface scattering, more easily when you use path tracing setting than in rasterization . This means they need fewer Material parameters (e.g. additional thickness information, expected shape of the refractive object, ...). For that reason, some parameters have no effect in path tracing, and some parameters change behaviour. For more information, see [Refraction models](#). |
There was a problem hiding this comment.
It's not performing those effects "more easily" (as in, it's not computing them faster, or with less resources), but rather can express them "more naturally" given the different way of representing light transport simulation.
There was a problem hiding this comment.
Thanks for this! Is the following re-write more accurate?
"Path tracing performs light effects using a more realistic light transport simulation than rasterization. This means light refraction and absorption in transmissive objects and subsurface scattering appears more realistic."
There was a problem hiding this comment.
For the first couple lines, yes.
|
Thanks for your feedback @eturquin I'm implementing your suggested changes in a google doc so I can get it reviewed by docs editorial. |
Sounds good! :) |
|
any update? thanks |
This doc is currently in editorial review. |
|
Hi @Vic-Cooper, where are we at with editorial? Ideally we'd want to have this merged within the next week or so. |
3 participants
As per the changes made in #5551