[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Search] [Home]
Settings Home
/3/ - 3DCG

Thread archived.
You cannot reply anymore.

File: coatedrocks.png (390 KB, 503x320)
390 KB
390 KB PNG
Is messing with the reflection strength / weight setting of a renderer allowed at all, if you want to get physically accurate results?
Pic related seems way too shiny to me. (100% reflection weight, 90% roughness weight, IOR 1.3)
Am i allowed to turn it down if i want to achieve photo realistic results?
You can't crank it past 100% and stay 'physical based' since it'd be impossible for a surface to reflect more light than it recessives. Dialing it down is of course valid since surfaces can absorb and diffuse incoming light.

You probably wanna read up on what 'PBR' is and why it's becoming popular compared to legacy shading models to better understand what you can and can't do and why that is.

Typically when you write a PBR type shader it won't be possible for the user to enter any values that would violate physics.
File: 568679780780352.jpg (42 KB, 326x236)
42 KB
I've professionally worked on a lot of models in substance painter before, so i understand the general PBR workflow, but never had to mess with any of the variables.
Right now i'm messing around with c4d physical (for fun), which super biased and gives you tons of ways to mess up the photorealism. That answers my question btw. Thanks.
While we're on the topic of PBR does any tr/3/pot know why the diffuse texture is commonly called 'albedo' in PBR?

It redlines my inner autismo, in physics albedo is the sum of all the light emitted from an entire surface, like "how bright is the planet Jupiter in my telescope?".
I keep correcting it to the more descriptive term 'diffuse' in my shaders but people keep coming up to me and ask "herpa derp anon, here in this node don't u mean albedo?" and inside my head I be all like "DON'T U KNOW AL-BEAT-U?!?"
Since you mention reflection weight, I assume you are talking about the reflection pass contribution to the final render, not about the reflectivity of a surface. In this case, you generally want to keep reflection weight to 1.0 and play with roughness to determine how glossy a surface is (for rock this would be quite high, ~0.8, although bear in mind that rock may have shiny spots, so a custom texture may be needed). If you were to reduce glossiness by tuning down the reflection weight, you'd be tuning down the Fresnel effect, which would be physically inaccurate. Regarding Fresnel, you chose an IOR of 1.3, which is probably inappropriate for rock. Unless you are doing water (IOR = 1.33), start with IOR 1.5, or even look up actual indices for the material you are trying to reproduce.
I believe it is because, in PBR, "diffuse" textures are supposed to encode reflectivity, hence the proper name would be albedo. In short, a diffuse texture may have shadows/AO, while an albedo texture should never, on principle, have that kind of information.
Adjusting specular strength/roughness on a per-shot basis is perfectly fine. Think of how actors powder their faces or not depending on the lighting conditions.
File: Untitled45745.png (636 KB, 1151x590)
636 KB
636 KB PNG
This is where i'm at right now.
A physically accurate IOR (about 1.6?) at 100% reflection weight makes the rock look like someone coated it with car paint.
Can you post an example of how it looks with reflection strength at 100% and IOR at 1.6? Also, are you using a bump map instead of actual geo detail for the ground?
File: 4374696789.jpg (674 KB, 1280x720)
674 KB
674 KB JPG
Here's the whole room with the same settings as above but reflection strength at 100% and IOR 1.6.
The ground has a bump map and 2 cm of displacement. I disabled the displacement, and turned down the samples for pic related though, so the GI wouldn't take forever to calculate.
File: 2365856.jpg (159 KB, 1214x719)
159 KB
159 KB JPG
The material look fine when rendered with the same settings in Redshift.
I guess this is a problem with the way c4d physical calculates reflections then?
>I guess this is a problem with the way c4d physical calculates reflections then?
I'm not familiar with the Physical engine, but it appears as if it's ignoring the bump map when calculating the Fresnel reflection. Since Redshift works as one would expect, it seems like the problem is in fact with the render engine, or with the way it interprets the reflection parameters, which perhaps doesn't align with PBR principles. A separate reflection pass would help with finding the cause.
File: 4572346.png (2.1 MB, 1280x1439)
2.1 MB
2.1 MB PNG
Here's the reflection pass with and without displacement.
It really seems to be a renderer specific problem. I guess i'll work around it.
You can do whatever you want.

Reminder that "Specular IOR" in render engines is a meme and you can often get better results by adjusting it to tweak your edge falloff and overall specular level. These idealized BRDF models like GGX are way too inaccurate to try and pretend the IOR parameter in the render engine matches the physical one with any degree of accuracy. Just look at Figures 10 and 11 in this paper if you don't believe me. http://www.cs.cornell.edu/~srm/publications/EGSR07-btdf.pdf

The only time you need to worry about accurate IOR is for refraction, e.g. with glass lenses and caustics. In that case, the effect of IOR is the same as the physical term. However If your material is opaque all you're doing is modifying the specular lobe shape you should choose IOR with artistic liberty.
>"diffuse" textures are supposed to encode reflectivity
pretty much, this is the texture map you use when you bake everything into a texture.
I usually dont make a big fuss about IOR. I usually just have active shade on in 3Ds Max and adjust it accordingly. usually my opaque objects stay around 1.0 and if i wanted it to be super reflective id go less than that. im not exactly sure what going pass 1.0 does, then again i dont understand this angle of incidence thing. If i wanted there to be more refraction i would enable coating and adjust the IOR on there.

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.