V-Ray Dome Light with Affect Specular disabled, Max vs. Standalone?

[Echadwick]

If a V-Ray Dome Light has a texture, and the Affect Specular option is disabled, V-Ray 3ds Max will render reflections but Standalone will not.

Is this possibly a bug in the Max renderer, or maybe in Standalone?

Glass ball on the left, mirror ball on the right.

Max render:


Standalone render:


The Affect Specular option:


Example file:
vray_domelight_specular.zip

3ds Max 2020.3 (22.3.0.3165)
V-Ray 4.30.00 (build 00001)
V-Ray Standalone, version Next Standalone, update 3.2 for x64, Build b4e50a4e from Apr 8 2020, 23:18:02, V-Ray core version is 4.30.02

Thank you!

[nikoleta.garkova]

Hi Echadwick,

Not able to reproduce this, getting the same results in both Max and Standalone render. V-Ray versions are the same.
Can you give it another test? First result may have been with Affect specular enabled - let me know if that's the case.

[Echadwick]

Thanks for the test.

Yes it is disabled in both my renders. I will try again and arrange the screenshot the same as yours.

The help file says:

Affect specular – Determines whether the light affects the specular portion of the materials. The Multiplier value on the Intensity rollout controls the light's contribution to specular reflections.

Affect reflections – Specifies whether the light source appears in reflections.

How is Affect Specular different from Affect Reflection? What exactly does Specular control in a dome light? Only the specular highlights, the brightest parts of an HDR image like the sun? Or all reflectance contributions?

In your renders, it affects all reflectance, not just highlights?

[nikoleta.garkova]

You can refer to this post for an explanation on what the updated Affect Specular vs Affect Reflections do, where you can also find some visual examples.

[Echadwick]

Thanks for the link!

Can you provide more information about why this change was made? The change doesn't make sense to us.

Before 3.1, Affect Specular used to control the brightest parts of the HDRI in a Dome Light, and Affect Reflection controlled the rest of the Reflection contribution. You could control them independently, or disable all Reflection by disabling both.

Before:


After:


The only light source here is a V-Ray Dome Light, and the material is a perfect mirror, with varying RGlossiness.

I know we're not going to change current development, you can't stand in the way of progress. But the changes don't seem to match the parameter names, nor do they seem to improve functionality?

[Echadwick]

Is there a good method to disable the sun in a HDR image-based light, so we can use VRaySun instead of the HDR sun?

We prefer the controls available in VRaySun, specifically the Size Multiplier which affects shadow softness.

Disabling the Affect Specular option used to give us this control in the dome light. But with the update we're not able to figure out how to disable the HDR sun.

[kosso_olli]

Clamp the HDRI used in the domlight with a value of 5. This should eliminate all the strong lighting caused by the sun captured in the HDRI

[Echadwick]

Thanks for the suggestion. Unfortunately that greatly darkens the rest of the image at the same time, often seen in-camera in our shots.

A couple examples where the HDRI is visible in the background:
https://www.wayfair.com/outdoor/pdx/...000098093.html
https://www.wayfair.com/outdoor/pdx/...001376614.html

We could use an inverse curve I suppose, but that's going to be pretty inconsistent.

[^Lele^]

(optionally paint and) Blur the lighting probe, keep the reflection one as it is.
Two maps instead of one, sure, but the lighting one doesn't need to have all those pixels: if you take away the main directional light source it's only going to contribute to general hues and lighting levels.

[Joelaff]

It would be slow, but you could use a transparent plane with some refraction glossiness (and affect shadows) as a diffusion flat and place it between the sun and the subject just like you would use a diffusion flat (“silk” “frame” etc) in the real world.

Alternately, use a soft circular mask to blur the sun part of the HDRI in 32but app. Much faster.

[^Lele^]

It would be slow

This is an understatement: the lighting would be entirely made up of caustic rays, some of the hardest to compute and converge.
I prefer the second option by far. :P

[Joelaff]

This is an understatement: the lighting would be entirely made up of caustic rays, some of the hardest to compute and converge.
I prefer the second option by far. :P

Yeah, no doubt. I have done a lot of experimenting with traditional photographic lighting techniques in 3d programs. Grid spots usually work pretty well, with some sort of grid or hex pattern made with black geometry. Definitely slower than using the Directional parameter, but very controllable. Geometry based snoots work well too.

When you start getting into the Transparent light modifiers it does slow down a lot. I have used transparent mapped planes with transparent gradients on them for product stuff (keeps the existing lighting behind it vs a plane that goes from black to white, which has a hard edge).

I’d love to see more reasonable speed options in this regard. There is so much you can do in the real world with light modifiers. A lot of it can be simulated with HDRI maps on the lights, but the real time interaction is so nice. The change in polarization of the light with some modifiers is pretty cool too!

[Joelaff]

So I go that backwards on the Affect Shadows when trying to simulate a light modifier. I just tried this again. You actually need Affect Shadows off or you just get hard shadows with attenuation. The refraction gloss has no effect on the shadow quality (I assume some optimization thing). I think this is the case. The IPR was not consistently updating with changes to that setting.

I put thickness to the plane as well. (shell mod)

Anyway, it is very slow, but it does look pretty.