LC+IM -> LC pass much darker than IM - light portal bug?

[Micha]

@Lele: Thanks for the infos around PPT and LC.

Now I hope somebody help me for the light portals.

[studioDIM]

Aye, forgot those :P
Visualise them as a "smart" way to pull in the background in an interior scene.
VRay will do a better job at identifying apertures, this way, as the light object has some sort of priority when calculating GI.
A vray light set as portal, and a black background will produce no visible change to your scene, in other words (bar, maybe, a slight slowdown).
If turning portals on produces big changes compared to when they were off, it means that vray had issues before distributing the BG light properly into the scene.
Overall, having light portals on windows and doors makes for a more precise solution.

Lele

[vlado]

Re: LC+IM -> LC pass much darker than IM - light portal b

Secondary engine - light cache 1200-0.01 - GI multiplier 0.8

It is probably because of the 0.8 multiplier. You use the light cache as a secondary engine, so you are, in fact, asking yourself for V-Ray to give you darker light cache.

When you change the multipliers from 1.0, you are deviating from physical correctness. This may lead to this kind of issues with the light cache, which is very sensitive in that regard. If you really want the GI to decay faster as it bounces, simply use darker materials.

Best regards,
Vlado

[Micha]

Hallo Vlado,

I use the multiplier, because I want to use white colors with value 255 and textures with full color range 0 ... 255.
I suppose, most "white" textures are created, so that white is 255. For example a client give me a texture called "polaris" - a bright 255 white with some grey spots. An other example are colors of standard systems like NCS or RAL. My clients and my color conversation tables works in color range 0 ... 255. I don't want to set a multiplier 0.8 at each color or texture.
Or is the color range internal limited at 80% reflectance? This would be fine. Maxwell do it.

Vlado, what did you think about the light portal? If I use it, than it strong change the light distributation. I feel, a window should not work like an emitter plane. It looks like it cause a wrong decay.

Best regards
Micha

[studioDIM]

0-255 is 0-1 in float.
That is multiplied by HDR lighting.
And further it can be exposed through a physical camera.
You can have pure 255 white in the rendered pixels even if you cut that white to a quarter (as i suggest in the videotut i prepared-tips and tricks section), and have a better light distribution and renderspeed to go with it.
Since you know maxwell, you should also know about the reciprocity law and 255 whites warnings in the manual (for the beta. wouldn't know for 1.x).
Vray is no exception, but for the fact it caters for it automatically (pure white reflection will override the diffuse coefficient).

As for light portals, since they are only there to optimise light distribution, maybe the BG intensity is too high, and they distribute that "wrong" lighting in the correct way.
Without a scene to sink theeth in, i/we can only guess.
In any case, no, light portals DO work properly here.
There's no bug (in the max version).
However, as i said, settings may vary greatly, and the issues in my case almost always sit between keyboard and chair (that is it's me doing it wrong, not Vray).

Lele

[vlado]

I use the multiplier, because I want to use white colors with value 255 and textures with full color range 0 ... 255.
I suppose, most "white" textures are created, so that white is 255. For example a client give me a texture called "polaris" - a bright 255 white with some grey spots. An other example are colors of standard systems like NCS or RAL. My clients and my color conversation tables works in color range 0 ... 255. I don't want to set a multiplier 0.8 at each color or texture.
Or is the color range internal limited at 80% reflectance? This would be fine. Maxwell do it.

Possibly, although I'm not sure. Even if so, it is more correct to apply the effect to the textures, and not to the GI solution. You have a texture multiplier and everything...

Vlado, what did you think about the light portal? If I use it, than it strong change the light distributation. I feel, a window should not work like an emitter plane. It looks like it cause a wrong decay.

Can you try that with a multiplier of 1.0 for both primary and secondary GI? It might be a bug in the light portal for the rhino version, so I'll have to check it anyways. I'm sure it works correctly in 3dsmax.

Best regards,
Vlado

[Micha]

....

Vlado, what did you think about the light portal? If I use it, than it strong change the light distributation. I feel, a window should not work like an emitter plane. It looks like it cause a wrong decay.

Can you try that with a multiplier of 1.0 for both primary and secondary GI? It might be a bug in the light portal for the rhino version, so I'll have to check it anyways. I'm sure it works correctly in 3dsmax.

The test show it - only one difference between the images: light portal lights on/off. I posted the problem at the vfR forum now.

tests with low settings and reflections disable


@Lele: Thank you for the infos and the tutorials. I will study your 0.25 method. Looks very interesting.

[Micha]

Vlado, what did you think about the light portal? If I use it, than it strong change the light distributation. I feel, a window should not work like an emitter plane. It looks like it cause a wrong decay.

Can you try that with a multiplier of 1.0 for both primary and secondary GI? It might be a bug in the light portal for the rhino version, so I'll have to check it anyways. I'm sure it works correctly in 3dsmax.
...

I let it test at 3dsmax - same effect: light portal and IM dosn't work right together. No special Rhino bug.

Good luck for hunting the bug.

[ene.xis]

other than being surprinsingly noisier...it doesn t look nothing near to ur "tests"...so id check ur "bug" consistency...

[vlado]

I let it test at 3dsmax - same effect: light portal and IM dosn't work right together. No special Rhino bug.

Good luck for hunting the bug.

Ok then, I'll check it out.

Best regards,
Vlado

[studioDIM]

Ok, i think i got where the bug may lay:
I opened the testscene, Micha.
You have a big VRay light in covering the window, with a multiplier set to 1.
And you have a png as the GI background, with a multiplier set to 25.
Switching lightportal will hide the light and use it as a pass-through for the environment.
Setting the light to normal, but not invisible, will block the BG out, and light with the 1.0 intensity coming from the light.
I replicate the issue on your scene, and it seems to be working in a similar fashion as to what you described.

Lele

[ene.xis]

ahhh....these damned bugs!

[Micha]

...
Setting the light to normal, but not invisible, will block the BG out, and light with the 1.0 intensity coming from the light....

I'm not sure what you mean with "normal". I placed a light portal in the window and turn on/off the complete light. An other test is, to turn the light portal on and render the scene with IM/LC and QMC/LC.

[percydaman]

he means, 'normal' as in, not a 'skylight portal'.

[Micha]

Ok, so don't set the light at normal, this is not my bug.