Please let me verify my understanding/assumptions on the material and volume parameters:

kt and kr define how much of the light is reflected and how much in transmitted through the irradiated surface. So I expect the two to sum up to 1. (100%) per color channel. But isn't this ratio depending on the angle? So are the values defined for square angle, and adjusting for the actual angle?

absorption is the lost of power of light by traveling through the volume, right? What is the unit? 0.01 mean 1 percent loss per 1.0 distance?

And what is the remaining meaning of transparency then?

What would be realistic values for window glas? kr=0.04, kt=0.96, absorption=0.01 (for 1. is 1cm), transparency=1. ?

Did I miss any documentation about this?

Thanks

## glass parameters

### Re: glass parameters

Nope, the sum of a "perfect" glass is 2, not 1, because, as you say, the amount of light reflected/refracted is angle (and IOR) dependent. So kr and kt are just 2 scale factors for the amount of light reflected and transmitted(i.e. some amount of light is absorbed by the material in the process of reflecting, transmitting).mick wrote: ↑Mon Jun 25, 2018 1:11 pmkt and kr define how much of the light is reflected and how much in transmitted through the irradiated surface. So I expect the two to sum up to 1. (100%) per color channel. But isn't this ratio depending on the angle? So are the values defined for square angle, and adjusting for the actual angle?

Yes.

kt is an attribute of the surface, the borderline between 2 volumes. The absorption is an attribute of volumes. The amount of light transported is subject to kt when going from one volume to the other (and can be just 1 = no effect) while absorption affects the transported light while traveling inside a volume (i.e. between 2 borderlines).

### Re: glass parameters

Thanks for the reply.

And does LuxCore adjust the actual reflection and transmission depending on the angle?

I understand that there is a loss (neither transmitted nor reflected but absorbed at the surface). But transmitted plus reflected cannot be higher than 100%. The max for kt and kr is 1. Lets assume a loss of zero, for 20% reflected and the rest transmitted into the object, what would kr and kt be. (1.6 and 0.4???)the sum of a "perfect" glass is 2

And does LuxCore adjust the actual reflection and transmission depending on the angle?

This is clear, as already confirmed for the previous quote. But what about the "transparency" parameter?kt is an attribute of the surface, the borderline between 2 volumes. The absorption is an attribute of volumes. The amount of light transported is subject to kt when going from one volume to the other (and can be just 1 = no effect) while absorption affects the transported light while traveling inside a volume (i.e. between 2 borderlines).

### Re: glass parameters

Yes. If the angle is close to 90° (equal to the mesh normal), nearly all light is transmitted and nearly no light is reflected.

If the angle is very shallow, almost all light is reflected and almost no light is transmitted (aka Fresnel effect).

The transparency parameter is not specific to the glass material, it is a parameter that every material has and it controls the overall transparency.

It is mostly used for unrealistic tricks, so you can just leave it at 1.

### Re: glass parameters

mick wrote: ↑Mon Jun 25, 2018 4:37 pmI understand that there is a loss (neither transmitted nor reflected but absorbed at the surface). But transmitted plus reflected cannot be higher than 100%. The max for kt and kr is 1. Lets assume a loss of zero, for 20% reflected and the rest transmitted into the object, what would kr and kt be. (1.6 and 0.4???)

Nope, the total is "80%" * kt + "20%" * kr. Like I wrote, kr and kt are scale factors, not absolute values. If kr = 1 and kt =1, the result is just "80%" * 1 + "20%" * 1 = "100%". If kr = 0.5 and kt = 0.5, the result is "80%" * 0.5 + "10%" * 0.5 = "50%".

### Re: glass parameters

I think I begin to understand. So based on Fresnel equation the transmission and reflection for perfect glass is calculated, and then multiplied by the respective factors kt or kr. Right?Nope, the total is "80%" * kt + "20%" * kr. Like I wrote, kr and kt are scale factors, not absolute values. If kr = 1 and kt =1, the result is just "80%" * 1 + "20%" * 1 = "100%". If kr = 0.5 and kt = 0.5, the result is "80%" * 0.5 + "10%" * 0.5 = "50%".

Can you give me rough values for kt and kr for let's say normal window glass? 0.95, 0.9? I'm doing physical simulations where this matters.

### Re: glass parameters

Yes, correct.

It is heavily dependent on the type of glass used but, yes, it is likely to be something over 0.9.