Glass refraction with bottle+water Tweet

[FLXmedia]

Hello guys,

since i am new to octane, i find myself in a tricky situation, where i need to render a bottle of gin.
With the standard Cinema4D physical renderer, the refraction in my bottle (and on the edge) looks much more realistic than my rendering on Octane.

I attached three images to this post. 1. is the original Cinema4D Physical Render. (Water index 1,33)
2. is the Octane render with a standard glass material from the Live DB (Index:1,65) and a water material at Index: 1,01.
3. is the Octane render with the same glass material (Index: 1,65) and the water material at Index 1,3.
4. a real picture showing the desired refraction.
Can you please look at them and tell me, what i did wrong?

The bottle itself is constructed with a lathe-object with the liquid being another lathe object intersecting with the bottle.
Is this modelling technique wrong?

Thanks in advance for your help and time!

Greets,

Felix

[bepeg4d]

Hi,
difficult to say with your screenshot, there is not too much to reflect/refract right now compared with your reference, but in OctaneRender, you need that the liquid is slightly overlapping the glass, to have the correct effect.
Or modelling in a different way, like explaned in the picture below:

IMG_0369.PNG

ciao beppe

[FLXmedia]

Hi,
difficult to say with your screenshot, there is not too much to reflect/refract right now compared with your reference, but in OctaneRender, you need that the liquid is slightly overlapping the glass, to have the correct effect.
Or modelling in a different way, like explaned in the picture below:

IMG_0369.PNG

ciao beppe

Hi beppe,

thanks for your help.

I am looking for the refraction especially on the back label refracting to the bottle sides. I modelled it with a small overlap.. so you think i should model the liquid completely inside the glass?

Thank you!

[nuno1980]

Hi, FLXmedia

If you deserve "real life" then all specular materials must have dispersion.

-- Glass settings:
reflection - 0.930
transmission - 1.000 (greyscale color)
ior - 1.509* (it's ior dispersed and the original ior is 1.53)
dispersion coefficient - 0.006*
rest default

-- Water at +/- 18 ºC settings:
reflection - 0.930
transmission - 0.998 (greyscale color)
ior - 1.323* (ior dispersed and original ior 1.333)
dispersion coefficient - 0.003*
rest default

* - see 2 formulas (convert from "abbe number" to "dispersion cofficient (= B)"; calculate from original refractive index (ior) to
ior dispersed) ---> viewtopic.php?f=25&t=61949#p316752 (look my message on Jun 24, 2017 11:37 am UTC)

--> I'm recommended you use PMC kernel for almost real life, if your videocard is very fast (eg: GTX 780, 980, 980Ti...).

[nuno1980]

LAST EDIT: water only (look my previous message) or

-- Water with 45% alcohol at 20 ºC settings:
reflection - 0.930
transmission - 0.982 (greyscale color)
ior - 1.350 (ior dispersed and original ior 1.360)
dispersion coefficient - 0.003
rest default

[FLXmedia]

Hi, FLXmedia

If you deserve "real life" then all specular materials must have dispersion.

-- Glass settings:
reflection - 0.930
transmission - 1.000 (greyscale color)
ior - 1.509* (it's ior dispersed and the original ior is 1.53)
dispersion coefficient - 0.006*
rest default

-- Water at +/- 18 ºC settings:
reflection - 0.930
transmission - 0.998 (greyscale color)
ior - 1.323* (ior dispersed and original ior 1.333)
dispersion coefficient - 0.003*
rest default

* - see 2 formulas (convert from "abbe number" to "dispersion cofficient (= B)"; calculate from original refractive index (ior) to
ior dispersed) ---> viewtopic.php?f=25&t=61949#p316752 (look my message on Jun 24, 2017 11:37 am UTC)

--> I'm recommended you use PMC kernel for almost real life, if your videocard is very fast (eg: GTX 780, 980, 980Ti...).

Wow, thanks for your help man! I'm using the PMC kernel, since we are working with a 980ti.

But, have you done something similar before? How do i have to model? Slight overlap in glass/liquid or small gap between that?

Thank you very much for your help!!