[Eric Lagman]

Forget the reason of why in all these discussions. All you need to understand is how to do it. Look at last diagram and ignore everything but the black and red lines. From a simplified modelling standpoint its two profiles that get revolved around a centerline. The black line is your closed volume that you would have as glass. The red is your OJ. There is a very tiny gap of .1mm or so between the glass that surrounds your OJ and the top of you OJ volume shown in red. This avoids the triangle overlapping issue in this area, and gives you accurate refraction. Hope that clears things up. No pun in intented.

Keep in mind the end result is an optical illusion because of refraction. The liquid does not really go all the way to the edges of the glass in your geometry so you dont have to worry about overlapping polygons if you model it the way he shows in the diagram.

[MS]

any ideas

And a thread:
http://www.maxwellrender.com/forum/view ... 884#275884

[Hervé]

depending the conditions (lighting mostly), Orange juice and milk for instance need about 23 samples to get acceptable noise...

turn sky system off.. prefer emitters..

[tom]

TOM....r u here Tom?

Could you simplify the scene down to a single glass of orange juice that is still causing the same and email me?

[frosty_ramen]

Tom
I just sent it off to you, thanks for the help

[tom]

Theoretically there is no problem with your scene but it will be noisefree after years using this setup. Besides, it will show thickness of glass which is not physically correct. First of all, your glass/liquid model should apply Thomas' definitions:



...and it's the only thing you need to change.

[frosty_ramen]

Thanks Tom for the info

[Fernando Tella]

Tom, every time I see that diagram I have the same feeling about FigA and B; both conditions are approximately equal on the infinitesimal level when we talk about IOR but they are definitively not the same in the finish of the materials. Is that correct? Is Fig.B what it should be?

[Eric Lagman]

Tom, every time I see that diagram I have the same feeling about FigA and B; both conditions are approximately equal on the infinitesimal level when we talk about IOR but they are definitively not the same in the finish of the materials. Is that correct? Is Fig.B what it should be?

I think that is just a really complicated way of saying the gap is so small it will not be noticeable to the visible result. Or do you mean the Top of the OJ will not look like OJ, but glass material instead as far as roughness setting etc? That is a good question.

[Fernando Tella]

That's what I mean. If the liquid has more roughness than glass will it show correctly with option A?

Edit: A test would solve my doubts but... I have some renders going, well in fact, I'm so lazy

[Hervé]

... I have some renders going, well in fact, I'm so lazy

no offense, but that's honesty Fernando !... rare enough to be noticed..

[tom]

Tom, every time I see that diagram I have the same feeling about FigA and B; both conditions are approximately equal on the infinitesimal level when we talk about IOR but they are definitively not the same in the finish of the materials. Is that correct? Is Fig.B what it should be?

I think that is just a really complicated way of saying the gap is so small it will not be noticeable to the visible result. Or do you mean the Top of the OJ will not look like OJ, but glass material instead as far as roughness setting etc? That is a good question.

Yes, that's how it works. Theoretically, it's surely being refracted but impossible to notice.

[frosty_ramen]

just a quick question,
does my orange juice have to be smaller than the inside of the glass?

[frosty_ramen]

Here is what is happening now,
the juice and the glass top are interfering with the cup.


any ideas, do i scale both down... will that negate the whole idea of a glass top on the juice??

-Dan

[Hervé]

just a quick question,
does my orange juice have to be smaller than the inside of the glass?

there is no inside glass..