[deadalvs]

the model is very simple, and I used Thomas An's method: The liquid is enclosed in the glass object.

why do you use two fillets for the surface tension effect ?

is this why there's a thin (annoying) double line at that edge ?

[Maximus3D]

Was it really necessary to quote with all those images included ? kinda waste of space as they're now duplicated..

Anyways, the milk may be simple to you Kurt but to us normal humans it's no easy task. Impressive and it looks very realistic! a true jawdropper'n'eyepopper

I'll be posting my first sss test image tomorrow when it finished rendering.. (i hope) :/

/ Max

[simmsimaging]

That milk material is bang on - nice work Kurt. Lot's of cool tests in here, but that is the one that really blew me away.

b

[3dtrialpractice]

the milk its self the sss works great and looks real pretty but the render dont look perfect- theres something not right about the crazy side repeating lines..the reflections of itself inthe glass.. definatly doesnt look physically accurate. nor close to perfect- great job w a quick test kurt milk truely looks good. .its just that the glass look way off and its maybe due to that method of puttign the milk geo under the glass geo

[Thomas An.]

the milk its self the sss works great and looks real pretty but the render dont look perfect- theres something not right about the crazy side repeating lines..the reflections of itself inthe glass.. definatly doesnt look physically accurate. nor close to perfect- great job w a quick test kurt milk truely looks good. .its just that the glass look way off and its maybe due to that method of puttign the milk geo under the glass geo

It is perfectly fine to make claims like that ... but you have the burden of proof. It is best to go ahead and take a photo of an actual glass, then go ahead and model it as close as you can and then compare the render from exactly the same vantage point.

Maybe the glass mxm he used needs adjustment ... or maybe things did get broken since a long time ago when we were testing glasses. Or maybe the SSS causes extra internal reflections in a pure dielectric. It is up to you to verify it for us though

PS: There are some stripes evident here (on the right hand side):

[Bubbaloo]

I think, and this is just a guess, that the "stripes" are due to poly's in the fillet. I don't know how well refraction in Maxwell handles poly smoothing. It's probably better to have as many poly's in a curve as you (or your computer) can tolerate for this type of situation.

[Eric Lagman]

I noticed the vertical lines in the glass test also. It looks like they occur in the same spots that the polygon mesh runs vertical.

[3dtrialpractice]

thomas- im glad u posted that pic- and ur right the variables of the situation do caus for certain looks.
and if the stripes occur likethat in that photo then it shows the power of maxwell -esp for just a quick test kurt put together.. i wasnt trying to be down onkurt/render cuz its a test ive never had patietnce to verntur into..
L

[pwrdesign]

I have to post something here

A porcelain corvette C6

24 min render, SL 13.14

10 nodes (core2duo and dual quad cores mixed)

[Maximus3D]

Here's my test of sss.. but i'm not satisfied with it because it rendered for 12 hours 31 minutes and 3 seconds and it's still very noisy i expected it to be clean. And the sss on the main body is not really visible, it's as translucent as the beak of this mesh is yet you can't see through it. The main body of the mesh is ofcourse hollow.



/ Max

[MetinSeven_com]

Wow, looks great despite the noise. Nice lighting!

The new SSS is amazing but noisy indeed. I hope NL manages to optimize the new SSS code in a next version.

[KurtS]

why do you use two fillets for the surface tension effect ?

is this why there's a thin (annoying) double line at that edge ?

The liquid (milk) is completely enclosed by the glass volume. This means that there is a very thin glass layer above the milk surface. I don't know of any other way to avoid intersecting surfaces and still have correct refraction.



cheese test:

[deadalvs]

@ KurtS


as far as i've understood my physics teacher a few years back in optics it was like that:

every area that is the edge of two different materials has a certain fresnel (n) value.
so between air and water, that's 1.333
between air and glass it's in the 1.5 neighborhood.

so the main point i want to stress is that there is nothing like two surfaces that have to be modeled. it's one and the same surface between glass and milk that is responsible for the realistic look. thus an offset surface is physically incorrect.

i have not tested stuff like this with maxwell myself so far, but i am not sure if maxwell really can handle this fresnel boundary between an SSS material and a glass material since the shared edge surface is ONE AND THE SAME... (a realistic material MUST incorporate all these values, maybe with a coating or something to mimic the fresnel value)

i have encountered this problem already a few years ago, but still did not try it myself.. lack of time and especially cpu power over here.

AND i've never seen a PERFECT PHOTOREAL rendering of that situation.

* * *

btw. is the milk shader one of the presets ?

[deadalvs]

cheese test:

give it a little more color .. our swiss cheese called «emmentaler» is a little more yellow !

cool model though !
-->
http://images.google.ch/images?q=emment ... rge|xlarge

[KurtS]

re deadalvs:

I suggest that you have a look in this topic: http://www.maxwellrender.com/forum/view ... 6888#26888

(the milk is not one of the presets)