[tom]

so, that would be one single closed object with a multimaterial, that is opaque on one and clear glass on the other hand (minimally right above the top surface of the liquid), both with an nd of let's say 1.51. this would give me correct results refractionwise, right?

I haven't tried yet but, it should.

what if i'd seperate the top surface of that closed glass and make it an own object? without changing any vertices' postition, while fulfilling the material premisses, you mentioned. i'd have the same positions for all vertices as before... only 2 seperate objects. one surface object, clear glass with nd=1.51 and and one "rest-body", opaque, nd=1.51. would that also work?

No, that will absolutely fail. Having those triangles member of object is a must.

what i need to seperate for my understanding: is it the closed-volume-context (as in sss-topics), or is it the use of a continuesly same index of refraction, which is important to get correct results?

I'm sure, same Nd multi-material should not be different than the closed-volume-context.

[mashium123]

...
No, that will absolutely fail. Having those triangles member of object is a must ... I'm sure, same Nd multi-material should not be different than the closed-volume-context.

there it is!
i've been looking for that one for so long now but never was able to find it as such a clear statement.
thank you very much

[alexxx_95]

I'm totally lost in the explanations ... it's due to my low english level... any little details sketch would be welcome...

[mashium123]

me, i've been interested in this liquid-in-glass topic for quite a time now.
there has been a discussion about it a few years ago, that can be foun on this forum; tom recently made a sticky topic about it:
http://www.maxwellrender.com/forum/view ... 15&t=17147

there you'll find - not in little scetch manner - material about this topic that had been provided by a member with the nick "Thomas An." and discussed by talented people.
if i had any talent at all, i'd allow myself to judge the quality of the material, that he provided. but i haven't, so i don't.

what i've been doing: i always got back there and tried to understand - again and again... but there was a point in the practical implementation, that i never found explained in a clear manner (well at least clear enough for me).
now again, with tom's help, there's another point that i can check as internalized...

go there, have a look at it, try to understand...
i really hope, this "Thomas An." will return to the forums some day (haven't seen him for a long time now).
such an awesome origin, providing knowledge to all. for free!

[alexxx_95]

thanks mashium, I already know this topic and thomas's method that I used when needed. but it's the "open top glass " version taht I don't understand... I will take a look closer to the thomas' topic again and try to try to understand it deeper

thanks

[tom]

There's no such open-top-glass version. It was my quick faulty workaround for opaque stuff. Later, we evaluated the idea of having it as a multimaterial.

[Maximus3D]

I made some small modifications to the liquidmesh inside the glass, scaling it up and down very little to test what would produce the correct effect. But i been staring myself blind at this so can you guys see what if any of these look correct ?

Rendertime varies between 15 to 30 and to 40 minutes in these two tests with 3 renderings running at the same time on my machine.





And a photoreference of the real thing from my kitchen, shot with my cellphone camera at a approximate angle and position.
It's lit with sunlight coming through a window at the right side of the glass. No other lightsources present in the 'scene'.
Second referencephoto was shot at the same location, from a similar angle as the renderers above and it's lit by one single fluorescent tube.


/ Max

[rusteberg]

thanks fernando. yes, the arch images are straight from maxwell. animated the car with motion blur enabled with 120degree rotary shutter set in the max camera parameters
tom, agreed! i'll work on nailing that tire material, when i have a little down time, i'll start a wip thread with the car

max....... i think it's time to remove that watermark how about a "Get Max V2 Fund", if you donate, max can share his pizza recipes!

edit: max, what's up with the funky faces in that one render?

[RK_art]

I adapted my skin-shader to V2, now it's getting to be fun.

Rendertime: 18 min on my Quadcore 4 Gb , no postwork.







Still need some adjustments, a way too much sss overall, but close to the desired result and of coarse FAST

And here with some psp-postwork:

[3dtrialpractice]

WHAT RK!! wow 18 mins!!! awsome!! looks great..
can you share what you did different form your 1.7 wip thread to these V2 skin adaptions?
are these pics all from 1 render (multilight?) or multiple IBL setups?
----way to go.. good to hear your not frustrated anymore with V2!

- I too got the hang of sss after some tests and making some sss wizrad tries..
here I have renders of a quick 5min sculpt i did testing 3dcoat newest build 3.1.08
NOT 18min renders..these took over 2hrs on my 3ghz machine but got to 15 samples:
wip of shader that renders derived from:http://www.filefactory.com/file/a0ec097 ... wShine_mxm

finals to be posted online mxm library resources.maxwellrender(what happened to the mxm section on the forum?) site l8r
-Luke

[tom]

Great renders out there! Although, I'm having hard times deciding whether fast render times or these creatures are more frightening.

[RK_art]

Very nice SSS-shaders, 3dtrialpractice. Well done.

All pictures above show one rendering with different multilight.

For human skin, there is no need for usual volume-SSS.
So I switched from normal volume-sss to single side sss. That makes things so fast.

This one is 1.5 hrs, reduced SSS compared to the ones above, no postwork:

[3dtrialpractice]

For human skin, there is no need for usual volume-SSS.
So I switched from normal volume-sss to single side sss. That makes things so fast.

ahh--DOH! of coarse thinSSS wow duh.. i cant beleive
ive been boxed in betfor to not consider thnSSS-- i gues we never had it w this engine anyways..
SO thinSSS isnt restricked to open geometry only i guess it does work on closed surfaces-i had done a test first day wen i got v2 on a ball but the settings didnt change the look-so i guessed no go for closed geo--but i see it does work.
WELL thanks for that TIP! thats a fun factor for sure!!
-Luke

[tom]

Yes, Single Sided SSS is nice shortcut. Although, the satisfaction highly depends on illumination. Because, it'd not work prefectly for backlit ears and/or they will look hollow for sure. But, still very useful for simulating non-extreme situations and the best plus you can have is being able to use trans and scattering textures at the same time.

[Maximus3D]

rusteberg: Hehe, but i'm starting to like the watermark in my renderings. It adds a certain sexyness to them donations would be awesome for sure but i still got around 3 weeks left to play with 2.0 and in that time find cash somewhere or win the lottery so i can upgrade. We'll see what happens.. btw, my pizza recipe is worth alot more than the price of a Maxwell 2.0 upgrade.

Those funky faces you see there in one rendering is because of the liquidmesh overlaps the glas mesh polygons and that causes that meshproblem during rendering. The polygons share the same space. I included that picture to illustrate the problem better, unfortunately i forgot to explain that in my post, doh! the other examples they have the liquidmesh offset either inwards or outwards by a tiny amount so there's either a tiny overlap where the liquidmesh overlaps the glass or a tiny gap between the liquidmesh and the glass.

RK: those are beautiful results you managed to get. Well done sir! but i'm curious what the sss looks like without the colortextures applied as the colormaps can fool the eye a bit. I too think the thinSSS could cause you trouble in the long run depending on the lighting, camera angle and more it could look like the guy is made of paper.

3DTP: Hehe what a cute little critter you made your sss also looks successfull, good job!

Here's a test i ran this morning on one of my Modo materialscenes i recently created. I built the scene to help people easily create cloth type materials, carpets and such things and then render them nicely all within Modo. So i ran this through Maxwell and threw on the tracingpaper mxm from the inclulded library and hit render. Still a bit noisy but not too bad..


/ Max