Maxwell Render | Official public forum

thanks Mesut Capkin!

w

you're welcome.

wimver wrote:sorry I have no time to search for a place to upload a zipfile

Don't. Just email me >> tom@nextlimit.com

wimver wrote:1. create new material
2. in wizards, select AGS
3. amount of reflection 100
4. type normal
5. color white 255 x 3
6. set roughness to 20

Simply, impossible. Bullet 3 and 5 is enough to make it full mirror, bets off...
Where do you run this wizard? In a Plugin or in Studio/MXED?

But it doesn't look like a mirror!!!! Would we not see the brickwork and plant this side instead of the shadow behind?

wimver has emailed me explaining it was a confusion. That render he has posted is a frosted glass, not an AGS.

Oh what a shame, looks good! In that situation with light from both sides it would certainly clean quickly anyway so I guess an AGS not needed.

I know in the example I posted as you suggested earlier Tom the background didn't blur, but it was originally good enough for me for that scene!

sorry for not responding sooner, I am teaching in evening classes, just got back.
I am still confused, and I feel I am not the only one, but it is almost midnight now, tomorrow I will post some screendumps and another simple render with my "not AGS frosted glass" to clear things out...

but about the confusion with Tom, here is what I forgot to mention in the above list:

I also altered the transmittance to some grey, attenuation to 30 cm and ND to 1.2. I also activated the force fresnel, but this does not seem to have an effect (haven't checked the manual yet what this thingie does)

so basically, if I create an mxm with the AGS wizard, and then change some parameters, the material is not worthy of being called an AGS anymore...

goodnight
wim

You can make chicken soup with fish, and still call it chicken soup if you want
If you change the ND and/or transmittance, the "ghost" component of the AGS material isn't invisible anymore, which is the whole point of it. To have reflections and transparency but without any refractions (to avoid caustics calculations), you need one BSDF which is the shiny, opaque reflective part, and one completely invisible BSDF, which will provide transparency. In V2 you can even do it simpler by just having the shiny opaque BSDF and lowering the Layer weight for transparency. It depends on the look you want. By changing the transmittance from 255 to lower and changing the ND from 1 to something else, you've created a glass material, which refracts light>caustics>defeats the purpose of the AGS approach.

wimver wrote: I also altered the transmittance to some grey, attenuation to 30 cm and ND to 1.2. I also activated the force fresnel, but this does not seem to have an effect (haven't checked the manual yet what this thingie does)

so basically, if I create an mxm with the AGS wizard, and then change some parameters, the material is not worthy of being called an AGS anymore...

what mihai says !
& In maxwell there as no SHADERS like other apps(ie mental ray have fastskin,blinn,phong) that have unique properties to them...allmaterials in mawell share same properties and the material looks like metal/glass/water/AGS to how those properties are set(not like other aps where you choose a glass shader and then tweak it) so i can use the wizard to make what it calls a metal..but change the properties around enough and have something that will render and look like a fuzzy velvet..
So AGS isnt a "shader" its just a acronim for Architectural Glass Solution.. and its just a combo of the bsdf settigns so that it renders like a glassish look w/o the noise and casutics of a dielectic(glass-liquid,etc) material(its just a setup which FAKES glass).. so when you use wizard you are not making "ags shader" to tweek.. its just making something of a fake glass and its important to note that it makes a material that has a nd of 1 ..

a nd of 1 = "air" meaning NO REFRACTION -- as soon as nd>1 (even1.01) then you have refraction (and casutics) and more noise(calculation) to clear up... the reason that your render u posted still looked clean/nice is due to advancments in dielectic caculations in V2 engine

also AGS Preset has transmitance of black (rgb000) and as soon as you go rgb>000 then you make a dielectic so this also kills the AGS "preset properties"

-Luke

wimver wrote: but hey, I think I just found what went wrong: When I create a new material with the AGS wizard, the opacity mask is set to 12.5 %

the opacity mask for AGS being set to 12.5 is not "wrong" its acctually what makes it the AGS "trick" (in version 2 of maxwell- version1.x materials didnt have opacity property so a AGS trick had to be done with 2 layers weighted togther to make a fake transparent look)

so an opacity of 12.5 makes a material transparent without any refraction or casutic "sideeffects" so it renders clean faster(IF transmitenc is pure black(solid object) and nd=1) .. the only other way to make a object transparent it to set its transmittanc rgb values >(greater than) 0,0,0.. example: rgb256,256,256 (white) and IN COMBINATION with the proper attenuation distance (so that light will travel through the object and reach the other side to make it look transparent)

so when you made your opacity of the ags preset 100% then there should have been NO (fake/trick)tranparency..it would rendered/looked like a mirror...

you said you changed the transmittance color and raised the nd..and THAT IS what made the 100% opacity material transparent (Not the fact that you started with a ags preset wizard material)

Thank you Mihai and Luke, it is getting all clear now... looks like AGS

so to finalize this, here is the complete picture...
I did not touch the subsurface properties.
the only thing still to figure out is how to avoid the dark shadow cast by the glass and seen through the glass.

cheers!

wim

We seem to be going in circles This last picture is a glass material, which shows exactly the problem that AGS approach solves, that dark shadow you see through the glass, which eventually will solve and show caustics through the glass, but it will take a really really long time.

Haha, once you change the nd to anything other than 1, it ain't AGS any mo...

I have had some luck with a faked frosted glass solution on single planes using the thin sss option. You might want to look into that. I'm not sure what would be faster, thin sss or letting caustics through glass clear up?

T_RNLDesign wrote:I have had some luck with a faked frosted glass solution on single planes using the thin sss option. You might want to look into that. I'm not sure what would be faster, thin sss or letting caustics through glass clear up?

Care to post some details??