All posts related to V2
User avatar
By tom
#310389
Maximus3D wrote: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.
None of these methods will work. You should enclose the liquid inside glass geometry.
Btw, nice paper cloth there!
#310390
So the liquid should be part of the glass geometry.. that sounds weird! that means any Realflow simulations rendered will be inaccurate and unrealistic looking. I have to run a new test today to see what your suggestion would look like Tom. But enclosing the liquid as a part of the glassmesh means the liquid will have no innerfaces. Hm..
Btw, thanks! :)

/ Max
#310392
Tom wrote:
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.....
Yeah, but only if you use it as the only material-component ;)

Look at this:

Image

That's the same setup with single-side SSS, but I don't use it as the only component.


Here is the final result after ca. 3.5 hrs, no postwork:

Image


and with Postwork:
(I wanted to look him as if he had a hard night ;) , so I used cooled colors and high contrast)

Image

Max: I will do a rendering without the albedo-texture so we can see what this will look like.
#310400
Max, here are multilight-variations of exactly the same setup, but instead of the albedo-texture a uniform light grey is used:

Image
Image
Image
Image
Image
Image

rendertime is 19 min.
Remark: The grainy one (No.4) has mostly HDRi-lighting, but it's the same render as the other ones.
#310419
Max i think your very first water render you posted looks the best.. there are plenty of simularities to the real thing.. but i guess its soo hard to get it 100% since it depends on exact angle of straw. thikness of glass,and angle of light
Maximus3D wrote:So the liquid should be part of the glass geometry.. that sounds weird! that means any Realflow simulations rendered will be inaccurate and unrealistic looking.
ya.. isnt that weird.. first thing i though about wen the realflow/maxwell combo nextlimit rekleased before i got to use v2 was that the liquid in glass stratagy surely must have changed otherwise they wouldnt push the realfow/maxwell connection..

but i AM glad they released V2 .. (since i dont own realflow nor do my projects require it for now) but maybe one day they will come up with something that works with liquid geo and glass geo.. I guess for fun/non exact product animation/renders no one will miss not 100% physical corect glass look if modeled out like real workd water geo ontop of glass... I mean your renders look great even though its not the perfectionrelaistice waterglass look..
#310423
Thanks for posting those grey images RK, i wanted to see how strong your sss was and it looks very faint. I would say it's almost too little sss now. And fix those creepy eyes!

3DTP: Yes that's what i also think, the first one feels the most realistic. To try to match it even more to the real thing i spent this day matching the cameraangle, lighting, materials, models and enviroment to one of my referencephotos of the real glass of water. You can see the results i've got so far here below, still some more tweaking to do especially to the lighting. But it gives you a general idea of what Maxwell can do compared to the real world. :)
True that based upon that liquid inside glass mesh theory you will not get that fully realistic result. However i think also that the fact the liquid is in motion humans don't care about the differences which are there. As long as it looks good enough or close enough to what happens in the real world.
Anyhow.. thanks! :)

I will try to make this the last of the glassrenderings i post in this thread now. :D promise!

Left side shows the Maxwell rendering and right side shows the referencephoto.
ImageImage

/ Max
#310428
huh.. i DEFINALTY like this latest render of yor glass MAX.. what i love is how the ref photo squiggles the tile reflection in the bottom of the glass..and So DOES the render..its got a squiggle too! so to me.. if i hadnt seen photo.. and especially if the glass was just a object amongst a stilllife or as a prop in a active scen like you say i wouldnt care..in fact its beautiful.. i see rendering as showing a artistic version of what you wanna express and maxwellV2 really helps ppl be artistic yet stay gronded in physicaly working light!
MAX you should just start a WIP thread for further developing these glass renders if you do anymore..

btw here is a link to the mxm that was WIP basis of the rendered clay baby monster pics before: http://www.filefactory.com/file/a0ec097 ... wShine_mxm
Ill post the final mxm's on the resources online maxwel website l8r
_luke
User avatar
By tom
#310434
Maximus3D wrote:But it gives you a general idea of what Maxwell can do compared to the real world. :) True that based upon that liquid inside glass mesh theory you will not get that fully realistic result. However i think also that the fact the liquid is in motion humans don't care about the differences which are there. As long as it looks good enough or close enough to what happens in the real world.
Now, this is not true. Clear water in a clear glass has a very simple characteristic and it's simply all about the refraction. Not only Maxwell but, any engine is able to render exactly the same look as soon as you provide the parameters/models correctly/accurately. Your setup is approximate and that's why it end up with an approximate look ;)
#310459
tom wrote:
Maximus3D wrote:But it gives you a general idea of what Maxwell can do compared to the real world. :) True that based upon that liquid inside glass mesh theory you will not get that fully realistic result. However i think also that the fact the liquid is in motion humans don't care about the differences which are there. As long as it looks good enough or close enough to what happens in the real world.
Now, this is not true. Clear water in a clear glass has a very simple characteristic and it's simply all about the refraction. Not only Maxwell but, any engine is able to render exactly the same look as soon as you provide the parameters/models correctly/accurately. Your setup is approximate and that's why it end up with an approximate look ;)
"Correctly" maybe but not necessarily "accurately." I don't render a lot of liquids in glass but it looks like to get a realistic result the solution is to place a layer of glass over the layer of liquid. That's neither accurate or intuitive. In general though, of course you're right. The more accurate the model the more realistic the result. The rendering engine might not make any approximations or shortcuts but the modeler certainly must.

-Brodie
#310467
3DTP: :) thanks alot! i'm glad you like it. Maybe it was a good idea to attempt to replicate the photo with Maxwell to see if and what way the liquid actually renders correctly and now i think i got it. It took a while a bit of work but the results are pretty sweet hehe. It sure is a artistic interpetation of the photo, as close as i could match it to the real thing. Thanks a bunch for your mxm, i'll be dissecting that to see how you did it :) i can admit i haven't got a good grip yet of sss so this will be helpful.

Tom: Aha, alright that i did not know. Thanks for the explanation, it's all about approximations isn't it.. :)

Ok i know i said the previous image was supposed to be the last one but i just gotta post this one, i made a ton of adjustments to materials, lighting, models and also the camera to try to match it even better. There's a couple of bugs still left to squash in this, one of them is the thickness of the glass which is not correct, the other is the lighting which could be a little bit better, and then there's the general shape of the glass itself which is not 100% accurate to the real thing. However.. this is pretty damn close as it is now! :)

ImageImage

Btw, here's what the scene looks like. I rebuilt much of my kitchen only with primitives but to bounce the light around correctly.
Image

A new tweaked version is in progress with many more fixes :) here's a early preview..
Image

/ Max
#310509
MAX.. the new one has a Line of what looks like intersecting geometry (running right thru the maxwel watermark)
did you scale up/down the water geo cuz your prev render ddint have this line..and it doesnt look like a refraction element either..

and good job on the kitchen proxys very good idear to help with the light..
besides the weird geo line i think the laters test is really great..and i like the reflections of the emmiter better in it too..
the straw will never look refracted like the photo since the water geo is seperate from the glass geo but even withthe straw aamplified like that is still says water to me.
  • 1
  • 12
  • 13
  • 14
  • 15
  • 16
  • 58
render engines and Maxwell

"prompt, edit, prompt" How will an AI r[…]