All posts related to V2
By wow001
#328188
Attach an image of a glass handrail lit from below and was unsure of the best way to recreate this effect.

Image
User avatar
By Bubbaloo
#328192
You can fake the effect with emitter edges and up-direction lighting onto glass with a bit of roughness, or you can try to model it correctly and see if Maxwell handles the caustics ok. The second option will probably clear slower though.
User avatar
By tom
#328221
These glass panels have rough edges and a clear surface. It shouldn't be a problem if you model and materialize accurately.
By wow001
#328893
thanks for the replies.
what do you mean by modelling accurately?
creating the edges separate from the main glass?
do we need to use sss in the glass and two different glass material edge and body?

our first attempts are not really working.
User avatar
By Bubbaloo
#328894
wow001 wrote:what do you mean by modelling accurately?
I meant model the glass panes, and shine some emitters up at the glass like the actual product. Have you tried this realistic approach? What methods have failed so far?
By wow001
#328935
attched our progress so far.
we have modelled the glass as a solid 16mm thick and 1100mm high.
then we did a second version of the glass handrail with 3mm solid edges so we could apply a second material.
we had to use a glass material with sss and i placed the emitters under the glass but not touching to get the our closet result.

attached the maxwell studio file (2.1) to see our construction so far.
http://www.wow.co.uk/downloads/maxwell/2stair_test.mxs

Image

Image

Image

Image

Image

Image

Image

Image
By brodie_geers
#328965
I gave it a shot yesterday with fairly similar results. I wasn't able to duplicate your photo. Using a glass with roughness or an SSS material I could get something approximating a sandblasted glass (if you let it render for a very long time), but it still wasn't like your photo. With the SSS, I think there is a known bug right now so it may simply be impossible with the current version of Maxwell.

The main problem I had was that the glass in your photo doesn't appear to be sandblasted. You can see through it pretty well, however the light is still clearly visible in the glass. It's hard to tell since the photo isn't the best quality, it almost looks like it's covered in a layer of dust which is actually catching the light, I'm not sure.

I did get the best results from using 2 different materials though. For the edges I used a glass with a higher roughness which seemed to catch the light better.

I'd post images but I never was able to produce anything worth saving.

-Brodie
User avatar
By Voidmonster
#328973
This seemed like a fun challenge, so I took a stab at it and I'm reasonably pleased with the results.

The 'accurate' glass version probably isn't practical, I let it render for 15 hours and it's still unacceptably noisy.

Image

I did a second version using AGS. I'd say it's practical, this render only ran for 2 hours or so.

Image

I learned a few things in the process, the most important being that the edges should not be a separate material. At least not a separate poly-group. Doing so, with the physically accurate glass, creates an open volume which renders very wrong, takes a lot longer and is generally nasty to look at.

What I did was to create a mask texture for adding roughness to just the edges of the glass (in the case of the curved glass I'm using, that meant UV mapping, but for planes projection mapping should be fine). The key is making the texture have the same proportion as the modeled part so that it doesn't get stretched in one axis.

Instead of using SSS to get the slight haze I just added a second diffuse layer at about 10%.

(also, yes, I know that the light balance is very different in the two renders above, I tweaked the look using multilight and aimed to have the same apparent glow on the edges instead of the same over-all lighting)

Edited to add:

It's way too hot here today for any rendering, or I'd redo the scene with a lot fewer lights. Looking at your example I see that each glass panel has only 3-5 lights beneath it, where my render has 32 per, which makes the glow very uniform. Also, I didn't use IES lights, so each of the ones in my scene are uniformly shaped and installed.
By brodie_geers
#329017
Tora, I think you win the contest. It's far closer to the intended effect than I was able to get and the noise clears quite well. I let a render run over night for fun. Got to SL 18.6 in 8 hours on my humble Intel Core Duo E8500 (3.16 ghz). I lit it with skydome and did a few light settings...

Image

Image

Image

-Brodie
Maxwell Rhino 5.2.6.8 plugin with macOS Tahoe 26

Grazie Ferdinando, prima di aggiornare il Tahoe 26[…]