Wine bottles

[hotep]

Wine bottles. HDRi lighting only.

[jure]

sweet! It's looking really nice! Congrats!

[salvatore]

very good render!!
i like it

[thomas]

Very nice. Two questions:

1) What was the rendertime (at what resolution)?
2) What material setup did you use? Dielectric? Nodes? Just plain old LW channels with gradients on incidence angle?

Thanks for the info!
Thomas

[hotep]

Very nice. Two questions:

1) What was the rendertime (at what resolution)?
2) What material setup did you use? Dielectric? Nodes? Just plain old LW channels with gradients on incidence angle?

Thanks for the info!
Thomas

1) On a triple core AMD Phenom the rendering process took about 8 minutes @ 600x600 pixels resolution.
2) For both glass and wine I used a very simple dielectric: just color, absorbtion and IOR.
Light setup: piece of cake, only an HDRi on Image World, no lights at all.
Kray did the magic

[thomas]

Wow! That's really good results and rendertime. I tried Dielectric last month but gave up because it was soooo slow in native LW. I didn't know Kray rendered Dielectric as well. Too bad there's just no decent way to render caustics in LW.

I don't supposed you feel like sharing the scene so we can learn?

Cheers,
Thomas

[hotep]

I don't supposed you feel like sharing the scene so we can learn?

Cheers,
Thomas

Sorry, I can't.
It's a test for a product shot. Anyway, I've just done another test, and with the same settings the rendertime is 4 minutes and 13 secs.
About dielectric: Kray CAN render dielectric, you have only to crank absorption WAY up.

[jnom]

Looks good. I also have long render time with glass with airpolys.
I think I will try to render with dielectric next.

[jure]

Looks good. I also have long render time with glass with airpolys.
I think I will try to render with dielectric next.

Don't use airpolys! This is ancient method that does not work with Kray or new LW versions.

[Janusz Biela]

btw DIELECTRIC
This is perfect Node very fast and realistic , but with some "bugs" which Kray can not accept (white light leak)
I very suggest use Dielectric ... but still we have problems under render.

hotep: did You use caustic?

[hotep]

hotep: did You use caustic?

Yes. You can see a caustic near the first bottle, it's delicate.

[thomas]

Janusz, I'm consfused. You say:

"btw DIELECTRIC This is perfect Node very fast and realistic , but with some "bugs" which Kray can not accept (white light leak)
I very suggest use Dielectric ... but still we have problems under render."

Help me out here: are you saying we should or should not use Dielectric?

p.s. I missed the caustics in the image. Cool that they are in there, and they look okay. It's probably impossible to get really nice sharp caustics though, no?

p.s.2. So what is the new way to do glass, without the air polys then? I've uploaded a JPG with two setups. The left one makes rays make a transition from one material to the other, the right one makes a ray 'exit' the material it's in before entering a new material and probably letting it enter a tiny tiny tiny space of air before entering a new material. Which one is correct now, for Kray and LW? Also, do I leave a tiny airspace in between masses (between glass and wine, for example), or do I let the points coincide? And if I let the points coincide, can I merge them? I always seem to have a lot of problems with internal reflections, way to bright white areas and noise when I do setups like this...

Thanks for any help / info!

[hotep]

Thomas, let's try with something more... easy.
The best way (in my opinion) to get a perfect glass is the following: build a sphere, now build a second sphere a bit smaller than the first one. Place the second inside the first. You're done!
Assign a clear dialectric to the first ball and a coloured dielectric to the second. Don't forget to assign a proper IOR.
Let us see what you get

[jure]

Yep keep it simple.

Outside bottle make regular glass material, inside bottle cut the bottle at the height of the liquid and make it liquid material. Done. No airpolys no air space, no double geometry.