Lighting Challenge #4: Bottle Collection


The glass (of the glass bottles) is looking good! The window glass needs work, that “crackle” texture isn’t very convincing, it’s very regular, and then it melds with the twigs of the tree you see through it and looks confusingly like a reflection of the tree outside in places. Speaking of “outside” the tree seems to come through the window a bit on the left. Then the metal parts of the window, that should be opaque, have a blurred shadow of the tree on them. It looks like maybe that was a compositing error, maybe putting a glow around the sky that bled over onto the metal, but still it looks strange.

Apart from the strangenesses of your test, it is looking good. Maybe some of the glass bottles could be going too dark, for bottles sitting right by a window like that, but the bump mapping is working on them. Why do you think the reflection of the 3rd from right bottle looks so bright and clear in the bottle behind it, when the bottle itself is so dark? Is it a shadow with too low a ray depth limit maybe? Or the shader needs more light absorbance?

(I’ve just been playing with some of those Maya raytracing parameters for glass. Funky stuff! Like the way any non-zero “surface thickness” cancels out the “light absorbance,” but light absorbance is mostly only seen in reflections and refractions of the bottles, not the bottles themselves…)



I am not certain my refraction is working.

The tutorial I read said (in lightwave) to copy and paste the geometry, flip it, and then assign the flip the air refractin value. The first set of geometry would have the 1.52 value.

I will go back tonight and try to isolate this before I work on the items Jeremy said to look at.


The bottles are already fully modeled in terms of having an inside and an outside, so no additional copies should be necessary. Just render them with some subdivision or polygon smoothing, as double-sided surfaces (if there’s a choice to turn off back-face culling and render front and back, although that might happen automatically for transparent objects.) 1.52 is a little high for glass, I see some refraction but not that much in your render, maybe because you did something else to counter-act it. You don’t need to worry about doing anything for the refraction level of air. A single IOR value should be used on the surface, and the entrance and exit refraction angles should will be computed by the raytracer as long as the bottles are rendered as double-sided surfaces. (Yes, even the Klein Bottle. :wink: )



Thanks Jeremy, I will go back to the original file, isolate the bottles, and make them double sided.

As far as the etching, I appreciate the comments. However, that was an accident if it wasn’t in the original geometry.

EDIT: Here is that site I was checking out


Thanks for the comments and suggestions Jeremy. I definitely agree with your comments and will be hopefully getting a chance to work on it a little bit more this coming weekend or maybe in my evenings at work…Thanks again!



Some simple test…



Jeremy,…at the moment all my shadows are simple 512x512 dmap shads. That accounts for a lot : ) Also the reflection coming from inside the room is a plane that leaves black space on the sides to be reflected.

I initially wanted to use rman for Maya and deep shadows to test the eval version, but the watermark was a bit of a pain. So i thought i might “fake” the whole thing. Like project fake caustics from a top light and so on.
A few of my lights also have gobos on them so thats what creates the light variation on the window ledge and the shadowing on the window frame (which right now is only a default blinn).
As far as the Maya raytracing parameters,… theres indeed funky stuff going on. A slight difference in shape from bottle to bottle produces totally different results. I have yet to play with light absorbance.

I will post some screenshots. Even with dmap shads there is at least one bottle that sort of works already : the first one on the left and right. I am still getting dark edges on it, but i kinda like where its going.

Some screenshots

and links to full res


Okay here’s my first attempt at this one. I just grouped the bottles together and threw a glass material on them to test the light setup. I’ll definitley put some more work into this.


sjmcc -

Thanks for inviting other people from the C4D Cafe. Your image looks a little too early to say much, just because I can’t tell if it’ll be indoor or outdoor lighting or what kind of setting you’re going for, and it doesn’t have much in the way of backgrounds or reflections yet. If the bottles are going to stay that green, then the shadows where light has filtered through them should be going green as well.



whoaw…master zap…very nice image you got there.

okay, heres my first take on the challenge. so far the shader works on some of the bottles. i only used one shader for all of them. when i come back tomorrow,hopefully i could texture each one of them right.
the lighting i think needs more tweaking. i used maya software render for this image. its a club scene.


also, (in maya) I had to manually turn on “visible in refractions” and “visible in reflections” in the render stats tab of each bottle…so I’m not sure if that was just the way my settings default to for some odd reason or if this is is turned off in the downloaded scene.


thanks jermy for one more nice challenge, i m going to start work on this n try to put some good woks.


I’m still having some problems with the glass material. I’m still not good with refraction setup. Gettin kinda fedup when i see something i don’t like after a 10 - 20mins render.
I’m gonna treat the bottles as opaque to check out their interaction with the environmental lighting for the time being.

Heres my trial run…


I like your comp and your lighting so far! You said you are using scanline Max render which is taking 10-20 minutes…I’m guessing from your times, that you are using advanced lighting w/ light tracer or radiosity with that? If that’s the case, have you tried Mental Ray? (I just recently switched over to Max from Maya about 8 months ago) so i’m not an expert by any stretch of the immagination, but from my rendering experiences in Max, a scene with opaque surfaces, such as yours has now, is quite a bit faster using MR rather than the scanline render. So if that’s the case, you might look into trying Mental Ray…and if you don’t know MR, it will take a couple days of frustration to learn, but in the end I think you’ll be happy that you struggled through it!

-Birn, not 100% sure how these challenges work so if you are supposed to be the one/main guy giving the crits/suggestions just tell me and I’ll shutup haha. I’m working on one too, so it’ll be in here soon hopefully!


I tried using a dome light setup for the scene along with some lights at certain key positions to ensure the contrast in the scene.
I didn’t use Mental Ray, LightTracer or Radiosity but I did turn on the raytrace global antialiaser and added a blur offset for nicer reflection… so that will acount for the longer rendering
Come to think of it, i shld have turned those off while i was still making admenments to the scene…
Silly me


Max - Good start. I agree with Justin that you could try using MR if you got a free license of it with Max, you ought to be able to find some software solutions that lets you develop the look of reflective, refractive glass bottles.

I appreciate everyone who posts feedback, it’s not my job to monopolize that. I’m setting up a scene myself, that should render tonight and I’ll post a version tomorrow hopefully.



Ok, I have learned alot about surface and part selection and object format conversion recently.

To summarize, Lightwave 7.5d does not read in obj with the surfaces/parts separated properly. I found that opening in Messiah first then saving as a LWO object is part of the solution.

If I do this first, then all the different items are visible for individual surfacing. However… and this is something I am still testing… if we have to copy and flip to set up exit refraction for air then that will still need to be done.

Jeremy gave us physically correct objects. However, based on the reading I have done, Lightwave does not “know” to bend the light ray again when it exits the glass.

I will be glad to post the Messiah “converted” version for anyone that has Lightwave.

ALSO, we can use the stats panel to manually select the parts/bottles to subpatch ect…

Let me know if I need to post for anyone.



HI Jeremy - can’t wait to see your results !
i’m working on the scene just to get some experience with caustics…
On my first test i’m getting weird caustics behaviour with filled bottles
(filled by me) so i need time to tweak on.
I’ll post something tomorrow -



Nice, you definately got that “club” feel. The red and blue light is working. You might even be able to stick with the bottles all having the same shader. The panel of light from underneath could be more convincing, though. I’d like to see more of a gradient from the white surface running up the window trim, from bright right above the white panel fading to dark a foot or so above. Most of the window trim could probably go darker, when you are away from the light. The glass bottles could also use some white near the base - maybe from a rectangular volume light scaled to include just the lower parts of the bottles, with ambient or other illumination in a gradient, and also white reflecting on the glass.



Here is an updated image. I still need to go back and subdivide some of the bottles and work on the window frame surfacing. Also, more refraction testing is needed. Comments and suggestions always welcome.