I'm doing a work project creating a series of stills showing a glass lab beaker half-filled with water, then adding white powder, then the water being stirred and becoming slowly clearer.

I was looking in the libraries on my Mac's A:M over the weekend and there were maybe six glass materials, but when I started work at my office pc this morning, there are only three glass materials in the library, Fake Glass MR, Liquid Glass MR and Perspex MR.

the assignment is tricky because at most points in the process everything needs to be transparent. I'm going to composite a toon line for definition of the beaker but I wonder if someone can make a good recommendation for a glass material that will approximate laboratory glassware.

Well, I know Robert has an awesome one from seeing his syrup tests. Bug him for it cause it's a cool glass!

glass is easy.

model your beaker so the walls have thickness.
make it slightly less than 100% transparent.
give it some very slight color like green or blue
give it Index of refraction more than 1.0
give it very slight reflectivity.

You test the object in its setting and vary all of those to suit the look you want.

Ooh, that index of refraction is the thing! I was already on my way with a gradient material with an edge gradient and added Robcat's settings to the transparent part. Works pretty well I think! And the toon line won't be necessary after all.

EDIT: one other change was to treat the interior of the beaker slightly different from the exterior so there wouldn't be double specularity. I think that helped too.

Reduce the spec size and increase the spec intensity a bit. Right now it looks a bit milky. It might be the transparency isn't high enough.

If your background will be blank like that then IOR may not be a factor. Your gradient combiner approach may get you more.

the IOR definitely makes a difference in the way the grey shading happens around the lip. I'll keep experimenting.

Try more IOR...like 1.5 or 2!

I tried that, and it's way too much. I've got it at 1.02. Any more than that and it starts getting wierd.

How the final render comes out also really depends on how you set everything else up. If you get it too realistic, the render times go up pretty fast. I didn't even finish the render below since it takes 7 minutes on my laptop for every pass. Thought I'd give it a shot though.

The IOR of actual glass (Well crown glass (ie lens glass)) is about 1.33, air is 1.000000001 and vacuum is 1.

Here is the super awsome best list ever of IOR's :
CG Talk IOR Thread

Photoman

Here's the next version, with a little more detail. I would sure like to get some motion blur on that paddle. What I've tried so far is to rotate it, convert to Euler interpolation, give it a good spin and render at 9 passes with 50% motion blur and this is what I got. Should I just increase all these settings?

In the process I'm animating the paddle is supposed to be spinning at 200rpm throughout.

humour me on this. Render it as a short animation and see if the blur works. Something I've found is that if I'm rendering a single frame only, a motion blur does not render, but if I'm rendering multiple frames, the blur does render. Then if you render to multiple TGA files or something, you could pick one with a nice blur on it.

That's worth a try! I'll post something later in the day.

EDIT: Actually it's working just fine without rendering a sequence. I just cranked up the revolutions and motion blur.

Here's the final sequence. It worked out just fine for our purposes but I'm sure it will change!

or perhaps this list...Table - Index

i particularly like the fact that they list a value for beer!

i had this problem with glass a few years back took me like 2 weeks to figure it out
this was pretty much the result

you can find the beekers and tubes here