[jfrancis]

There seem to be two kinds of neon-style tube lighting. One where the tube is opaque, like a solid emitter,(I think it might be internally coated with some kind of phosphorescent powder), and one where the glass tube is transparent and the wall thickness is apparent and it contains a soft-edged glowing gas inside.

I'm doing the former right now with an extruded tube as an emitter (with a reflective AGS layer on top in case reflected highlights are visible), but I'd like to try the latter.

Any suggestions?

I guess I need to avoid enclosing a tube-shaped emitter inside a pipe of dielectric glass. So AGS glass?

I haven't played with particles or volumetrics much yet but dare I ask if I can make a light-emitting gas contained in a walled glass pipe?

Obviously there are composite-based solutions, but I'm curious as to rendered options.

[jfrancis]

work-in-progress, but it's looking like this now and I want to increase the realism as much as possible.

[eric nixon]

Ok, what you want is three tubes, the inner tube is an emitter, the middle tube is glass, the outer tube is an emitter, which is hidden to cam, and also if possible make it have invisible back faces so that the middle glass tube picks up reflections from the scene.

I would need to look for my old files to check all that though..

Let us know if it works...

[jfrancis]

Sounds interesting; I'll check it out.

Thanks!

[jfrancis]

Here's an experiment based on your suggestion...

A fairly thin tube emitter in a thick tube of rough, red glass of rapid attenuation.

No outer emitter (yet)

[eric nixon]

I had an attempt, at what i suggested, renders quickly enough, but doesnt really look right;

[rusteberg]

I guess I need to avoid enclosing a tube-shaped emitter inside a pipe of dielectric glass. So AGS glass?

i don't think you need to worry about that....















you could add some lens diffraction to give it a little bit of glow if you wanted to....

or, you could put a little bit of sss in the glass ( between 5-20 Scatter Coeff with ~ -.75 Asymmetry and you'd get something like this: (but i don't know that you necessarily need to do that...)

[jfrancis]

nice

[eric nixon]

If you look at this photo (which has prob been post processed) most of the red outline glow comes from the damp air, (something which can be best achieved as a post effect OR a motion blurred ring emitter) apart from that there is a tiny red outline on the neon itself which shows up more at the dimmer areas, like the ends of the tubes. I think the three tube setup could give that outline effect, by making the tubes very simialr in size and using very short attenuation. The fade for the ends needs some emitter mapping.



Note; there are lots of different types of neon with different looks ofcourse.

[rusteberg]

If you look at this photo (which has prob been post processed) most of the red outline glow comes from the damp air, (something which can be best achieved as a post effect OR a motion blurred ring emitter) apart from that there is a tiny red outline on the neon itself which shows up more at the dimmer areas, like the ends of the tubes. I think the three tube setup could give that outline effect, by making the tubes very simialr in size and using very short attenuation. The fade for the ends needs some emitter mapping.

to me it looks like that reference was shot at a low shutter speed with a hand held camera and crappy glass (lens) based on the blur i see..... based on that observation, it looks like there's a whiff of smoke passing through the upper portion of the neon sign which is streaking more light (notice the gaps) than if it were a still - shot on a tripod at let's say 5 o'clock am once the bbq pit had been closed for a while on a damp southern night along the bayou.

you could mimic that effect with a volumetric bounding box, and even amplify it with lens diffraction and scattering if you wanted to (as shown below), but........

volumetric render without lens diffraction/scattering


volumetric render with lens diffraction/scattering

[eric nixon]

Thats looking good, its just missing the paleness of the neon with its saturated thin outline.

Just had a thought that you could use MB to achieve the glow, not with a moving emitter like I said before, but with a geometry plane (using a soft edge alpha map) and have that plane pop in to view behind the sign just long enough to give some glow.

A volumetric surround seems easier and more promising though.

[jfrancis]

Interesting.

[jfrancis]

I haven't played with particles or volumetrics much yet but dare I ask if I can make a light-emitting gas contained in a walled glass pipe?

A photo I found online has that 'clear glass / soft-edged gas' look that is also nice.

Works well in closeups.

[hatts]

rusteberg I think your last two tests are really good, and could serve two different purposes; the one with SSS looks really good up close and that volumetric one looks great from a distance

[eric nixon]

soft-edged gas' look

I hope the next version of maxwell volumetrics has a 3d soft edge functionality (increased density per depth).