[rusteberg]

ok, so here's a different approach without sss or volumetrics

[jfrancis]

Some nice ideas here. I'm not sure my framing is close enough to try them properly this time around...

[eric nixon]

I tried a MB object 'catching the light'. It was a bit tricky to setup the movement + timing, and the result is not very good, so this is more of a warning, dont try this, kind of a post;

[eric nixon]

Another test with rough glass outer tube (hidden to gi) and a hack emitter map which only works for neon shapes that are 2-dimensional, which gives a darker saturated edge.

[eric nixon]

Had to tweak that setup by adding the 'hidden-to-cam emitter outer tube', and deleting that other 'hide to gi' tag that was on the glass before, also dropped the glass roughness to 72, and the ND to 2.3

Basically re-creating the setup mentioned in my first post.... here is a packed mxs file;

http://www.mediafire.com/download/svy5r ... olution.7z

8min render sl14;


That stripe you can see on the sign-board is where the material changes from metallic to diffuse, to show that the lighting works for both surfaces.


Finally, just found this very old maxwell 1.1 render (no idea why its upside down..) which shows just the 2d emitter mapping hack + scattering. Will render slower but its an easy setup.

Francis that thai girl looks seriously stoned/ Rohypnol.. The neon (how did you make it) looks great btw.

[eric nixon]

More detailed model with the same setup as above, tiny tweak to glass transparency;



Model by 'the pixel lab' - thanks for that

[jfrancis]

Very nice

[eric nixon]

Thought this would make quite a nice asset/RnD image. I think I chose the wrong shade of blue for the word COLD..
Making the letters was actually fairly tricky because I'm using an mxi type emitter (minor fakery to boost the saturated egdes) which means that the polygon density will affect the emitter brightness. So it took about an hour just to build those letters...

[jfrancis]

The intermittent pattern in the orange is particularly nice.

I always forget about mxi emitters, so hearing polygon density would affect brightness is a bit of a surprise to me.

[eric nixon]

Yeah, that pattern was a total fluke, it is because of the uneven poly sizes. It looked nice so I just left it that way, the way the neon fades at the end was done on purpose, and is controlled by the scaling down the extrusion at towards the ends.