Ring World

[dandelO]

All very cool. I especially like Ulco's aeroplane race track. I scrapped most of what I was playing with but I have managed to get a working clouds/atmo solution to start again with. Here's some failures, some with working atmo, some without. I'm starting again, I've learned a bit here...

[TheBlackHole]

You might need more samples in the atmosphere and clouds. I'm seeing a bit of grain.

[dandelO]

* A lot of grain. They're just muck-arounds, I paid no attention to render settings here, just set 0.5 quality and have been playing with various settings in this testing.tgd. Nothing more. I agree they're just nasty screenshots.

[cyphyr]

Looking good !!
Have you noticed how you can't see the tops of the clouds once they start climbing the opposite curve of the ring?  I don't think there's a real solution to that (although it just occurred to me that 2d clouds might not have this "optimisation").
Are you working with planets, spheres or planes? I'm back on planes now as they cast propper shadows (well mostly!)
Keep em comin'
Richard

[Dune]

Very nice, Richard! I like the way it seems to float in 'space'. And the greenery is getting great.

I've rendered another version, but cut it off, as it was getting too late. I added a concrete living quarters' wall (just a projection on a plane). The inverted sphere works perfectly to complement the tunnel base, except for a tiny seam, which I still hid here behind a concrete wall. I'll try something light with glass roof as well. I added a car on the road, but due to the immense scale it's hardly visible (ah, and it wasn't rendered yet, here). It's title is now 'Perpetual Hot Pursuit' (the ring, you know). I'll also try a bending/curving ring  

I found that keeping the size of the light source very small (0.1) the soft shadows work good. But the 'glass' indeed stops all light from emitting. Too bad.

[Tangled-Universe]

Very creative work here Like that last one Ulco!

Martin

[dandelO]

Another play with Jan's method with 2 planets for ceiling/floor surfaces. A third invisible planet provides the atmospherics.
Racers from Google 3D Warehouse.

Frame 2 of 3 for the blur effect:

[attachimg=#]

[attachimg=#]

Great fun!

* Probably edit this for some contrails and such...

[Zairyn Arsyn]

F-zero rules.   

Another play with Jan's method with 2 planets for ceiling/floor surfaces. A third invisible planet provides the atmospherics.
Racers from Google 3D Warehouse.

Frame 2 of 3 for the blur effect:

[attachimg=#]

[attachimg=#]

Great fun!

* Probably edit this for some contrails and such...

[Dune]

Nice one, Martin. Did you hang the invisible planet underneath the other two too?

Here's my next one, with a procedural roof (some blue nodes and a default shader). I wanted to have glass in it, but that didn't work as I wanted. The glass reflected the inside of the tunnel, so became very dark towards the top (only the sides were transparent). With less reflection it didn't look like glass anymore. When adding another roof, turned inside out, it wasn't visible... of course.
I did another interesting experiment, but TG went totally bazurk; when I made the refraction index a negative 2, some weird prism like colors were spread over the inside of the tunnel. But already the preview rendered it completely different each time I changed almost nothing in something else; sometimes resulting in some totally black areas, or dark blue, changing location. Very weird.
I'm getting a bit fed up with this now; I just got another idea for a free falling waterfall this morning, so here's probably my last. Unless I change my mind. Anyway, this is not really a ring world anymore.

[dandelO]

Nice one, Martin. Did you hang the invisible planet underneath the other two too?

No, the planet is the same coord's as the ceiling/floor planets but it has only a 1m(if I remember correctly) radius. I suppose I could have just used the ceiling planet to add atmospherics to and just edited the heights of the cloud layer for that but I found it easier to make it a separate parameter. The ceiling/floor is 2x1000m radii planets, I used the third planet's cloud layer height fixed at about 900m and a thickness of (close to) half of that so, the atmosphere is like a band that fills the tube, this way the top of the cloud layer is well above the ceiling of the tube all the way around(it can't be clipped by that annoying bug this way) and can be expanded/contracted by using just the depth control. I scrapped using an actual atmosphere node in favour of a cloud layer because, it's easier to adjust and you don't have to invert the blue sky/haze parameters to make the atmo work 'correctly' at the south pole. A cloud layer with a subtly contrasted density fractal provided a nice and easy haze, better than any atmosphere I'd played with.

It's a fun idea to play with, I like the idea of using independent planet nodes as sculptable surface objects.

Richard, you mention using plane objects. How are you bending those into a curve, by using a displacement shape much bigger than the size of it before making it concave/convex with more shapes?

Great thread! Cheers to everyone here, especially Richard, Jan, Ulco and of course, Avkhatri for bringing it about in the first place.

[cyphyr]

I think were exploring the outer reaches of the Terra-verse here, that would make us "Terranaughts" ?
I've really not been able to give this the attention I would like, I'm at my folks helping with hospitals and other equally tedious stuff

Yes the Planes are distorted by shape shaders, one for the basic curve of the ring and one for the cross section (I'm working on one now that uses three for the cross section). Of course being on the outer edge of terragen physics not everything works as one would expect. Altitude based shaders don't really work at all. No mater what flavour is used altitude based shaders always behave as if "use y for altitude " has been ticked. There's also the cut off of atmospherics (which is not a problem if you use an enclosed "roofed" ring.

The "up" side of using planes is that they can cast proper shadows since no fakery is being used with pseudo transparent surfaces.

I've certainly not finished with this but I still expect progress to be slow

Attached is a basic set up showing where I'm at at the moment. Maybe I'll get a render done on this ol' dog of a laptop some time soon !



Richard

ps: By the way I spoke to my father who is a physicist (really) and he said that in order for a world like mine to work the gravity would have to be induced by centripetal motion, it would have to be spinning.
The formula to work out how fast is:
rΩ²=f
where
r=radius (measured in metres)
Ω=speed  of rotation
f=force (measured in m/sec/sec)
Therefore my 500km radius ring would be spinning at ... .. . ≈ 70.7 m/s or 158 mph!!

you can tell I waiting for a render to finish ...

[choronr]

Thank you Richard, this unknown territory (for me) is very intriguing; and, look forward to working with it. If I run across any new discoveries, you'll hear from me.   

[Dune]

Altitude based shaders don't really work at all.

Try this: locate a camera at the centre of your planet and use a distance shader... it works perfectly.

[cyphyr]

It works perfectaly for planets, I've used it
Unfortunately it does not work for the plane system I'm using at the moment.

... .. .

The best solution I have found so far is a grey scale gradient projected in plan y.

Richard

[Seth]

"Tunnelworld-5top-23-10-10.jpg" was great Dune !!!!