I moved the rendering to here:

http://isometricland.com/gearhead/ghpanotour_spinner_test.php

Right now (not in the above image yet) I'm experimenting with radiosity 
and a bit of atmospheric haze. What would good values be for scattering 
media? I'm assuming the density should be greater toward the edges of 
the cylinder, but by how much? Should the density increase linearly or 
exponentially?

Thanks!


-- 
http://isometricland.com

Post a reply to this message

Le 2010-06-04 17:04, SharkD a écrit :
> I moved the rendering to here:
>
> http://isometricland.com/gearhead/ghpanotour_spinner_test.php
>
> Right now (not in the above image yet) I'm experimenting with radiosity
> and a bit of atmospheric haze. What would good values be for scattering
> media? I'm assuming the density should be greater toward the edges of
> the cylinder, but by how much? Should the density increase linearly or
> exponentially?
>
> Thanks!
>
>

For the density. It would remain almost constent near the axis of 
rotation, then increase somewhat faster as you get nearer to the outer 
surface. I'd go for proportional to the square of the radius. At least, 
it looks reasonable to me. Use a cylindrical pattern:
density{cylindrical poly_wave 0.5 density_map{[0 1][1,0.3]}}

Your cloudy "sky" looks out of place here. A large transparent area, 
while good to get some natural light, is a great structural weakness, 
not to mention that that area is lost from the ocupancy potential of the 
station. If it's just a decoration for the inabitants, it's even worst. 
Also, as the illumination seems to come from that bright axial beam, it 
should be constant everywhere.

If you want your buildings to taper toward the top, I suggest using 
prisms instead of simple boxes. But, unless a building is particularly 
broad, there is no real reason that they should'nt have a constant width.



Alain

Post a reply to this message

On 6/4/2010 9:05 PM, Alain wrote:
> For the density. It would remain almost constent near the axis of
> rotation, then increase somewhat faster as you get nearer to the outer
> surface. I'd go for proportional to the square of the radius. At least,
> it looks reasonable to me. Use a cylindrical pattern:
> density{cylindrical poly_wave 0.5 density_map{[0 1][1,0.3]}}
>
> Your cloudy "sky" looks out of place here. A large transparent area,
> while good to get some natural light, is a great structural weakness,
> not to mention that that area is lost from the ocupancy potential of the
> station. If it's just a decoration for the inabitants, it's even worst.
> Also, as the illumination seems to come from that bright axial beam, it
> should be constant everywhere.
>
> If you want your buildings to taper toward the top, I suggest using
> prisms instead of simple boxes. But, unless a building is particularly
> broad, there is no real reason that they should'nt have a constant width.
>
>
>
> Alain

Thanks for the comments. Do I need to scale the density file to the 
scale of my scene?

The "sky" is also meant to place the opposite side in shade in order to 
produce "night". I've tried other methods like placing a sleeve around 
the lamp, but due to radiosity the opposite side is then never fully in 
the dark. I.e. the "bright" side casts reflected light onto the supposed 
"dark" side.

As for loss of occupancy, you're assuming people want to live in zero G. 
Only the areas near the surface have suitable "gravity".

The "sky" does look ugly though.

-- 
http://isometricland.com

Post a reply to this message

Le 2010-06-04 21:52, SharkD a écrit :
> On 6/4/2010 9:05 PM, Alain wrote:
>> For the density. It would remain almost constent near the axis of
>> rotation, then increase somewhat faster as you get nearer to the outer
>> surface. I'd go for proportional to the square of the radius. At least,
>> it looks reasonable to me. Use a cylindrical pattern:
>> density{cylindrical poly_wave 0.5 density_map{[0 1][1,0.3]}}
>>
>> Your cloudy "sky" looks out of place here. A large transparent area,
>> while good to get some natural light, is a great structural weakness,
>> not to mention that that area is lost from the ocupancy potential of the
>> station. If it's just a decoration for the inabitants, it's even worst.
>> Also, as the illumination seems to come from that bright axial beam, it
>> should be constant everywhere.
>>
>> If you want your buildings to taper toward the top, I suggest using
>> prisms instead of simple boxes. But, unless a building is particularly
>> broad, there is no real reason that they should'nt have a constant width.
>>
>>
>>
>> Alain
>
> Thanks for the comments. Do I need to scale the density file to the
> scale of my scene?
>
> The "sky" is also meant to place the opposite side in shade in order to
> produce "night". I've tried other methods like placing a sleeve around
> the lamp, but due to radiosity the opposite side is then never fully in
> the dark. I.e. the "bright" side casts reflected light onto the supposed
> "dark" side.
>
> As for loss of occupancy, you're assuming people want to live in zero G.
> Only the areas near the surface have suitable "gravity".
>
> The "sky" does look ugly though.
>

Yes, you need to scale the cylindrical pattern to the radius of the scene.

For the "night", in an environment such as that station, there is never 
one, and it's quite realistic that there should be no dark side.
If they want a "night", they should do it by turning off the central 
glowing beam, or dim it down. There is absolutely no reason to have 
areas dark while other are brightly illuminated, it's the same hour and 
time of the day everywhere. Have you seen Babylon 5? It's a very good 
example of a spining cylindrical space station.

For the loss of ocupancy, I mean loss of living surface due to the "sky" 
area, not going all the way up to the axis.


Alain

Post a reply to this message

On 6/7/2010 2:22 PM, Alain wrote:
> If they want a "night", they should do it by turning off the central
> glowing beam, or dim it down.

Good idea.

I uploaded a new render (same URL). I'm still having problems with the 
scene looking "artificial", or like a miniature model. I can't quite put 
my finger on it.


-- 
http://isometricland.com

Post a reply to this message

Still looks washed out... :(

-- 
http://isometricland.com

Post a reply to this message

Attachments:
Download 'gh_scene_spinner_cubemap_h.png' (444 KB)

Preview of image 'gh_scene_spinner_cubemap_h.png'

Le 2010-06-07 16:58, SharkD a écrit :
> On 6/7/2010 2:22 PM, Alain wrote:
>> If they want a "night", they should do it by turning off the central
>> glowing beam, or dim it down.
>
> Good idea.
>
> I uploaded a new render (same URL). I'm still having problems with the
> scene looking "artificial", or like a miniature model. I can't quite put
> my finger on it.
>
>
All those cars and the streets don't help. I'd expect a comunal 
transport system, part on grount, part suspended, to accomodate the 
cityzens transportation.
I still have a dificulty with your clouds and "blue sky". If there are 
some clouds, they would tend to gather into a cylinder around the center 
beam, but never touching it.
Any blueish area due to the athmospheric scattering would be in the 
direction of the central beam relative to the observer. Now, it's about 
120° from where it should be.


The location IS artificial by definition. You may add some light fog, it 
may help to give a sense of scale.

You may try puting the camera on the roof of a building or at street level.


Alain

Post a reply to this message

I adjusted all the pigments to compensate for assumed_gamma. The scene 
is no longer as washed-out and bright.

Another problem I haven't fixed yet are the streets. For some reason 
they show up as black when they instead should be gray. I don't know 
why. I thought I fixed all coincident/overlapping surfaces.

-- 
http://isometricland.com

Post a reply to this message

Attachments:
Download 'gh_scene_spinner_cubemap_i.png' (406 KB)

Preview of image 'gh_scene_spinner_cubemap_i.png'

Better indeed.

From looking at the buildings, I feel that the cylinder should have a much 
larger radius. People will get claustrophobic here, with the impression that 
the buildings are falling on their heads. :-)

I think that the sense of scale would be improved with a better balance 
between cylinder dimensions and building dimensions. I suggest a larger 
cylinder and (much more) scaled-down buildings/living space. Add some 
atmospheric media or a subtle fog for suggesting distance and you will get a 
much more believable scene.

Thomas

Post a reply to this message

On 10/06/2010 8:28 AM, Thomas de Groot wrote:
> Better indeed.
>
>  From looking at the buildings, I feel that the cylinder should have a much
> larger radius. People will get claustrophobic here, with the impression that
> the buildings are falling on their heads. :-)
>
> I think that the sense of scale would be improved with a better balance
> between cylinder dimensions and building dimensions.

I agree.

I suggest a larger
> cylinder and (much more) scaled-down buildings/living space. Add some
> atmospheric media or a subtle fog for suggesting distance and you will get a
> much more believable scene.
>

If Mike does that then it may be realistic but it will be boring to look 
at. IMO realistic space scenes show one feature while hiding most others 
due to differences in scale. Do you remember the posts from a few years 

Having said that the ground could be s bit further away ;-)


-- 

Best Regards,
	Stephen

Post a reply to this message