Screen Shots for next version

[Urwumpe]

I don't know if a global climate model would really be needed. The question is just, how much realism we want. I think for good procedural clouds, it would be enough to have a enough textures available to map all tiles without a repeating pattern.

If you look at satellite images, you find so many repetitions and similar patterns, it should be realistic to create a procedural cloud cover with a pretty limited number of textures.

I have to find this paper on procedural forests again, the drawings inside it explained the aspect better.

EDIT: Wow - Google is scary sometimes. The third hit was already the right paper.

http://www-evasion.imag.fr/Publications/2004/DN04/

 

[martins]

On this subject, will the next version include functions to set a local atmospheric parameter for vessels? The only thing I have to say is missing in Orbiter is that the wind is always zero. It would open the door for many atmospheric improvements by 3rd party enthusiasts (such as myself :hotcool: ) if we could set the local pressure, temperature and density returned to a ship's call to functions involving these parameters by overriding what is fed by the global planetary model. By doing so, it would then become possible to have a module generate zonal temperature and pressure gradients, on a very rough scale at first, so that the temperature is not always 15 C everywhere on Earth everyday of the year.

Something along the line of:
vessel->setAtm(temperature, pressure, density);
vessel->setWind(vectors...);

Then, the weather modules would be responsible for checking on the vessel's positions (lat, long, alt), lookup/generate the proper values and set them with the functions in my example above. If someone wants to attempt fancy stuff such as getting the data from the Internet in real time, or generate clouds around vessels based on water vapor content from a model or real data, it's up to the weather module code to deal with it, from -outside- Orbiter's core.

All good points, and worth keeping in mind, but probably not for the next version - unless you have the patience of waiting for another couple of years for the next update ...

 

[Urwumpe]

All good points, and worth keeping in mind, but probably not for the next version - unless you have the patience of waiting for another couple of years for the next update ...

99.9% of spaceflight is just waiting for the next maneuver.



Just kidding! :lol:

 

[martins]

If you look at satellite images, you find so many repetitions and similar patterns, it should be realistic to create a procedural cloud cover with a pretty limited number of textures.

I think you are correct that on a local scale there are indeed many repetitions and similarities between cloud patterns. But my point is that this randomness doesn't extend to global scales. On a global scale, the clouds are everything but random, and in my opinion a cloud layer that has been generated from a procedural algorithm on a local scale would not look realistic on a global scale.

But I am of course more than ready to be convinced otherwise, if someone can come up with a demonstrator ...

 

[Chipstone306]

It looks beautiful so far Dr. Schweiger! My only question is whether the physics (NEWTON) will change in the new version. Keep up the great work!

 

[Artlav]

I think you are correct that on a local scale there are indeed many repetitions and similarities between cloud patterns. But my point is that this randomness doesn't extend to global scales. On a global scale, the clouds are everything but random, and in my opinion a cloud layer that has been generated from a procedural algorithm on a local scale would not look realistic on a global scale.

One way to give the clouds at least a perceivably higher resolution could be a sequence of microtextures applied over a main texture, like that:

 

[simonpro]

Actually I think Orbiter could be substantially improved (visually) by adding a second cloud layer. The main thing that bugs me about the clouds is that they're all flat and all at the same height.
A little bit of variation (maybe a higher layer with some cirrus) would make things look nicer.

 

[Mission_CDR]

There is one more screenshot for the next version! http://orbit.medphys.ucl.ac.uk/images/Florida_lvl14_1600x1200.jpg

 

[markl316]

Thanks for your reply Dr. Schweiger.
I liked Artlav's idea about constantly more detailed microtextures as you got closer. The microtextures could be repetetive.

Also, here's what I meant about a more turquois ocean.
http://images.google.com/imgres?img...rev=/images?q=space+shuttle+orbit&gbv=2&hl=en

and

http://images.google.com/imgres?img...rev=/images?q=Space+station+orbit&gbv=2&hl=en

 

[Scarecrow]

I'm wondering about 2 little things with the DX7 graphics engine that I hear is coming along:
1. Will there be support for multiple light sources, such as night lights on a launchpad?
2. Will vessels be able to cast shadows on themselves, the same way I could look at the sun, and then cast a shadow on my face with my hand? Might they even cast shadows on each other? Mabye?

 

[willy88]

Idea for next version: bump mapping, specular reflections & ground textures

Orbiter, admittedly, has somewhat dated graphics. Three fairly major things that I think would improve it quite a bit is bump mapping, specular reflections and ground textures.

Bump mapping would look quite nice on old beaten up spacecraft, or perhaps the Space Shuttle ET.

Specular reflection would look excellent on windows or maybe the water.

By ground textures, I mean alpha-transparent textures that would blend in with the surface-tiles and would appear when you get very close to a planet's surface. For example, Earth could have a grass texture, like this:

Of course, it would be a bit strange having grass in the desert, so there could be different textures for different areas of a planet.

 

[jedidia]

I think Artlavs Orulex has taken promising steps into this direction allready, and will probably get better. And since all that stuff like bump-mapping etc. will be possible in the OVP, there's not much sense in it with plugging it into the core just right now. Let's be pacient and wait for a stable OVP-release.

 

[SkyCaptain]

By ground textures, I mean alpha-transparent textures that would blend in with the surface-tiles and would appear when you get very close to a planet's surface. For example, Earth could have a grass texture,

Or have a complete synthesizing ground texture system with sampled data sets like FSX:
http://www.fsinsider.com/developers/Pages/GlobalTerrain.aspx

 

[tblaxland]

I think Artlavs Orulex has taken promising steps into this direction allready, and will probably get better. And since all that stuff like bump-mapping etc. will be possible in the OVP, there's not much sense in it with plugging it into the core just right now. Let's be pacient and wait for a stable OVP-release.

Ditto for multiple light sources and shadows on vessels.

 

[aldarion]

Future Orbiter with lvl14 textures!

http://orbit.medphys.ucl.ac.uk/images/Florida_lvl14_1600x1200.jpg

It will be better than I thought. Enormous amount of data will be required, though.

I'm ready to dowload it!

 

[simonpro]

Level 14 is only over specific areas.

 

[ryan]

Can you guys imagine how much MB or maybe even GB will be, lvl 14 textures will be quiet big, i want to know how the Doc is gonna do that.

 

[cjp]

Wow... I want to have that...

Where do you get free maps with that amount of detail? I found NASA's blue marble, but this is a lot more detailed.

 

[simonpro]

Lots of places. MODIS is more detailed, so is MSG, landsat and a tonnes of other stuff.

 

[markl316]

Will the landsat team (www.sethden.com... I don't know the exact url) make a level 11 texture in addition to the level 11 that will be on the orbiter download site? Their level 10 looks completely real from about 850 km.