Ideas for next version of Celestia

[Mikeydude750]

2 words: Death Star

It would be so fun to blow up a planet....to relieve some stress.

[Redfish]

I'd like to see a bigger database for stars. 100,000 is nice, but i like to fly at high speeds, like 100 ly/s and then i'm out of the small cluster of stars around the sun. Why can't you use the tycho database with over 1 million stars? OR does only hipparcos' database include distance to stars?

Galaxies: It'd like them to consist out of stars only :) That would rule!
And heck i don't mind. Why don't you fill the rest of the galaxy with stars? It doesn't have to be realistic. I just want it filled. Realism can be added later when new data is available.

Local zoom function: When selecting stars it's quite hard to pick them with one try. I'd like to create a rectangle with my mouse around a certain number of stars, and then see a little screen with that part of the selection zoomed in, so you can select the stars more easily.

No sounds please. There aren't any sounds in space. And i too like the God's eye perspective :)

I'd like to see in addition to ly/s au/s and all the other speed scales Your speed expressed in lightspeed: of course it's nice to travel at ly/s but i have to do calculations constantly and still don't know how much 1 au/s is in lightspeed. Maybe it would be nice to see parsecs/s too?
Finally: Bigger textures, extreme detail on planets etc :P They just make me drool.

[marshead]

For a larger star database, there is always the Tycho 2 catalogue, with 2.5 million stars in it.

This is a revision of the data released at the same time as the Hipparcos catalog. Tycho was a second "mission" that used data from the scanners on the Hipparcos satellite to create less precise parallax measurements for many fainter stars. Initially, Tycho (1) was about 1 million stars, but a rereduction of teh data was conducted, and in the past year or so, the Tycho 2 database came out.

For those interested, it can be accessed at:

[url]http://cdsweb.u-strasbg.fr/viz-bin/VizieR?-source=I/259[/url]

The spectral types so conveniently contained in the Hipparcos catalogue do not exist for Tycho (1 or 2), but the color index (B-V) can be used along with the absolute magnitude (calculated from apparent magnitude and distance/parallax) to get a rather accurate size and color for each of the stars.

An increase from 100,000 to >2,500,000 would be spectacular! You could probably begin to see the local structure of the Milky Way based on that data alone!

Hope to see this in a future release. It's probably one of the easier integrations into the existing program.

Marshead

[chris]

For a larger star database, there is always the Tycho 2 catalogue, with 2.5 million stars in it.

This is a revision of the data released at the same time as the Hipparcos catalog. Tycho was a second "mission" that used data from the scanners on the Hipparcos satellite to create less precise parallax measurements for many fainter stars. Initially, Tycho (1) was about 1 million stars, but a rereduction of teh data was conducted, and in the past year or so, the Tycho 2 database came out.

For those interested, it can be accessed at:

http://cdsweb.u-strasbg.fr/viz-bin/VizieR?-source=I/259

The spectral types so conveniently contained in the Hipparcos catalogue do not exist for Tycho (1 or 2), but the color index (B-V) can be used along with the absolute magnitude (calculated from apparent magnitude and distance/parallax) to get a rather accurate size and color for each of the stars.

An increase from 100,000 to >2,500,000 would be spectacular! You could probably begin to see the local structure of the Milky Way based on that data alone!

Hope to see this in a future release. It's probably one of the easier integrations into the existing program.

Marshead

As far as I can tell, there's no parallax information included in the Tycho catalog. Am I just missing something? If it's there, I'll get to work immediately on converting the catalog for use with Celestia.

--Chris

[Rassilon]

Someday I would like to see more variance in star appearence...Or maybe script them in like planets...apply color for glow and flare...

Check out this recent hubble photo I clipped...


This is what I thought of for a red-giant like those...


Having script override the original source code for the stars...

[Stargazer_2098]

Another thing I wish in Celestia, or rather in the forum, I must say: Avatars, the little picture under our names giving a little decoration.
If there isn't enough space for avatars, I understand. Just maby....
The forum is allready great, but it could be enhanced even further with the use of avatars.


Stargazer.

[chris]

Another thing I wish in Celestia, or rather in the forum, I must say: Avatars, the little picture under our names giving a little decoration.
If there isn't enough space for avatars, I understand. Just maby....
The forum is allready great, but it could be enhanced even further with the use of avatars.

I haven't tried it yet, but phpBB does support avatars . . . Just click on profile at the top of the forum page. You'll see a form that allows you to customize your preferences. Toward the bottom, there's an option to upload a custom avatar up to 80x80 pixels.

--Chris

[alexis]

As far as I can tell, there's no parallax information included in the Tycho catalog. Am I just missing something? If it's there, I'll get to work immediately on converting the catalog for use with Celestia.

In fact, there are parallaxes in Tycho, but they really don't add anything because the errors of the parallaxes exceed the parallaxes themselves... (except for the very nearby stars).

/Alexis

[Guest]

Mikeydude750 wrote :

2 words: Death Star

It would be so fun to blow up a planet....to relieve some stress.

True, true..... but what you really want for that is a Mimehunt-type screensaver !

Redfish wrote :

Galaxies: It'd like them to consist out of stars only That would rule!
And heck i don't mind. Why don't you fill the rest of the galaxy with stars? It doesn't have to be realistic. I just want it filled. Realism can be added later when new data is available.

Hmmm.... that would be nice, but how many stars would you need to not only show the spiral structure of the galaxy, but also show it from a very long way away i.e. a few thousand light years. Currrently the stars more or less disappear at those sorts of distances. Personally I'd prefer a better 3D model rather than fake stars but I suppose it's just a matter of taste.

Local zoom function: When selecting stars it's quite hard to pick them with one try. I'd like to create a rectangle with my mouse around a certain number of stars, and then see a little screen with that part of the selection zoomed in, so you can select the stars more easily.

I think that's a great idea.

No sounds please. There aren't any sounds in space. And i too like the God's eye perspective

In Celestia no-one can hear you scream, but I just think the program would "feel" nicer with a few sounds.... even if there were only a few blips or bleeps when selecting things.

Finally: Bigger textures, extreme detail on planets etc They just make me drool.

They certainly do.... of course you probably know lots of places to find bigger textures. I don't think they should appear in the standard version of Celestia, because those of us without very fast modems would be driven slowly insane waiting for it to download. Perhaps allow users to choose the resolution level of textures when downloading ? Also, I'd like to see smoother transitions when textures load up i.e. I have a 4k moon, which takes about 30 secs to load when viewing it. Perhaps all textures within the current star system could be preloaded into memory ready for viewing ? I think someone already suggested this.

(Mad Boris)

[Sum0]

How about 360 degree FOV? Or at least a panoramic screenshot option. Or even a >360 degree FOV... that would be weird.

[RohanTM]

A basic feature I'd like to see implemented is leaving the camera pointed where it is when the Track feature is toggled off. Track is a useful way of finding, say, a nearby planet when you're at one of its moons; but when Track is toggled off the camera snaps back to wherever it was originally pointing and you have to haul the camera around by hand. I am continually surprised at how long I can do this without actually finding what I am looking for.

Would be cool for rings to cast shadows onto planetary surfaces too, and for moons and other objects to fall into shadow when sunlight is blocked from another body.

I really like the idea of pre-programmed tours as well.

What bodies are currently bump-mapped? I notice that many moons (such as Mimas) don't seem to be. I'm sure that this is a huge job though.

N'cha
Rohan.

[ogg]

I'd like to see

1) a 'disable' checkbox for each graphical effect element. Some are already there in the options box, but I'm personally still stuck with strange lighting on planetary rings (due to crappy OpenGL extension support on crappy graphics card) and I'd like to be able to 'hack' the rendering on a feature-by feature (or extension by extension?) basis. This might also allow users to more effectively identify/debug their own hardware-specific problems and not clog up this forum (like I have been known to...)

2) A preservation of realism - or at least a clear option to do so. If we don't have the data we don't have the data. There are always suggestions made about cloning data to make planet maps look more 'realistic' but this is always just going to be an illusion of realism. Knowing the limits of what we know is as interesting as seeing what we do... so putting in extra stars or cloned map data *by default* is just a bit deceptive. There's a big difference between virtual and fake. On the same token, it would be nice to have all the data available on milkyway structure (Ie the radio-hydrogen maps), pulsar location and so-forth brought into the default data set. Pulsars, Nebulas, Galaxies, spiral arms, even if it's just the positions of these things optionally rendered as dots or blobs. Some stock-taking of publically available data and and prospects for it's future incorporation would be interesting in itself.

3) It might be useful to keep texture/data/script-sets under separate folders as though they were 'profiles' or 'skins'. You could always just use the basic, default data if you want. But those who want could maintain customisations separately or maintain different universes altogether in parallel data sets under different folders on their hard-drives: a scientifically 'pure' universe, a 'star trek' universe, and numerious other fictional, customised or 'airbrushed' universes. The scripts, datafiles and textures in the skin would *override* those files in the default of the same name (ie be loaded in their place), but where no replacement was offered it would revert to the default. This way the people who want fake stars can have them and the real data too. You could have one or more 'custom' folders but it would also let whole universes be created and distributed easily (as skin-like packages for download) that the user can install, explore and appreciate in their own right without worrying about them wiping anything they already have. I don't really like having to backtrack, undo or reinstall if I want to put Celestia back the way it was before I downloaded Rassilon's latest efforts etc. I know most users here are a lot less squeemish then me, but at the moment Celestia is still a bit too messy to effect changes for newbies to take the plunge. It seems a shame to have such 'wow' effect graphics and intuitive UI and yet require you to basically have to know how to hack it yourself in order to try out anyone elses modifications.

[Guest]

Just like to say I really like that skin idea - it would preserve Celestia's massive flexibility and customisation options whilst saving us from the irritation of manually altering files, finding they doon't work because you missed out a comma or renamed something slightly wrong... you know what I mean.

[Avetikus]

I'd like to see mip mapping.

I was fooling around the moon the other day with time compression at about 1000 (heh.. that phrase can only make sense here) and every time the earth swung by I would take a huge frame hit.

I'm using a bumped 8k earth and the earth itself was only 50 pixels wide or so, but the card dutifully loaded it up each time it swung through the camera's lens.

Thanks for listening,


M Rodriguez

[ogg]

Request a new Milky Way. The old one looks very nice, but doesn't seem to fit the data as well as it might.

Here's a page about the shape of the milky way, combining latest published data and findings to give an approximate sketch of it. rest of the website is very good as well.

http://www.anzwers.org/free/universe/milkyway.html

note that the structure of the spiral arms is much more complex then the current blob-model in Celestia.

Plus:
the blobs are so faint as to require a serious contrast adjustment on some monitors (fine on my laptop's LCD, invisible when pushed through a TV monitor) that's detrimental to the rest of the objects rendered. Some way of adjusting the brightness of the blobs from inside Celestia would be nice.

Naive thought from a non-programmer:
There have been comments in the past about adding dust lanes or rendering the blobs as partly opaque. Perhaps this is too complicated. Perhaps it would be enough to not render more distant blobs once inside the disk, instead of having it calculate which ones overlap. Closer you are to the central plane of the disk, the less of it is rendered - perhaps only a cylander centred on your location, which did not extend fully to the central bulge. As you get further away from the galactic plane, more and more is rendered until you get to the edge, about 1000ly G-north or G-south (we're talking about moving along the z-axis here, perpendicular to the plane of the galaxy) and the whole thing is done. I think that's how Sky3D v2 approximated it; the only other program I've seen which tried rendering the Milky way in 3D.

thanks again for a great program
ogg

[Guest]

On the subject of rendering Galaxies:

I read an article in Game Developer magazine a few months ago on volumetric fogging techniques that was very interesting. This morning, I was thinking about the galaxy rendering issue and wondered how these techniques might be applied.

With fogging, the view vector's intersection(s) with the edge of the 3D fog volume is used to determine the opacity of the fog. This produces a pretty realistic effect both from inside and outside of the fog. Interesting things happen when the view vector passes in and out of an oddly shaped fog bank.

A similar method could be used to render galaxies, using a 3D modeled galaxy volume to modulate things like color and brightness as well as opacity.

3D Volumetric fog isn't yet supported in hardware by 3D cards (at least I don't know of any) and I don't know how much extra horsepower it would require CPU-wise, but it seems like a promising technique. You might need to use multiple volumes per galaxy or use more complicated geometric properties than simple linear intersection, but I bet the results would be pretty .

Just my two cents.

abiogenesis

Here's a link to a paper on Volumetric Fog in OpenGL with screenshots:
http://www.nothings.org/computer/vfog/
Probably not exactly what we'd need as galaxies can't really be rendered as sharp-edged boxes, but it's a pretty good demo of the technique.

[Guest]

How about real clouds for earth? Worldmap, which you can get from http://www.virtware.net gets real weather satellite images for the clouds which is pretty neat.

[Sum0]

Why just a star browser? Why not a planet browser, with radius, mass, orbital data, type (would require a new tag in ssc... 'Class GasGiant' or something), moons, etc? Or a galaxy browser...

[Paul]

abiogenesis wrote:

3D Volumetric fog isn't yet supported in hardware by 3D cards (at least I don't know of any) and I don't know how much extra horsepower it would require CPU-wise, but it seems like a promising technique. You might need to use multiple volumes per galaxy or use more complicated geometric properties than simple linear intersection, but I bet the results would be pretty .

I did something similar to this in my own computer program - it used a vertex shader, so was very fast. The main issue is that you need density variation - like the GL demo you mentioned, I simply modelled volumes of uniform density (it made the fog density easier to calulate from each vertex ray), so my galaxies were all elliptical/lenticular - no spiral structure.

It should be possible, however, to have your vertex shader calculate a ray's density based on an equation that takes the number and tightness of arms and size of central bar into account - it'll just be a lot more difficult. I hope that Chris decides to tackle it...

[slashi]

Would be nice to see such feature in Celestia to run scripts (from command line):
Celestia.exe myscript.cel