If someone was going to intergrate a sound API into Celestia

The place to discuss creating, porting and modifying Celestia's source code.
Avatar
t00fri
Developer
Posts: 8772
Joined: 29.03.2002
Age: 22
With us: 22 years 7 months
Location: Hamburg, Germany

Post #21by t00fri » 11.12.2005, 19:00

Malenfant wrote:
t00fri wrote:Anybody thought about the resources this takes away from Celestia's MAIN task: rendering the universe in 3d and exceptional precision...

I think that's a spurious argument if you consider that in all probability people do other things while using Celestia. e.g. I usually have a web browser in the background and an mp3 player playing some music, maybe an email client too or Word or Excel - all of these are supposedly 'taking away resources' from Celestia, but it still manages just fine.

Right of course, but this fact can hardly be promoted into a design criterion ;-) . The strategy will always remain to optimize the Celestia perfomance by itself (i.e. irrespective of people looking at what those Russian bots in the forum have to offer as "background entertainment" ;-) )

Also many applications like e.g. email load the system very little , since writing an email takes a (comparatively) long time and sending it is done quickly. Playing MP3 music is a continuous load, though...To have some narrator tell stories is presumably also a fairly economical affair in comparison to playing high bandwidth music.
The capability to play sound in Celestia would be great and very useful, especially for educational narrations - and to be honest I think education is Celestia's main task.
I agree, of course that sound would be fine. It's just a matter of compromise and the price may be quite high.

I can still see a number of other attractive purposes of Celestia besides education. Since I am kind of "educated" already, I would be bored with it otherwise ;-)
When it comes to "rendering the universe in 3D and exceptional precision" Celestia is really far from this goal at the moment - it doesn't even render planets with the realistic photometric functions and its multiple lighting algorithm is currently extremely flawed and so far these do not seem to be on the table for fixing, and nobody has even bothered to comment on these issues when I raised them.
While I think you have a point here, I would not go as far as saying that the overall precision of Celestia is far from outstanding. There are many aspects that I can judge very well and that are extremely impressive, indeed. The rest is hopefully to come soon, given that Chris is kind of "reactivated" ;-)
Your galaxies and binary stars are excellent and very useful add-ons and certainly allow for a more fully rendered universe, but there are other important areas that are still badly flawed when it comes to realistically rendering the universe.


Agreed. But let's see how we'll be doing in that respect in the near-future...Like I am really looking forward to a complete overhaul of the planetary atmosphere code, that --I know--is high up on Chris' todo list!


Bye Fridger

Malenfant
Posts: 1412
Joined: 24.08.2005
With us: 19 years 2 months

Post #22by Malenfant » 11.12.2005, 19:23

While I think you have a point here, I would not go as far as saying that the overall precision of Celestia is far from outstanding. There are many aspects that I can judge very well and that are extremely impressive, indeed. The rest is hopefully to come soon, given that Chris is kind of "reactivated" ;)


Yeah, I think what you're doing (importing catalogues into Celestia) and what I'm more interested in seeing (more realistic photometry and lighting) are two different sides of the realism coin. The problem is that importing the various catalogues (galaxies, nebulae, binaries, etc) is pretty much a never-ending task. For example, we don't - and actually can't import all the bodies in the asteroid belt because currently that would grind Celestia to a halt. We don't have all the multiple star systems either - we have a lot of binaries with your binary file now, but that's still far short of all the multiple star (3 or more) systems in the sky. So catalogue importing is pretty much an ongoing task that requires a lot of updating - it's certainly worth doing though, it's just a huge task ;).

However, the sort of improvements I'm more interested in are much shorter term and actually are probably easier in the long run to attain. I know they can be done too - look at the work that Bjorn Jonnson has done using his own renderer (e.g. http://www.mmedia.is/~bjj/renderings.html and http://www.mmedia.is/~bjj/3dtest/ ). As it stands, rendering of planetary bodies in Celestia is very far from realistic - all the detail in the world doesn't matter if the object isn't even reflecting light the way it should be. The multiple lighting issue is a serious problem too, as I very clearly demonstrated in this thread - and while it's certainly better than it was in 1.3 (when we didn't even HAVE multiple lighting) it's a very far cry from realistic.

If concentrated work is done just on getting the photometry and multiple lighting corrected, Celestia would be catapulted far beyond the competition when it comes to realistically rendering the universe. And I don't believe this would be that hard to do if we put our minds to it. The atmosphere code is part of this, but there are more bodies without atmospheres out there that need to be rendered realistically too. I think atmospheres should be done AFTER we've got the 'airless' bodies rendered realistically, since that would be laid over the surface photometry.
My Celestia page: Spica system, planetary magnitudes script, updated demo.cel, Quad system

Avatar
t00fri
Developer
Posts: 8772
Joined: 29.03.2002
Age: 22
With us: 22 years 7 months
Location: Hamburg, Germany

Post #23by t00fri » 11.12.2005, 19:52

Malenfant,

all fine, but what you are interested in is /exclusively/ Chris' domain in coding. This is sophisticated OpenGl stuff, where he is the Guru.

If you can rhetorically manage ;-) to reshuffle somehow his priority list, these things will be addressed faster.

Clearly, my field of activity is far larger than importing catalogs into Celestia. Yet anything related to 3d graphics is Chris' business. 3d coding is what he gets paid for, professionally...

Bye Fridger

Malenfant
Posts: 1412
Joined: 24.08.2005
With us: 19 years 2 months

Post #24by Malenfant » 11.12.2005, 19:57

t00fri wrote:If you can rhetorically manage ;-) to reshuffle somehow his priority list, these things will be addressed faster.

Clearly, my field of activity is far larger than importing catalogs into Celestia. Yet anything related to 3d graphics is Chris' business. 3d coding is what he gets paid for, professionally...


Well I've been trying... but so far nobody's bothered to comment on the multiple lighting thread - on the list Chris said he would but there's no sign of a response from him there yet. I do think these should be a bigger priority. Yes, it may involve more indepth fiddling with the code by Chris, but I think it'd be a much more meaningful improvement if the core of the program could be improved.

That's not to say that the rest of us (for all I know that may just be myself) who know something about photometry shouldn't pipe up and say something about it though ;).
My Celestia page: Spica system, planetary magnitudes script, updated demo.cel, Quad system

Avatar
t00fri
Developer
Posts: 8772
Joined: 29.03.2002
Age: 22
With us: 22 years 7 months
Location: Hamburg, Germany

Post #25by t00fri » 11.12.2005, 20:17

Malenfant wrote:...
Well I've been trying... but so far nobody's bothered to comment on the multiple lighting thread - on the list Chris said he would but there's no sign of a response from him there yet. I do think these should be a bigger priority.
...


I am of course aware of Chris' reply in the list...

For me --as a less ficticiously minded Celestian-- multiple lighting issues have more of an academic interest, since "at home" we only deal with one sun. But as you know I have just developed myself some theoretical framework about comets travelling in multi-sun environments ;-)

Bye Fridger

Avatar
Chuft-Captain
Posts: 1779
Joined: 18.12.2005
With us: 18 years 10 months

Post #26by Chuft-Captain » 05.01.2006, 04:52

t00fri wrote:Selden,

did you perhaps misunderstand? I was taking about CPU power nothing else. We are spending a hell of a time to keep the fps rare up through clever culling techniques. If I just switch on some music player in the background in addition, Celestia stops being fun, even with my fast graphics card and 3GB of RAM.


Fridger,

With regards to your comments above in this post: http://celestiaproject.net/forum/viewtopic.php?t=7628

I was wondering how the workload was shared amongst processors in Celestia. I'm assuming that any rendering in Celestia would be handled by the processor on the dedicated graphics card, and that audio playback would be handled by a separate audio card, and that the only extra load on the main CPU would be a few clock cycles for the triggering of the audio sample, which at worst would slow down a few orbit calculations etc... I'm surprised that audio could have that much impact (especially on your machine).
My machine does grind to a halt if I run Celestia and Anim8tor at the same time, but that's probably because they're sharing what little graphics capabilities I have (I haven't got a decent graphics card).
Of course, I've never seen the Celestia code, so please excuse and correct me if these assumptions are wrong out of ignorance.

Regards
"Is a planetary surface the right place for an expanding technological civilization?"
-- Gerard K. O'Neill (1969)

CATALOG SYNTAX HIGHLIGHTING TOOLS LAGRANGE POINTS


Return to “Development”