Is Orbitersim good enough? For some Celestia users, definitely not!

[diegorodriguez]

When I did put forward the subject on gravitational models at the Celestia Forum (you probably know that the otherwise beautifically crafted Celestia does not know what gravitation is) I got as reply all sorts of objections. The most interesting from t00fri, an old time Celestia bodyguard.

"...As I wrote, is is fairly trivial to implement SOME time-dependent gravitational interactions, but they are bound to be badly WRONG in many configurations. This surely holds for Orbiter, too.

One undesirable consequence is that users will not be able to judge for themselves WHEN things are getting wrong. For us it is absolutely no reason to implement something that is largely incorrect, simply because space games and other software claim to have such features. We prefer NOT to implement some physical aspect altogether, before we do it INCORRECTLY.

What we emphasized is that a CORRECT implementation would strongly interfere with Celestia's main design strategy, also because of the resulting complexity.

I am ready to discuss these issues in PHYSICS terms either with you (if you are competent in this field) , or with anybody else who feels competent, e.g. from Orbiter

The design goals of Orbiter are quite different from those of Celestia, although both simulations may superficially seem to address similar tasks..."

So Orbitersim core could be 'largely incorrect', according to this Senior Scientist in theoretical Particle Physics and Cosmology...

... and if not, who could prove it?

 

[ar81]

Can you ask him what exactly is "incorrect"?

If it is about "unrealistic DG" (overpowered and fuel being stored in a subspace container) or some vessels (old addons or spacecraft.dll vessels mostly) not having the proper aerodynamics, or atmospheric flight not being modelled to the last detail, that's a known issue. DG is not intended to be real (more aimed at being a user friendly craft for newcomers to the world of physics), but the physics of the core are real AFAIK, except for lack of relativistic physics. If it is about addons, blame addon makers.

Or perhaps he might refer to propagation of trajectories using orbital elements and no perturbations when you change date. But you won't do that very often.

In one point he is right. Both softwares have 3D, but their goals are different.
Orbiter is about "fly your craft", Celestia is a viewer. Orbiter has solar system only, while Celestia has the whole universe.

I came to Orbiter thanks to Celestia community as I suggested the idea of turning it into a space flight sim.
So it is thanks to them that I came here.

What I am certain is that Orbiter is by far the most realistic space sim I have ever seen in therms of physics.
If it is not I am all ears.
Celestia is the most complete universe simulator I have ever seen.
I love both.
And if I had to add some other nice softwares to the collection, I would add Orbitron and Stellarium.

I do not see any reason to have confrontation between communities. So unless it is just for the sake of friendly entertainment for all parties involved, I do not see a reason for a heated debate. I have both installed in my computer.

 

[diegorodriguez]

Yes, yes, I know all that about solar system focus versus universe focus, and it doesn't make any difference for our topic purpose.

The point is: in an specific time and space frame, is the Orbitersim gravitational system faulty or unrealistic?

 

[ar81]

Probably Martin could correct me if I am wrong...
As I understand it, vessels in Orbiter are subject to orbit perturbations, but planets and moons are not. So you may get some small variations in time between reality and modelled reality that would go unnoticed unless you plan a real trip in real life.

Also, modelling twin planets like Pluto and Charon is not possible in Orbiter, AFAIK.

As for Celestia, I have found some bogus behavior in a previous version if you change time to year 10,000 AC. Moon was not looking towards Earth anymore... Haven-t tested lately. It seemed to be caused by the ephemeris data that is used to feed Celestia, if I am correct. Ephemeris data makes movements valid for a certain period of time.

This is my understanding. But I would suggest to confirm with a more reliable source.

 

[simonpro]

Why don't you give us a link to the topic on celestia? That way we can see exactly what he's on about.

I wouldn't care too much if he is a senior scientist in the field of cosmology or whatever either, doesn't mean he's right. IMO the people who are usually wrong are those who spend more time advertising what they are rather than coming up with the goods

 

[ar81]

That situation of cosmology reminds me a weird man I met long ago.
If you tell him he is right, he will tell you he is right and you are wrong.
He has gathered an impressive collection of papers on space missions (non technical newspapers, magazines, books mostly) and he claimed to know a lot about space.
I found orbiter and once I refuted him, and he called me "astronaut wannabe who thinks that a sim makes you technical".
In the first moments I found it offensive, but later I discovered that he did it to other people in no space related fields. Now I see that man was in need of serious psychiatric services. He advertises himself as a local authority on UFOs (because he has papers too)...

 

[cjp]

I would like to see the original Celestia thread, but in the mean time:

When I guess what this is about, I'd say it is about the accuracy of the orbits in situations that are more complicated than a simple 2-body system. Maybe this guy doesn't trust the way how Orbiter handles certain perturbations.

It would be a bit unfair to judge Orbiter on that. Maybe it's true for a particular version of Orbiter, but things can always be improved. I don't see why the movements of celestial bodies can't be as accurate as in Celestia. And about spacecrafts: I think it's better to compare that to flight simulators than to astronomical software. I doubt any flight simulator can be as accurate as the (non-atmospherical) motion in Orbiter.

Orbiter has a disclaimer that it's not supposed to be used for actual flight planning & so on. I think it should be judged in this way. If you see something weird in Orbiter (or even whenever you think you can draw conclusions from what you see in Orbiter), there's always discussion possible on how accurate Orbiter actually is.

It looks like the discussion was about whether such simulation should be added to Celestia. The arguments should be judged in that context too. So I guess it's more about Celestia (and the Celestia community, or maybe just this guy) than about Orbiter.

 

[2552]

Also, modelling twin planets like Pluto and Charon is not possible in Orbiter, AFAIK.

FTR, there are 3 different Pluto/Charon addons available (2 on Orbithangar, 1 on AVSIM).

 

[diegorodriguez]

Well, you can find the Celestia thread at

http://www.shatters.net/forum/viewtopic.php?f=2&t=13141

 

[ar81]

FTR, there are 3 different Pluto/Charon addons available (2 on Orbithangar, 1 on AVSIM).

Yes, but it is modeled as planet and moon, not like a twin planet system.

 

[simonpro]

Diego, you're very wrong I'm afraid.

Highly successful Orbitersim is entirely based on a gravitational model.

No it's not. Orbiter (as celestia) uses VSOP for planetary calculations, hence how your starting MJD can be any date, past present or future.


-----Posted Added-----

Yes, but it is modeled as planet and moon, not like a twin planet system.

Nonsense. The Earth/Moon system is modelled as a twin planet (go look at the barycenter scenario if it's still there)

 

[ar81]

Well, you can find the Celestia thread at

http://www.shatters.net/forum/viewtopic.php?f=2&t=13141

I have an idea for Celestia developers.
Since motion of bodies is given by a set of data that is generated somehow, and since they do not want the physics engine inside Celestia, how about a separate physics tool that allows to generate trajectories for objects using gravity computations for vessels, comets, probes or whatever?

 

[to be]

I have an idea for Celestia developers.
Since motion of bodies is given by a set of data that is generated somehow, and since they do not want the physics engine inside Celestia, how about a separate physics tool that allows to generate trajectories for objects using gravity computations for vessels, comets, probes or whatever?

Because that would be exactly orbiter?

 

[Urwumpe]

I hate cross-posting between forums. if diego wants to create mischief between the two communities, he should leave both. t00fi did not attack orbiter at all, he just delivered a clear and scientifically correct answer why Orbiter is not 100% accurate. But let's start somewhere else:

Double planets are possible in Orbiter since the 2006 version. Maybe even earlier, but the 2006 needs less fiddling. Sure, you need a DLL with an ephemeris module to implement them, as you can't describe their orbits using Kepler elements. Of course Orbiter does not simulate the planets. Why simulate something in a bad way, which ephemeris modules with precalculated data can do more accurate?

Orbiter does gravity calculations on vessels very accurate - there could be more accuracy, sure. We don't have all J-terms of nonspherical gravity supported yet, and we use only Newtonian physics.

Orbiter solves the time problems in the planet positions by using a ephemeris module for the planets and only allows to propagate vessel state forward in time.

Still better as Celestia - Celestia can't afford such cheats as Celestia has a different mission. Celestia is a astronomical visualization tool, and needs to jump forward and backward in time. Orbiter is a spaceflight simulation toy and only needs to propagate state forward in time.

Sure, Orbiter is the best and greatest and more funny as Celestia. But we can't compare them from a physical point of view.

PS@Diego: Learn what Orbiter is, before you discuss it in other communities. Your posts in the Celestia forum are only one step away from being insulting for BOTH Orbiter and Celestia Fans.

 

[ar81]

Well Orbiter is more fun for me. But for older people or very very young kids Celestia is.

The problem, Diego, is that I see you are very energetic, but following that path you are likely to say things that could start a flame war. If you are used to the ways of speaking and debating in the spanish way (I know what it is to engage in a heated spanish debate), and you translate them literally to english, without considering cultural differences, you could end up being a bit insulting for some people.

Orbiter and Celestia developers have dedicated a significant amount of their life for free. And the least they need is to feel diminished.

My advise: Encourage and empower developers instead of comparing them.

Because that would be exactly orbiter?

:study::weird::shout::blackeye::blink::tumbleweed::lol::speakcool:

 

[Urwumpe]

I see no problem in comparing Orbiter and Celestia, but the comparison should remember the nature of the things compared.

For example: A Lamborghini Gallardo and a Piaggio Ape are both Italian cars. But try to use the Gallardo for delivering goods to a small restaurant in a small side street. Or drive 280 km/h with the Ape.

 

[diegorodriguez]

Urwumpe, I don't know you, even when you happen to always come in the middle of my threads (do you live inside the forum?), but you happen to play same role as t00fri plays for Celestia. A Watchdog barking at people that may seem to be critizicing your personal possession.

I am really growing tired of people like you that seem so usual on these communities, defenders of the essence of the holly code, guardians of the official standpoint.

Don't say I do flamebait: I don't.

I wrote my 1st gravitational simulator early in the 80's with an Amstrad 128, and since then, I ve been an enthusiast and advocate for Celestia, Orbiter and a long list of remarkable space related software projects, so I would really appreciate you to refrain from commenting my threads anymore unless you come up with something focused on the topic, please please.

For instance, pros and cons of the current Orbiter model and how feasible would it be for a Celestia enthusiast to adapt it.

Verstehen Sie?

 

[martins]

Well, this thread (in combination with its sister thread on the Celestia board) seems to highlight the different design philosophies of Orbiter and Celestia quite nicely, once you cut through the polemics .

Celestia's main idea (as Chris pointed out) is instant access to any point in spacetime. As such, the simulation must be entirely deterministic (i.e. closed-form solutions, or at least parametric solutions as a function of time). This means no numerical time propagation of dynamics, and (presumably) no user interaction that changes the behaviour of any of the simulated objects that would break determinism.

Orbiter's main point is real-time space flight, i.e. interaction-based, and therefore fundamentally non-deterministic. Analytic and semi-analytic solutions are used where possible (e.g. planet ephemeris), but it does use numeric calculations for dynamic time propagation for any objects whose spacetime trajectory is under user control (i.e. spacecraft). The price is the loss of random access in spacetime, but that isn't a main concern for a real-time simulator anyway. And it has the advantage that we can simulate non-trivial trajectories, like the perturbed trajectory of starship in the Jovian system that was used as an example in the other thread.

There you have it - the universe is big enough for Celestia and Orbiter!

 

[diegorodriguez]

haha, the universe is big enough... So true.

I already told the guys at Celestia to gain the random time access design request keeping deterministic trajectories, and build on top of them a gravitational model for selected scenarios.

Celestia deals with imaginary worlds, as well as poorly known binary stars systems that would benefit from this approach. There have been proposals in the past, and they will keep coming until the goal is achieved.

By the way, a bit off-topic, but I published the Spanish version of an interview with Chris Laurel at my blog http://aerialphenomena.blogspot.com/2008/10/detrs-de-celestia.html, where he discussed the main goals of the project he heads.

Martins was also invited to submit his answers, but so far I have not received any feedback. Maybe on his spam box. It would be so nice to gather these two outstanding personalities, at least in the cyberspace!

 

[to be]

Urwumpe, I don't know you, even when you happen to always come in the middle of my threads (do you live inside the forum?), but you happen to play same role as t00fri plays for Celestia. A Watchdog barking at people that may seem to be critizicing your personal possession.

I am really growing tired of people like you that seem so usual on these communities, defenders of the essence of the holly code, guardians of the official standpoint.

This is way out of line. You do not own 'your' threads, this is the point of a forum. If you don't want discussion of a topic, don't post it. Orbiter and Celestia are not the same thing. Orbiter simulates spaceflight, Celestia simulates Space. They are not the same thing. Neither are 100% accurate, but trying to say which is 'better' depends what you want to do.