An idea for a new feature

[Oshyan]

Cyber-Angel, your ideas about simulation are nice but a bit grandiose - TG is not now and probably never well be an astronomy or celestial object simulator. TG is fundamentally an artistic tool so it's usually going to be up to the artist to create a realistic effect. We might provide a system to create planetary rings for example, but it's unlikely we'll do much beyond setting reasonable defaults to really let people know "Hey, that's not realistic, the simulation says your rings should be bigger". I can't speak for Matt, Jo or John on this, but I just don't think that's really the approach Terragen takes, generally speaking.

Mind you it's always good to provide tools to enable realistic effects, and particularly to provide realistic presets and defaults. But in terms of actual simulation that's probably taking it a bit far for most things. Where it might make sense is, as you said, in a particle system, but I think the forces simulated there would be fairly basic - gravity, wind, collision. Not sophisticated enough to produce a planetary ring through actual physical simulation. In fact I think the best simulation engines in the world have a bit of trouble with those kinds of scenarious. So no sense trying to compete with them when essentially the same result can be created artistically through more simple but powerful tools by a knowledgeable artist.

I think this is the sort of approach we'll end up taking with things like depicting planetary-scale weather systems. Sure you could setup a sophisticated system to simulate global weather and make the National Weather Service jealous, but who's got the super cluster to run it on? We're much more likely to find a good, versatile approach to making realistic-looking global clouds that involves little or no simulation IMO. It's just more practical and functional in most cases.

- Oshyan

[§ardine]

Enjoy.

[attach=#1]

[attachurl=#2]

[Will]

Nice, its using populations on a disk right?

Regards,
Will

[FrankThomas]

damn . . . I tried that but couldn't get it to look right.  I'll have to have another go

[Cyber-Angel]

I have had an interest in the universe we are one vary small part of, all my life,  times are changing in the world of both hardware and software these are existing, dynamic times we live in right now and shows like Siggraph are testament to this.

I may come of as an academic stuff, this is not my intent at all; I am Academic, Philosopher and Artist both I have been around computers since the Commodore Vic-20, I remember seeing my first computer animation, Luxo Jr as it goes one of the first uses of Ray-tracing that is well known I still have the same feel now as I did then about what a computer can do and what they could do in the future, in other words the magic never went away.

Think how far we've come in twenty years when computers had 500k of ram if you where lucky, when companies like Magi Synthavision worked with early 3D software (the precursor to ray tracing started off as a method to track radiation fallout) it took a week just to render a simple image and you had to hand code every thing, and you could not rotate an object in any direction like you can now due to the limits of the computers back then.

The point I am trying to make is that I would expect a certain level of realism in any ring system that maybe developed at a certain point in the future, and that the necessary information required to-do so is in vast amounts of literature out there.

Like any project, you need to see what is out there and then read carefully what is there and find commonalities and trends within that data and write futures based on those trends.

If you want to have acceptance in the Motion Picture and Production markets as I suspect that the high end products of the TG2 range will be targeted at in the future then a certain degree of realism is and will be expected that is to say High-end Clients = High-end Features.

Over all then I understand that yes I can go overboard some times (I call it thinking outside the box) but then again the early pioneers of the industry didn't play by the rules either and look where we are today my point is that now the Tech Preview is out in the wild the competition will be looking to play the catch up and pass game that gets played, there is a vary small window for Terragen to become the Industry leading software application look to the users of this community to see that happens.

Regards to you.

Cyber-Angel       

[Will]

Cyber-Angel I see your point but I would also like to say that I would like terragen 2 to be bug free before we start asking for somthing as big as an realistic calculation of plantary rings This is a neat idea and I completely back it but there are higher prioritys for Planetside to focus on. Though if there are any programmers out there looking for a challenge this would be a very intresting project to take up, you could possibly have the output from the code to be placed some how into the distrubution shader that plugs into a population of rocks. I have seen sevral planary collision simulators around the web so I know it can be done Things such as this is why Im trying to learn C and C++ becuase the extent of my knowage right now is some limited assembly.

Regards,

Will

[Cyber-Angel]

 He-He,  you will find not stronger Advocate of TG2 becoming bug free or of Terragen in general or indeed where it could go in future then me. I will repeat what I said on Ahundar last year and that is Planetside comparatively specking (In industry terms) is only a small company with finite resources, and I understand and respect that.

I suppose I like to see where Terragen could be instead of where it is, and times and business models in the software industry have changed or at least are beginning to with the advent of the internet the closed developer model is beginning to give way to a more open community lead development model, Silo 3d from Nevercenter and the Brazil Render System are examples of this model in use today.

Perhaps one day to speed up development maybe Terragen could be made open source but still have in house development as and when required (Just me thinking aloud).

As Adam Savage (Mythbusters) once said "I reject your reality and substitute my own." any way the way I look at it is what we can not do today with time and hardware we can do tomorrow, and if we can not do so then the competition will.

There are many things I wish Terragen had but this is not the post for the discussion of those nor are they fully throat through at this juncture and as Yoda said "Do or do not, there is no try". Any way what ever happens with some future incarnation of Terragen receiving Planetary Rings (Which should include Ring tilt angle and ring spacing) I look forward to. Any way "Onwards and Upwards" as they say. End of Line.

Regards to you.

Cyber-Angel     

[Will]

Heh the adam savage quote is one of my favorites Anyway I would like to add that its 3d party plug-ins and community based projects that get some of these features added, it may start with a planetary ring script that someone makes and eventualy turnings into a valuable resoures that people look for. Its like Garry's Mod from Half-Life 2 it started out as just a little coding project and turned into a retail game after the community joined in so all we need is someone so write a "simple" script that we can all add on to in are own way. So who will take the intiative? (looks around) Will not one of you take this idea and form it into reality..ok the shpeal (shpeel?) is over

Regards,

Will

[swiftstream]

The problem with a realistic simulation of planetary rings is that it essentially an n-body problem with millions of bodies. The n-body problem does not have general closed-form solutions for n>2, and even the fastest general simulations run in time at best O(n log(n))[1]. I did a research paper about the n-body problem, once upon a time.

I suggest that attempting to simulate planetary rings should not be something that Planetside should worry about--at all. The vast majority of applications for a program like Terragen require not realism, but plausibility--it has to look "good enough." Very few movie-goers, for example, will really care if planetary rings are not physically realistic, as long as they are visually plausible. Sardine has already demonstrated that it is possible to create visually plausible simple planetary rings with the tools we currently have, without excessive difficulty.

What of more complicated planetary rings, though? For example, what if you want to do an animation of a planetary ring? (I assume that planetary rings rotate... I'm not an expert on them specifically, though). Terragen does not currently allow rotating populations as a whole--if you could do that, while preserving the specific instancing (which makes it even more complicated), then it would be an acceptable way to make animations of planetary rings on sufficiently small scales (or from a great distance), and over sufficiently short time frames. And for any yet more complicated or realistic requirements for planetary rings, I would suggest that Planetside let somebody else write a plug-in for that. This sort of third-party extension--both free and commercial plug-ins--of 3D programs has a long and productive history.

Maybe my long history with Terragen--it's been 8 or 9 years since I first stumbled across it, now--has given me rose-tinted glasses, but I honestly can't say I find any fault with the way Planetside has gone about developing it, or marketing it, in the past[2], nor the direction it appears to be headed in the future. Of course, like everybody else I would like a program that does everything I could possibly wish for the low, low price of $4.95; but Terragen, for $99, $199, $299, or whatever it might end up costing, is an exceptionally capable program, and for a company that has two main developers (Matt and somewhat more recently Jo) Planetside has done some truly exceptional work over the years I've been around.

[1] Due to the nature of a planetary ring system, I expect it could be roughly simulated considerably more efficiently with a dedicated algorithm; perhaps even O(n). However, for a ring with millions of objects in it, even that could take a good long time to simulate.
[2] Including Matt's trip across the ocean to DD and the ensuing delay, the delay in the Tech Preview's release which caused so many ill feelings, and so on. Obviously, it wasn't always the best possible PR, but stuff happens. Things go wrong. Perhaps I am patient and/or forgiving due to my long history of making mistakes, missing deadlines, and so on...

P.S. Will, it's spiel--from German, literally meaning game. It's taken on a somewhat different meaning in English, though.

[Sethren]

Perhaps Planetside should just take a look at Diard Sotware. There universe image creator plug-in makes some nifty rings.

http://www.studiotools.co.uk/images/plugins/universe.jpg

http://www.diardsoftware.com/contests/Jan05/destinationsaturn.htm

There concept for there 3D rings is not bad, all that is really missing is the ability to add alpha maps and texture maps. Texture maps and shader control can easily fake dust and ice particles of a ring system, even multiple rings like we see in saturn. For saturn all you need is a ring gradient pattern and of course your textures for dust-ice. I used to do this stuff in photoshop all of the time and it turned out fairly realistic.

[Will]

P.S. Will, it's spiel--from German, literally meaning game. It's taken on a somewhat different meaning in English, though.

Ah thank you Ive been wondering about that for a couple years now . And yea I understand that the entire N body problum is not something for planetside to be working on (I've been using terragen since I was 11 so 5 and a half years now) so thats why I think it should be a community inishtive (as I preluded to earlyer Im 16 so thus can't spell) to make such ideas come to life.

Regards,

Will

[Oshyan]

I suggest that attempting to simulate planetary rings should not be something that Planetside should worry about--at all. The vast majority of applications for a program like Terragen require not realism, but plausibility--it has to look "good enough." Very few movie-goers, for example, will really care if planetary rings are not physically realistic, as long as they are visually plausible. Sardine has already demonstrated that it is possible to create visually plausible simple planetary rings with the tools we currently have, without excessive difficulty.

Precisely! Although simulation is creeping into the mainstream of effects visualization and even hobbyist computer graphics (e.g. Blender's fluid simulation), the majority of 3D work is still done with approximations, "fakes", etc. It is made to be "good enough" - all it needs to do is fool the average person. 99% of people won't know an accurate simulation from an inaccurate but visually convincing fake. Producing the latter is usually much easier for a program to facilitate. More demanding on the artist perhaps, but then again often times pure simulation produces undesirable results artistically speaking, so you need explicit control within the simulation if you're going to implement it, which makes it again more complicated and difficult.

Ultimately planetary rings and other complex, specialized phenomena are probably going to be the realm of plugins. I am reminded of the Halosim program, entirely dedicated to simulating atmospheric halos. It is implemented in such a way that its output can be combined in post processing with that of other applications, and the effects it simulates are highly specialized and generally quite rare (most people see them less than 5 times a year). Yet an entire, complex program was created just to simulate this. The same would likely be required to create accurate planetary ring simulation.

I would be thrilled to see the Halosim program integrated with TG2 as a plugin, so that its effects could be incorporated easily in any TG2 render without compositing or other post work. The same is true of any planetary simulation; this seems a great project for an interested astronomy student and coder to tackle. TG2 is primarily a visualization platform, not a simulation engine, but that makes it the ideal partner for deep simulation coders to work with in creating graphical representations of their simulation output. I have high hopes for the plugin community once the SDK becomes available. Don't let me down guys.

- Oshyan

[Will]

Cyber-Angel ever do coding in your long history with computers? You seem very intrested in planetary rings so it would seem you have the enthusam. If I knew really anything useful about programing I would start it off for you guys but im afiad im still on the first chaper of "programing C++ for dumbies"

Regards,

Will

[Cyber-Angel]

I never really found the time to learn to code, my attitude towards software is "If they can think of it as a plug-in, then include that functionality in the main programme" what I understand is this, Yes Terragen is about artistic expression and conversely as a production tool (TG2 Deep and Deep + Animation any way) to this end then software should provide the artists with the tools they need to do there job, nothing more complicated or scary then that.

To that end then, when I read about how in some software used in film fx production that it takes six months just to set up the light rig, I think to my self the old adage "Work smarter, not harder" in other words there has to be a better way to do the same job in less time, in other words help the artists do what they need to do in the most productive way possible.

In the end then, the artists who use a software product weather they be a single individual producing art at home to the people do work in the motion picture industry are the life blood of any software company no matter what scale of operation they are, if artists can work in a proficient manor they are more then likely to be less stressed at the end of the working day but more over they are more then likely to maintain brand locality, its not just common sense its sound fiscal practice.

Yes, for now approximations are common practice; fine I have no problem with that, then again computer hardware is becoming more powerful all the time so with that in mind better ways need to be found to do the same thing.

Plausibility, my definition of that is to do with human perception of what is real and that means macro scale detail that is present in real life but most of it is filtered out by the brain to prevent information overload, however if this detail is missing the brain has a problem with the image as having some thing missing and thus been perceived as been unrealistic and fake, this has been the bain of CGI since the beginning.

You know I wish I could program, but what is, is, and so I will have to live with that some people are meant to do things-that others are not that's life!

As long as the Planetary rings are more sophisticated then those seen in some space sim games (Starlancer and Freelancer, come to mind) which look like a ring texture mapped on to a flattened disk with an alpha channel for transparency I will be happy. 

Regards to you   

Cyber-Angel

[Sethren]

True, we absolutely need something beyond texture maps and alpha transparency's.