Macross mod for Source Engine

[Aurel Tristen]

First off let me clear up that this is just a concept that has been rolling around in my head since I saw the first 3 seconds of video of Half-life 2 and the Source engine at E-3. I hope I didn't cause any heart attacks with that title (observe subtitle)

I wanted to lay out what I had in mind in front of some hard core Macross fans, and see what you all think. I know the talent is out there to make this happen, but its a mater of finding it. I'm not planning on actually trying to make this mod unless I find no less than 2 skilled programmers, but again, I'd like to get the idea out there.

I understand the Macross universe is MASSIVE, and there are hundreds of Valkyries and other crafts that fans have always wanted to pilot, but I think starting small is key when it comes to MODs. Community first, then additional content. It doesn't work the other way around.

My goals would be:

• Full cockpit piloting:

  The player never has a magical external camera. This is a simulator designed to immerse the player into the roll of a VF-1 pilot. Fully modeled 3D cockpit with consistent animations for various controls. Animated to experience transformations. Player is able to see the pilots body operating the controls.

• IR head tracker support:

  After trying this little baby myself, I now know NO flightsim is the same without it. Head tracking would be used to allow players to observe the cockpit, and track enemies.

• Space combat:

  I've seen it done on source before, so I'm sure it can be done again. Space combat would probably be easier to implement than atmospheric flight.

• 1st person shooter type play in Battriod, Flight Simulator type play in fighter, and a unique combo in G.
• Multi-player with up to 64 players (higher numbers would require more work on the netcode)
• Game modes with enemy forces controlled by AI (such as in Left 4 Dead, using AI to balance the battle) and PvP
• VF-1 with all major variants and fast packs.
• Take off and landing (full length and catapult)
• Full joystick support for G and F modes
• Chain of command:

  Players join a squadron under the command of a squad leader (like in BF2). Squadron leader designates primary goal (IE cover Blue Squadron, Bomb Primary target, Defend Sector XX)

• Visible interactive user interface on main screen of fighter cockpit
• Inverse kinematics for B and G modes
• True bullet trajectories (IE lead target, use targeting computer)
• Missile targeting and shoot down capability

I chose the source engine because its easier to make a FPS into a flight sim than a flight sim into a FPS. the Source engine can scale to slower computers, is highly moddable, and easily accessible through Steam. It has a massive existing fan base, and a lot of potential modders.

The goal for the first release would be a flyable VF-1S on a large desert map (easy to map, easy on the CPU) facing off against a few target drones. Player could take off, engage targets in 3 modes, and land. Player would have use of the gunpod, micro missile packs and basic long range missiles. A small city type range would be set up in the middle of the map for MOUT type engagements, and to test B mode.

The second major release would be the same map, 1 on 1 VF combat.

The third major release would introduce Space combat and AI Zentradi (spelling)

and so on.

[Cdr Fokker]

I would be up for it, but I have long since abandoned modding in favor of working from scratch. I'm also not particularly fond of Steam, or any coding Valve-related after my last run-in working with a Half-Life (1) mod.

Also, if you're truly interested in a simulation rather than a game, choosing a typical game engine is the wrong move. A sim needs a consistent real-time loop running at a known frequency for calculations, which most (if not all) game engines tend not to support. For example, most games don't support running an internal calculation RTI loop at 300 Hz (especially while the graphics loop is simultaneously running at only 60 Hz).

I have a full Macross flight sim planned after I finish my current side project. Actually, if anyone is interested in helping with this effort down the road, I will need some "media artists" (2D/3D, audio, etc). Not to pull anyone away from this, of course. As I said, my project will be quite a ways down the road time-wise.

[Aurel Tristen]

I would be up for it, but I have long since abandoned modding in favor of working from scratch. I'm also not particularly fond of Steam, or any coding Valve-related after my last run-in working with a Half-Life (1) mod.

Also, if you're truly interested in a simulation rather than a game, choosing a typical game engine is the wrong move. A sim needs a consistent real-time loop running at a known frequency for calculations, which most (if not all) game engines tend not to support. For example, most games don't support running an internal calculation RTI loop at 300 Hz (especially while the graphics loop is simultaneously running at only 60 Hz).

I have a full Macross flight sim planned after I finish my current side project. Actually, if anyone is interested in helping with this effort down the road, I will need some "media artists" (2D/3D, audio, etc). Not to pull anyone away from this, of course. As I said, my project will be quite a ways down the road time-wise.

Well this is what I posted here for, this is some good feedback.

Valve has actually gotten a lot better in recent, and I wouldn't call Source typical. Its quite flexible, and I've worked with it since its creation, and with Source Gold before that since its creation, and the early Quake engines before from witch it was built, but I have to say you defenetly know a lot more about this stuff then me. I'm not sure what you mean when your talking about real time loops and such. Is this programing related, graphical, AI, or engine hard code stuff?

I've loved the idea of a independent Macross game, but its hard enough to get a mod made, so by the time most engines are developed to a point where they are useful, the team is disbanded, everyone loses interest, or SOMETHING happens. Hell, I've been on the Valkyrie simulation project team for years, and am still waiting for the engine to catch up to a point where I'll be able to help out. I believe in the last 3 years, they've pinned down the flight model for a F-14, and are still tweaking the VF-1. Its still a HUD hovering over a green checkerboard. I guess I just don't have the patience for this kind of thing. I've never seen a independent engine make it past the early testing phases (except Infinity: Quest for Earth, witch is AMAZING by the way)

Hell, I just want what everyone else wants, a VF-1 flight sim with a little immersion and a damn 3D cockpit.

[Pawn_of_Evil]

I have been working on a Source mod myself over the last few weeks, although not Macross related. If you don't have experience with Source, it can be a pain to get custom content into the engine. You might have better luck with Unreal 3, which has vehicle support built in and it is a whole lot easier to get custom content in. You can also make it look better if you have the computer for it.

[Cdr Fokker]

Its quite flexible, and I've worked with it since its creation, and with Source Gold before that since its creation, and the early Quake engines before from witch it was built, but I have to say you defenetly know a lot more about this stuff then me.

I'm a Software Engineer who specializes in real-time simulation (particularly graphics).

I'm not sure what you mean when your talking about real time loops and such. Is this programing related, graphical, AI, or engine hard code stuff?

All of the above.

A typical game will run flat-out : as soon as it completes one pass of processing, it starts on the next. Depending on the current load - due to AI, graphics, what have you - the time that is required to perform this processing may fluctuate. This does not lend itself to doing precise calculations, as the timesteps are variable and not always easily calculated on the fly.

A real-time simulation, on the other hand, runs fixed-frequency loops (or at least attempts to... sometimes the load becomes too great and you can experienced dropped data, frames, etc). There are often several different loops running simultaneously at different rates. For example, mechanics (kinematics, aerodynamics, etc) code typically runs fairly fast. 300 Hz isn't unheard of. High precision is needed to ensure proper simulation, even though all of these timesteps are not being rendered to the screen (though it would certainly be nice to be able to!). Other simulation-related code - for example, network communication or low-fidelity physical simulation (simplified rigid-body kinematics, for example) might be running at a slightly slower rate - say 120 Hz. The slowest loop (or thread) is the graphics thread, which is usually run at 60 Hz. Depending on the implementation, though, this thread may actually be decoupled from the real-time loops and run at a variable rate, like games do. However, most stricter simulations enforce a fixed framerate to ensure consistency.

Most game engines are not set up to deal with this kind of loop/process/thread maintenance. Since this management is the central core of the overall engine for a sim, it's not like you can just shove it in a standing game engine - especially one that is closed-source. Games are often not optimized for such CPU-intensive processing either, with graphics being a big part of current games, which have the vast majority of their processing accomplished on super-fast (due to optimizations) GPUs.

It really all comes down to a design mentality. A sim operates differently than a game. My current side project ("side" only to denote separate from my paying day job) involves a sort of hybrid, with framework in place to support simulation design, but with out-of-the-box implementation more toward the line of games.

Distinguishing between a game and a sim is like distinguishing between functional and object-oriented design. You can't just say that one is the other with more stuff. C++ isn't just C with a bunch of extra things added on - it opens a whole new series of design patterns, a whole different mentality.

[Aurel Tristen]

Ok, I understand what you are saying now. I guess I never knew the technical side of what you have just explained, but now that you have, I think I can 'feel' what you are talking about in games like MSX, as opposed to a FPS.

I don't think this is as much of an issue if you are making a simulator as long as your focus isn't on trying to replicate the true to life arodynamic properties of the VF-1 airframe. I'm not talking about making a true to life hard core emulator. I'd think most fans would be happy with something that just lest them sit in the cockpit, and fly like they see in the show, witch is often times in no way realistic (for the love of god, the fighters are flying like aircraft in space.)

I was imagining this little mod being programed much like the early flight sims, and some of the slightly more arcade type games, where aerodynamics are simulated with basic principles, such as when the plane is level, and going fast, it maintains altitude, where as turning sideways will cause 'slipping' and pulling straight up without the proper airspeed will cause the aircraft to stall. I don't really think anything from an anime belongs in the hard core simulation environment because they are normally quite unrealistic as it is (though still a million times better than something like Gundum)

It again comes down to weather you want to build a flight sim into a FPS, or a FPS into a flight sim. Its hard to say, because I love both aspects so much. I think the future may be in procedural graphics (like Infinity) to allow a fighter to switch to B configuration and land at street level, and not be looking at a road texture 64 pixels wide, and no traffic. Otherwise, environments have to be hand crafted so that they are interesting and look good. Macross is about as challenging as it gets for a game, combining these two concepts, and I think its going to take a lot of revolutionary thinking to get a good game that is true to both aspects.

So are you saying that because of this tick rate issue (pardon my non-industry words here), are you saying that a highly accurate and realistic simulator is not possible in an engine like Source, or that you don't think a satisfying flight experience is possible in it?

[Cdr Fokker]

It again comes down to weather you want to build a flight sim into a FPS, or a FPS into a flight sim. Its hard to say, because I love both aspects so much. I think the future may be in procedural graphics (like Infinity) to allow a fighter to switch to B configuration and land at street level, and not be looking at a road texture 64 pixels wide, and no traffic. Otherwise, environments have to be hand crafted so that they are interesting and look good. Macross is about as challenging as it gets for a game, combining these two concepts, and I think its going to take a lot of revolutionary thinking to get a good game that is true to both aspects.

Try not to limit yourself so much. Don't think in direct terms of existing conventions.

Oh, and paging is nothing new. Scaling from continent-scale to a road-scale in real-time is entirely possible, and has been accomplished. It's all about memory management.

So are you saying that because of this tick rate issue (pardon my non-industry words here), are you saying that a highly accurate and realistic simulator is not possible in an engine like Source, or that you don't think a satisfying flight experience is possible in it?

You cannot have a simulation with a typical "game" engine, end of story. One of my biggest pet peeves is people throwing around the world "simulation" when they haven't got a clue as to the implications of the term.

Oh, and I myself don't even get a satisfying experience out of most "flight sim" games. That tends to happen when you are involved with real aircraft and real simulation.

YMMV.

[Aurel Tristen]

Try not to limit yourself so much. Don't think in direct terms of existing conventions.

Oh, and paging is nothing new. Scaling from continent-scale to a road-scale in real-time is entirely possible, and has been accomplished. It's all about memory management.

You cannot have a simulation with a typical "game" engine, end of story. One of my biggest pet peeves is people throwing around the world "simulation" when they haven't got a clue as to the implications of the term.

Oh, and I myself don't even get a satisfying experience out of most "flight sim" games. That tends to happen when you are involved with real aircraft and real simulation.

YMMV.

I wouldn't go so far as to say that I don't have a clue to the implications of the term. I agree with the dictionary in that Simulation means "the representation of the behavior or characteristics of one system through the use of another system, esp. a computer program designed for the purpose." In essence, all computer games are simulations of something, but just at differing degrees of accuracy (weather that be intentional, accidental, or due to limitation)

In the gaming community, the word simulation is commonly used when talking about a game built specifically to be enjoyed due to its accuracy to what it is simulating. IE, flight simulator is fun because it attempts to simulate flight, where as Half-life 2 simulates the process of loading and firing a weapon, be it not accurate, but this simulation is not the driving entertainment value of the game.

I know where you are coming from, and I feel the same way when I hear games called 'modern tactical shooters', or military sims such as ARMA. Being a soldier myself, and having been on the battlefield many of these games claim to simulate, sometimes I'd go so far as to say I am offended by them, but I understand that someone who hasn't been to war, doesn't understand, just as I would expect you to understand that someone who hasn't flown an airplane will not understand how poorly simulators simulate flight.

[akt_m]

Source is kinda old. I guess a mod for UT3 wouldn't be so dificult to be made, but the gameplay would be more arcade. (i have very little experience with UT3, tried to export Tifa to UT3, but i gave up when i came to export the hair properly). I think it would be cool a Macross mod for Crysis, my PC can barely handle this game and also i have no idea where to start a mod.

2 rendered shots from my tifa model. Sorry but i couldn't find any in-game screenshots from UT3, also i don't have the game anymore in my hd. The body looked better because of the post processing effects (HDR and other things). Just the hair i couldn't do well: 1) couldn't apply the physics to the hair, i knew how to do but i gave up as it would give me too much work, unless i changed tifa to use a ponytail because that's what the game could use out of the box. 2) the engine always lowered too much the alpha transparency texture resolution, making the hair look weird.

 

 

Used hardware render for both.

With this same model (older version) i was trying to do a Minmay, but never finished it. 

I should redo the textures by appling a layer of ambient occlusion rendered texture, would look a lot better.

[Avngr]

Me and Sakura are ( at a somekind "slow" pace of things ) working at a simulator for Macross. We will use an engine that is not yet release, that being the Targetware 2.0 engine. The Targetware 1.0 exists, but it is focused mainly at WWII airwarfare, lacking some of the componets needed for a high tech simulation. You guys can check it out here: www.targetware.net , and gather more information at the forum, wich can be found on that link.

Problem is that I started college, and my course demands a little more of dedication, so I am not contributing much ( sorry Sakura ) for the progress of the game.

As a WW2 simulator, targetware is an excelent engine, wich simulates the flying of the aircrafts way better than the other combat simulators around, even though it lacks a little in the graphic part. But this, and all other things missing will change with the release of the next version, wich will be very flexible, making basically anything the game modder thinks off. Talking with a developer one of this days, he said that we will have a full sized galaxy, space phisics, as well as planetary, where you can have a planet with mars gravity, as an example. Ships will be possible too, I asked him if it would be possible to have 2km long vessels, he said that it will be possible to have even 200km ships. A very good Macross game will be possible to make with this engine.

Ah, the current engine hosts a virtually limitless players number, although numbers only up to 100 were tested. Imagine the possibilities...

[evirus]

Chain of command:

Players join a squadron under the command of a squad leader (like in BF2). Squadron leader designates primary goal (IE cover Blue Squadron, Bomb Primary target, Defend Sector XX)

a way to encourage players working together in squads would be to have a player who stays near his leader gain points for how long he does so. give the player more points for completing a squad objective than just doing stuff on their own.

[Avngr]

a way to encourage players working together in squads would be to have a player who stays near his leader gain points for how long he does so. give the player more points for completing a squad objective than just doing stuff on their own.

That's too arcadish IMHO A better idea was to let the player do whatever he wants. In a space environment, in constant war, rogue players wouldn't last long, be sure Unless the enemies were made too easy. If the enemies are modelled to have real capabilities, or even better, if the enemies are to be players, then squad flying would be essential if one would want to come home alive. If the game designers make the plane count to work, player would be really encouraged to bring their kites home safely. Imagine you have 200 fighters on your ship, fighting in a campaign, away of any fighter factory. If the players start to lose planes, it will impact on your fleet's combat ability...

[Avngr]

There's orbiter, if anyone wants to give it a try... It's free, and completely mod-able

http://orbit.medphys.ucl.ac.uk/

[SpacePirateNeko]

well whatever any of you decide..lemem know id love to make some models and textures for this stuff

[Chronocidal]

Hmm... Well, I guess the big question is how detailed is the goal. I've worked with MS Flight sim for the past 6 years or so now, so I can tell you a couple things that might or might not work well for a sim layout.

Going from your list:

"The player never has a magical external camera."

-I'd actually advise against this point very heavily, since no sim I have ever used lacked an exterior view of some sort (they may have existed, but probably not since the days of the old "big blue rectangle above the big brown rectangle" sims in the earliest days of computers). No, it's not realistic in the pure simulation sense, but flight sim people LOVE to make movies, and people love to watch instant replays of battles from the outside, or just fly around for fun. Besides, what good is a full transforming Valk sim if you can't even go "Oooh, ahhh!" at the mecha pron? You can't tell me anyone is going to be satisfied with the fleeting glimpses of enemy mechs flying by your cockpit.

"Player is able to see the pilots body operating the controls."

-I've never seen this work well for one reason: phantom-neck syndrome. Yes, you can model the body in so that you can see it... but unlike a real head, sim viewpoints are usually either fixed or variable points that rotate about the camera's center point. You look down too far into your lap, and you're going to be looking down into your own neck. Yes, you can limit the view, but keeping the pilot's shoulders out of the view means there is a large range where the head won't be able to move. I don't think I've ever actually seen a sim that does this either. Everything I know of just uses an invisible pilot, with the controls animated as if someone was operating them.

"Missile targeting and shoot down capability"

-While this would be nice, I'm having trouble seeing how you could ever implement it. It would have to be automated somehow, because.. well let's face it. If you're trying to make this an actual sim, and not a sim/fps hybrid, you're going to be aiming the gunpod with a joystick, and clicking on the missiles to shoot them down just wouldn't feel right to me. The only way I see this working is if the whole gunpod works similar to an AH-64 Apache, where the pilot's targeting control is linked to the gunpod, and it autotracks whatever he selects (or in realistic terms, looks at).

On that point though, controls may be an interesting issue.. I can't think of any existing system that would offer good controls in all modes to be honest.. and short of building a custom HOTAS system, it's gonna take some creativity to come up with workable controls.

Control stick is kind of a given, and there's not much you need to setup for that. Same for the rudder/foot controls... provided you have a set of pedals for the computer, most of these include both rudder and braking functions. Unless I'm remembering incorrectly, didn't the pilots step on the rudder pedals to make downward thrust in battroid/gerwalk? Link this up to the brake pedal axes, and you could have a nice setup. What would be better is if you could move the two pedals independently, but that would be.. well, complicated at best.

The throttle is the biggest problem I see, since the VF-1's throttle is actually a 3 axis joystick.. it slides forward and back for throttle in aircraft mode, then rotates 90 degrees to become a second joystick in gerwalk/battroid. For this.. I'd almost suggest trying to allow support for multiple joysticks. That way you could have a spare left hand joystick to switch to when needed, instead of relying only on a throttle..

Gah.. it's so late, I need to be sleeping. Anyway, I've thought a lot about how to get these things working in MS Flight sim.. and actually, a good fully variable Valk in FSX may not be as hard to do as we all think (I've recently learned a system that actually lets you directly override the physics engine, and program your own flight model into the sim.. been using it for a repulsor lift system on an X-Wing, but I'm sure it could be adapted for a Valk as well). The biggest issue with FSX is the fact that you need to vary the collision points for the different modes.. and I have no idea if that's possible.

I don't know how much I'd be able to contribute to this project, short of maybe doing some models/textures. But I'd be happy to lend a hand when I can.

[mk16]

Ah i would love to see some kind of valk sim/game on the pc.

Theres this valk sim

http://valkyrie.sourceforge.net/HTML/main.htm

But its been ongoing for like 8years and not much progress is made really.

Mostly because the developer doesn't get a lot of interest from the fan base about a sim for macross

Also the main developer for That sourceforge sim is an aviation/sim programmer so hes got experience making sims

More people should show interest and he may prob get encouraged to finish it one day