Atmospheric flight

That’s been somewhat mitigated with some of our optimizations and newer codebase; as far as I can tell, you can now scoot pretty fast at ground level and mid altitudes.


The 4D procedural field isn’t for my ship, but for the planet. It’s the weather system that you’re after, but without having to model high and low pressure systems and figure out how they interact. It’s the same technique used for clouds without having to come up with moisture levels, evaporation patterns and air pressures. It’s the same technique used to have mountains and steppes and valleys without having plate tectonics, volcanism or erosion. It’s all client-side and localized, meaning that it can contain very high fidelity data. Just like clouds and terrain.

The real complexity comes in for things that @mattk50 mentioned: moving bits and pieces that represent a significant portion of the ship’s surface area. Those things should alter the ship’s aerodynamics, but would require real work to get right. The fact that Infinity ships seem to lack those things is a win for the team; they don’t have to worry about getting that stuff right. The ships are essentially bricks.

My preference is to just see small ships being more capable in atmospheres. If we assume that atmospheric flight in turbulent air requires either power or maneuverabilty, it is going to be the small ships that win out. They have the best acceleration and the quickest response times. So when mother nature starts battering a ship, the pilot will be able to regain control much more quickly in a small ship. Meanwhile, the large ship is going to be lumbering around. High up, that’s not a big deal, but lower to the ground, it could be a real problem.

So the more energetic the atmosphere, the smaller the ship I want to be in. I don’t have to add gear to it that allows me to operate in atmospheres. Rather, it is the nature of small ships that lets me do well. As always, I hope to avoid gear wars, where victory was already decided in the hangar.

1 Like

I expect we’re on the same page here, but just wanted to expand on this a bit.

While of course smaller ships would be much more manoeuvrable, I’d imagine them to be more easily tossed around by high winds. - But then as you said JB, able to recover more quickly.
I would expect bigger ships with higher mass and inertia to be less manoeuvrable but more stable in gusty wind, but if their pilot flies them into winds that are too strong and the ship gets set tumbling, they’d be more difficult to regain control of.

Maintaining a dead straight course in high winds in a small ship would be very difficult but if you just keep pointing where you want to go you’ll get there, but your course won’t be a straight line. The skilled pilot would react dynamically to changing winds and use the wind to their advantage.

Even a low-skilled pilot in a small ship would be able to easily out-manoeuvre a big ship in atmospheres. In high winds the small ship with low-skilled pilot would be difficult to hit as their natural manoeuvrability is complimented by the ship being tossed about on the wind.

Yeah, we’re on the same page. I prefer to see small ships being the vehicles of choice for getting things done on surfaces. They physically scale better against buildings and other surface facilities, and they complicate the task of doing anything on a large scale. That complication is gameplay.

So instead of a 1km-long ship landing and dropping off a 5GW nuclear reactor, that 1km-long ship stays in space and sends multiple 100m-long ships down with bits and pieces of the nuclear reactor that must then be assembled on the surface. Game designers usually just go with the first approach because it’s much simpler to implement (which players have gotten used to) - but the goal is not to present a fiction, but to create gameplay - stuff that occupies and entertains players.

WIth that in mind, I wanted to handicap large ships in active atmospheres (in truth, any atmosphere). All those ships flying down to drop off parts for the nuclear reactor have to face the tumult of crazy winds on each trip. On any given trip, they might lose any one of a dozen different types of parts, for which the 1km-long ship may not have any replacements or backups.

What if ships could transport stuff externally via a sling system? It would allow for much bulkier loads to be transported, but it would also mean that sufficiently violent maneuvers could lose that load. And the load itself may impose maneuvering restrictions. Handling of the ship may degrade with larger loads, especially at the limit of the ship’s capabilities. More flying challenges for players.

First off, welcome to the forums, you must be new :stuck_out_tongue:

The INE is all about procedural generation, remember? You pick a seed, sync a timestamp with the server, and each client generates an entire galaxy using various pseudorandom number generators, noise functions, and what not. No further synchronization needed - because there’s no such thing as truly random numbers being generated by the sort of computers that will be used to play the game.

A 4D procedural wind field is just one more layer to the galaxy generating algorithm, same as star systems, planets, orbital motions, and everything else.

We should try to minimise overlap between this topic and the Capital ships, atmosphere, planet surface and landing topic.

If we assume that big ships (size discussed in the other topic) can’t approach too close to massive bodies, in this topic we can discuss the gameplay surrounding ships inside atmospheres.

I’d prefer it to be called “returned after long absence” (although I did occasionally, once every two to six months or so, read up on some threads, just never had anything of substance to add to the conversation. Which makes it all the more regrettable, that my main incentive to post is because I, once again, have a different opinion than JB / misunderstood him)

I do think we want basically the same this time (seeing how his ‘4D procedural field’ is on planet scale), I just would prefer to have wind created using a semi-believable model. To prevent something immersion breaking like heavy rainstorm on a planet that is devoid of any liquid. As long as it is feasible in real-time of course, I can see how it might be possibly too much to ask for.

4D fields are not ‘mine’. They’re a standard part of procedural noise generation. Perlin noise. Simplex noise. And so on.

Also, @Runiat’s chiding is something you should take to heart if you’re going to comment on procedural generation techniques. You’ve seen the procedural terrain that various groups have generated. It’s really nice stuff. You’ve seen procedural clouds, and they too can be gorgeous. Procedural forests. And other applications of procedural generation. The results look good because they are controlled by the developers. That is, developers don’t just pour ‘procedural generation’ into their code and weird things happen. It is carefully designed to produce a specific range of results.

So there are no forests with trees extending off into space. There are no valleys so deep that they poke out the other side of the planet. And there will be no rain on planets without liquids in the atmosphere.

(Excepting obvious exceptions, of course.)

I think you’re missing my point. I am not saying that procedural clouds can not be gorgeous. I am just concerned, that just pushing yet another layer on top of it, without referencing to the other procedural layers, might end up in an immersion breaking experience.

I tend to agree with you dekaku.

Applying too many variables could end up with a total nightmare for any programmer. If a ship has virtually no wings, it would have to be flying extremely fast to retain any lift. Alternatively, there would be some form of vector thrust to keep it flying. If it was moving fast, the clouds would be of very little consequence since you would be through them in the blink of an eye, a slight shudder is all you’d really experience at most.

Of course, there is the possibility that flying too fast may result in severe damage. Water is soft to land in unless you fall from 10K feet, in which case it’s like hitting concrete. So too would the thick cloud base be more destructive if you hit it hard enough. On the other hand, if you’re travelling at those kinds of speeds, clouds are the last thing you need to worry about since you’d be unlikely to slow down enough to avoid becoming a large crater in the ground.

Sorry to necro, but it’s a relevant thread.

First of all, I realise there is still plenty of work to be done on the atmospheric flight models and the end result will ultimately depend on the resources available, but I just wanted to discuss something that perhaps may not yet be a part of the team’s vision.

When I watch the videos and streams showing off flying above the surface, I can’t help thinking that it doesn’t look much different from being in space. Again, this is of course likely going to change and it might actually feel different than it looks, but it got me thinking about the use of flight assist.

Currently ships can come to a complete stop and hover using flight assist. What I would like to see is a fair amount of “struggle” as the vehicle shakes and sways from the thrust it is producing to counteract gravity. It would be rather dull to just have people sitting motionless next to a factory shooting at it for a few minutes.

Yes I’ve already brought this up many times with the team, all very much WIP.

1 Like

I would hope that sitting motionless would be a good way to get dead. I don’t see a need to make spaceships struggle to stay stable unless there are forces actively destabilizing the ship, such as powerful winds.

Make installation destruction involve something like dropping charges down a hole to destroy a Death Star, with lots of flak towers to make that maneuver difficult. So instead of just wearing away at a target with a volume of fire for a period of time, structure the attack so that players must make difficult maneuvers, dodge static defenses and enemy ships. Nobody will be sitting still anywhere near that factory unless they have a death wish.

Well sure I would hope at least there would be some defences to encourage movement, but for a hypothetical where there is no resistance, the vehicles should still feel as though they are in an atmosphere.

It’s always good to have an analog kind of baseline roughness to simulation, even if getting just the right dose is a really subtle thing.

Another e.g. when firing RCS there ought to be a tiny barely perceptible deviation from input (duration, sputter, etc). Never enough to significantly affect a critical maneuver, but enough that it gives the spacecraft character. Just like with real vehicles (cars/bikes/etc).

Flying a commercial airliner yields no visible motion from the outside. Shaking is noticeable to the occupants when passing through turbulent air, but that’s about it. So unless a ship capable of multi-g accelerations on every axis is somehow being tested by the environment, I don’t believe there should be any wobbling or wiggling. Remember that the control system must be precise enough to keep the ship smooth in space.

If you want such things, take spacecraft out of the atmosphere and send in something that is designed to be the efficient solution for that environment - an aircraft. They’re cheap, they’re light, they can be challenged by the environment and they can be as streamlined as you like. You’ll end up with atmospheric dogfights.

When I say take spaceships out of the atmosphere, I mean that they should spend as little time maneuvering in the atmosphere as possible. It’s bad for the thrusters. Besides, the weapons designed for space don’t work in atmospheres anyway.

Another twist is to say that spaceships turn into aircraft when they’re in atmosphere. That is, the control systems change because the thrusters are damaged by use in atmospheres. Instead, the ships only use special engines on the back, and then rely on their aerodynamics to fly around. The guns switch over to machine guns instead of vacuum-loving plasma rounds. Instant jet aircraft conversion.

Another alternative is to make atmospheres so turbulent and thick that they really do challenge the ship controls. Selling that visually might be rather a difficult technical challenge.

In the prototype one is unable to change the heading of the ship when entering the atmosphere going really fast… This is the effect of aerodynamics. If the thrusters are made week enough, the same effect will be present at lower speed/thinner atmospheres. This would result in jet plane movement… higher speed = larger turn radius.

Yeah, that same thought occurred to me today and I was eager to get back here to suggest it - but you beat me to it. I think that having thicker atmospheres make thrusters weaker would provide an interesting point of game balance. The deeper you go into the atmosphere, the lower the efficiency of your thrusters. Diving deep into a gas giant would completely shut down your thrusters, dooming you to fall ever deeper until you hit crush depth.

Letting players sacrifice vacuum efficiency for atmospheric efficiency would be interesting as well. I may tune ships for atmospheric use, but they’re rubbish in space. So send the carrier a bit into the atmosphere, then launch the atmospheric fighters.

Some aerodynamic effect from structural asymmetries would be nice as well. Induced yaw, etc.

Very high speed and maximally smooth flying by pilots/autopilots, and a craft designed analogously to a cadillac, to maximize comfort.

Not the right comparison for this very specific thing I’m suggesting.

[quote=“JB47394, post:31, topic:955”]
unless a ship capable of multi-g accelerations on every axis is somehow being tested by the environment, I don’t believe there should be any wobbling or wiggling.
[/quote]But you should feel the aircraft, then and in as many other situations as possible. If the craft is in still air, then you might feel nigh-on imperceptible defects in the attitude thrusters. If the craft is in turbulence or being pushed on by wind, then you should hear, feel, etc, the RCS working like clockwork to nullify that interference with your job (piloting, putting a bead on a target, etc etc).

Remember what makes a good, blood pumping machine. That is, it has “soul”.

Listen to all those mechanical flaws… Are they not beautiful? Are such machines really worse off having all those adorable flaws removed for the sake of perfection? Why not have the best of both? Those flaws only need to be kept out of the way; they don’t need to be entirely removed.