Visuals over physics?

[boriz]

KSP 2 enjoys new visual tech. But does it do physics as well as KSP 1?

[Klapaucius]

7 minutes ago, boriz said:

KSP 2 enjoys new visual tech. But does it do physics as well as KSP 1?

Judging by all the interviews with the developers, I think it will do it better.

[intelliCom]

The presence of procedural wings implies that KSP 2's aerodynamic model is going to be better than the previous.

Additionally, it seems like KSP 2 may have weather, which can have a significant impact on an aircraft's flight.

[jastrone]

we have seen some gameplay videos and the basic physics seem about the same as ksp 1 but are probably better(less kraken attacs)

[SciMan]

For one, the presence of procedural wings does not imply better aerodynamic modelling, because the mod for procedural wings in KSP 1 is usable with stock aerodynamics, you don't need to install FAR. The two mods do work well together, but they're not dependent on each other.

For second, with this kind of game asking that question is kind of disingenuous. You can't have good visuals in this kind of game without having good physics, because having bad physics in the game makes the game look bad more than good graphics makes it look good.

[K^2]

Having a physics engineer on board is a good sign. They're a bit green as far as formal experience with game physics goes, but good acafemic background and I'm sure a number of side projects. So hopefully, they'll do a good job.

It sounds like KSP2 still runs GameObject system, rather than ECS, so we're stuck with PhysX as a base. That's the same physics engine as KSP, so some problems are likely to remain. However, it has been confirmed that we are getting continuous collisions, which helps a lot when two ships collide, and there has been talk of "physics LoD." We don't know exactly what this means, but likely, ability to simulate fewer joints by combining some parts into rigid bodies. If Intercept can manage to make that work without struts summoning the Kraken, it could easily improve performance and stability for larger rockets.

So we're definitely getting some improvements, but how much difference these will make remains to be seen.

[JoeSchmuckatelli]

14 hours ago, SciMan said:

procedural wings

I keep reading about these and assume I know what people are talking about - but then I constantly discover on this forum that my assumptions are incorrect. 

Will someone clear up exactly what procedural wings are and how they differ from the stock wings? 

(are they configurable or just a single set piece?  Etc.) 

[pandaman]

27 minutes ago, JoeSchmuckatelli said:

I keep reading about these and assume I know what people are talking about - but then I constantly discover on this forum that my assumptions are incorrect. 

Will someone clear up exactly what procedural wings are and how they differ from the stock wings? 

(are they configurable or just a single set piece?  Etc.) 

Essentially they are 'drag to the size and shape you want' rather than the current method of assembling them out of set prefabricated pieces.

[JoeSchmuckatelli]

5 minutes ago, pandaman said:

Essentially they are 'drag to the size and shape you want' rather than the current method of assembling them out of set prefabricated pieces.

Do you still have to attach control surfaces? 

[TheOrbitalMechanic]

Just now, JoeSchmuckatelli said:

Do you still have to attach control surfaces? 

I believe they've said control surfaces can also be "drawn" onto the wing parts (and resized).

(And not to turn this into a thread about procedural parts, but I hope the old stick-wing-parts-together method remains in some capacity for people who would rather build like that.)

[K^2]

@SciMan is correct, though. Procedural wings do not imply better aerodynamics or physics. They do encourage a sensible lift model, but even that would basically mean about the same quality as KSP is now. Which, honestly, I'm fine with. Some changes would be welcome, like transonic effects and lifting body forces, but there are more important things to get right in physics and simulation in general, IMO.

[Chilkoot]

On 2/13/2022 at 4:46 PM, K^2 said:

@SciMan is correct, though. Procedural wings do not imply better aerodynamics or physics. They do encourage a sensible lift model, but even that would basically mean about the same quality as KSP is now. Which, honestly, I'm fine with. Some changes would be welcome, like transonic effects and lifting body forces, but there are more important things to get right in physics and simulation in general, IMO.

I'm hoping the procedural wings lead to lower part counts, fewer joints, fewer individual drag calculations,  etc.  While I don't believe there's anything implicit about their inclusion pointing to improved modeling, there's a case to be made that as a part in general they could lead to smoother gameplay for most designs.

[K^2]

25 minutes ago, Chilkoot said:

I'm hoping the procedural wings lead to lower part counts, fewer joints, fewer individual drag calculations,  etc.

You still want to run drag and lift computations over multiple points along the wing, and that does still scale with the wing span, and you need at least one point for every aspect and sweep change along the length, but these are relatively cheap to compute compared to the rest of it. And yes, absolutely, on everything else! There will definitely be improvements to stability and performance of aircraft due to introduction and use of procedural wings, but as you say, not because of any specific changes to the actual physics.

[fragtzack]

think the most important point is that Kraken physics will be returning louder, bigger and better then ever before.

 

Edited February 19, 2022 by fragtzack

[SciMan]

Kraken physics, hopefully not. Kraken Tech, on the other hand, that should be returning in full force! Nothing beats a physically implemented hinge joint when it comes to making said joint act as it should IRL (with some friction and slop due to the various parts used, and most importantly not stretching until something actually breaks).

Of course, the only way to make a joint that's superior to that, is to have it baked into an animation on a part and that part not use the unity joints system at all except for connecting to other parts (like how landing legs have a deploy animation but you can't make the leg "fail" by putting something in the way of that deployment animation, instead you'll get bending of the Unity joints between the parts that connect the leg and the obstructing part).

Basically, if you want a huge hinge, you can't rely on robotics parts simply because of the way they're made. You have to make an actual bearing of some kind, the easiest example I can think of is a stick made of Mk0 fuel tanks placed in the hollow section of a Structural Fuselage (the 1.25m non-configurable one). Yes, it has plenty of slop, and likely too much friction, but if you add some tiny landing gear wheels to the structural fuselage suddenly you've made a functional ball-bearing hinge. More wheels used makes the hinge more rigid, but even then because of the structural fuselage that is used as a basis for the outer "race" of the bearing, there's an ultimate limit to the amount of deflection you'll ever see that hinge undergo, assuming all the other Unity joints stay as rigid as they are supposed to (but never really do).

[Chilkoot]

On 2/15/2022 at 11:15 AM, fragtzack said:

think the most important point is that Kraken physics will be returning louder, bigger and better then every before.

 

I think the Kraken has had its day.  It was frequently funny in the early days of an indie KSP, but with a fully funded and polished game like KSP 2, Kraken attacks will more likely be perceived as bug.  I don't expect a lot of tolerance for quirky, game-breaking physics sim mechanics in this release.

[K^2]

On 2/18/2022 at 7:11 AM, Chilkoot said:

I think the Kraken has had its day. 

This is still PhysX. There will still be Krakens. Though, hopefully, nothing much worse than current versions of KSP to start with, and maybe much better a few patches in.

And to be fair, I give PhysX a lot of crap, but regardless, it's very hard to have a physics simulation that's perfectly stable in all scenarios. And KSP is a sort of game where players will keep finding new ways to break things. But getting things down to a point where you don't encounter Kraken unless you try to build something truly giant or bizarre is achievable, and I'm cautiously optimistic about KSP2 in that regard.

[TT213]

I was wondering, will we have more advanced physics ? Compression lift for example, or improved bouyancy.

[Vl3d]

Modern unity volumetric clouds (physics based rendering, HDRP):

Also terrain demo:

But... this is Unreal Engine 5:

 

Edited April 4, 2022 by Vl3d

[shdwlrd]

@Vl3d OK, that is worth a years delay. Also the Unreal demo is weird. A hole pops into the terrain and the surface debris goes flying. Where is the debris from the ground itself? It's weird.

[Vl3d]

Oh, I actually agree that Unity is just fine for KSP2. Yeah, permanently destructible terrain would be nice to have but it's not necessary. But what we really NEED are those Unity volumetric clouds and weather effects.

[SciMan]

I'm one of those people that's played video games for so long that I don't really care what they look like anymore. Yes, good visuals are nice and all, but that's not why people keep coming back to a game.

For that, there's only one thing that fits the bill. That's the gameplay. And for a simulator like KSP (and as much as it's a game, it's even more so a simulator), the only gameplay that matters is "how accurate are the physics".

So yes, incorporating things like Compression Lift or some way of calculating buoyancy of a ship's hull that's made from structural panels and not fuel tanks (but I'll just take a roughly tenfold reduction in the amount of drag that the water creates on things, perhaps combined with making aerodynamic parts like fins have increased effectiveness underwater) would be things I'm interested in much more than something like "clouds and/or volumetric lighting".

All those graphics improvements seem to do for me is make my graphics card run hotter, they don't make me play the game any better, or let me build better rockets or other vehicles.

 

Why not both tho? The PhysX physics engine was originally made to run on a piece of hardware mostly similar to a graphics card, right? If a game like Dyson Sphere Program can use the GPU to simulate a gigantic interstellar factory, why can't it simulate the physics of one or more rocket-propelled vehicles such as spacecraft?

I know, you're going to say "it's not that easy!" or "it's more complex than that". At the end of the day, it's all just math. You can solve an incredibly greater number of math problems on a graphics card than you can on a CPU, simply because of the greater core count (that's one of the reasons DSP is such a good factory game, even if it is somewhat more simple than a game like Satisfactory in the graphics department).
So why can't you do that with a flight simulator? I'm pretty sure that for professional flight simulators that are used to train real pilots, that the simulation is largely being run on what's essentially a graphics card without the video rendering engine part of it (in other words whatever Nvidia is calling their purpose-designed GPU computation card these days, the last name that I can think of is "Quadro").

[shdwlrd]

34 minutes ago, SciMan said:

I'm one of those people that's played video games for so long that I don't really care what they look like anymore. Yes, good visuals are nice and all, but that's not why people keep coming back to a game.

I'm the same, but Kerbin always felt empty and any ground scatter that was added is poor at best. (I've seen better ground scatter in games from the late '90s.) 

41 minutes ago, SciMan said:

And for a simulator like KSP (and as much as it's a game, it's even more so a simulator), the only gameplay that matters is "how accurate are the physics".

I think the better question is "How accurate do you want the whole simulation?" The physics is only one part of the simulation, how the objects act in the simulation is the other part of the equation. You touched on that later in you post. 

[MechBFP]

5 hours ago, SciMan said:

At the end of the day, it's all just math. You can solve an incredibly greater number of math problems on a graphics card than you can on a CPU, simply because of the greater core count

This is plain wrong  just so you know. CPUs and GPUs are made to process different things in different ways. 

The parallel processing of a GPU is worthless for something like the physics of a ship in KSP1 because it’s not a process that can parallelized. 
As such a CPU with higher core speeds is more appropriate in this case. 

Edited April 5, 2022 by MechBFP

[darthgently]

6 hours ago, SciMan said:

I'm one of those people that's played video games for so long that I don't really care what they look like anymore. Yes, good visuals are nice and all, but that's not why people keep coming back to a game.

Bingo.  There was actually a paper on this way back.  Game players wanted graphics enough to provide useful information, beyond that, responsiveness was far more important.  There, of course, can be some big overlap in good looks and usefulness.  Trying to spot a good landing site while descending into mountains on Moho is a great example where good graphics could make a huge different.  They aren't mutually exclusive.  But when the looks overhead causes enough lack of responsiveness that it causes the player to get out of sync and make mistakes, then that is too good looking and not worth the overhead.  When the player presses a button or key, or moves a joystick, the game should respond very fast at the cost of graphics quality if necessary

Edited April 5, 2022 by darthgently
typos, spelign erorrs