Is there additional work on the flight modeling planned?

[Bosco19 198]

The other aspect of Arma 3 that makes helicopter flying so enjoyable - is on a server that uses a mod called Quicksilver, It features an outstanding 3D polygonal map that replicates objects in a convincing real world, in stunning detail. The map  is huge, and holds a group of large and small tropical islands with varying terrain, and a large assortment of objects, including small compounds and villages, and even a small city with skyscrapers. You can maneuver in and out and around the 3D terrain, landing on objects, down in between trees and rocks, really, anywhere your skill set allows you to go. These objects are fully detailed and textured, and maintain their details no matter your distance from them, unlike the slab-sided boxes presented in MSFS. There is an assortment of different helicopters, all based on real world models, each with different flight characteristics. I thought that if developers could bring this level of realism to a title that's really centered around a realistic FPS military sim, with the aircraft just an adjunct, then certainly a modern dedicated flight sim could be expected to do as well or even better! I haven't tried DCS, so I can't speak to that, but at present, if I want a taste of what what it FEELS LIKE to fly in the real world, at least in a rotorcraft, I still have to boot up ARMA 3. It's not really a flight sim, but it DOES simulate the FEEL of flight in a very convincing way. BTW, this will be of little interest to all the button pushers out there!

[jcomm 5,939]

Yet another interesting thread at the Dev Support of flightsimulator.com.

It's also the first time I saw An Petrovich post there:

How does the need for accurate geometry balance against the lack of adjustment? - MSFS DevSupport (flightsimulator.com)

 

Edited April 3, 2023 by jcomm

[Bosco19 198]

Very interesting comment. This is out of my purview. I am not a programmer. But the question in play here seems to be: What's more important - the geometry and behavior according to the use of accurate physical dimensions of the aircraft model, or whether or not the behavior in the sim feels convincing. I am obviously on the side of the latter, and I think this addresses the underlying philosophy of the sim design itself, and it's intentions. I realize that many users really get into what I have described as "button pushing," and there is a place for that kind of activity in a sim. might it not serve more users to have a sim with two different intentions - One being a "procedures" sim, and the other being a "flight" sim? Lets face it - at least at this point in time no consumer-level sim is ever going to truly simulate the actual real-world flight experience. It's just not possible (though a respect them for trying) The biggest missing part is gravity and physics. In a real aircraft, the forces applied to your body are just as important to what you are seeing and hearing. The pro's get some of that in their massive moving sim capsules. We are never going to get that on our desktops. People like me are more interested in the physical aspects of flying the aircraft themselves. Early out of the gate, when I was flying typical rental aircraft - Pipers, Cessna's, and Beechcraft - I realized that my interest was not in flight planning, managing a busy cockpit, and adhering to a strict set of rules and regulations concerning what I was allowed to do with the aircraft (I was also a radar air traffic controller). I was more interested in cranking-and-banking and attacking challenging landing scenarios in varying conditions and in a variety of different aircraft. Low-and-slow - that's me. Button pushers might most likely be satisfied with a good IFR experience - pushing buttons and following strict chart procedures while flying in the blind. People like myself want to get into an aircraft with predictable flight characteristics and see if we can put it down successfully on an impossibly short runway - amidst an impossible crosswind. To do that in a "sim" I think it would be better to assign rather basic general flight characteristics to the aircraft type- in essence - you fly a tail-dragger differently than you fly a mid or low-winged bird. If you know what it's like to fly an Interstate Cadet or a Piper Cub, you could fly one with rather basic characteristics in a sim, and after executing a landing, say to yourself - "That's kind of what it felt like in the real aircraft." It's because the programmers are trying to do it all - to do the impossible - that we have a sim like MSFS, that, unsurprisingly, comes up short in so many areas.

[Noel 4,507]

This is a weird discussion to me because after all a flight model in a desktop sim is a data table with a bunch of variables that impact the FM that can be modulated by the sim and SDK for plane development.   This is so old now, knowing the type of relationships those variables need to have in order to have a convincing level of inertia and various responses to inputs from controllers, friction and wind etc.  IOW, this is not new pioneering technology and effort it's all been done before and even in MSFS' legacy sim.  So this begs the obvious question:  how come?   Did the dev team leave out a few critical variables?   Surely Asobo is reading all of this of course they'd have to, so again, what's behind the perceived deficits?   

[jcomm 5,939]

13 minutes ago, Noel said:

This is a weird discussion to me because after all a flight model in a desktop sim is a data table with a bunch of variables that impact the FM that can be modulated by the sim and SDK for plane development.   This is so old now, knowing the type of relationships those variables need to have in order to have a convincing level of inertia and various responses to inputs from controllers, friction and wind etc.  IOW, this is not new pioneering technology and effort it's all been done before and even in MSFS' legacy sim.  So this begs the obvious question:  how come?   Did the dev team leave out a few critical variables?   Surely Asobo is reading all of this of course they'd have to, so again, what's behind the perceived deficits?   

None in the initial team was an aerodynamicist, to my knowledge, so, "just" having a degree in physics and/math doesn't provide the necessary insight into flight and flight dynamics modelling... 

They probably believed their approach was good enough... It isn't, not even by far...

Maybe the best option is to think of a Legacy version with no quirks from the new weather physics, also flawed, impacting it...

Edited April 3, 2023 by jcomm

[Noel 4,507]

1 hour ago, jcomm said:

None in the initial team was an aerodynamicist, to my knowledge, so, "just" having a degree in physics and/math doesn't provide the necessary insight into flight and flight dynamics modelling..

I'll buy the 2nd possibility that they felt their approach was good enough, and there is no question IMM that most users believe the must believe the same and well after one year in operation here on Avsim most RW pilots spent most of their time in MSFS over the others or so the poll indicated.  What makes you think there was someone at Asobo who had a degree in physics/math was employed to develop its FM?

[Sonosusto 681]

Yeah, there will be improvements in the future. 

Lately, I just adjust some sensitivity and reactivity settings for my honeycomb alpha. I have different settings for different planes. Not ideal but it works. 

Example: I have the 737-700 and 900 set with less sensitivity and less reactivity etc.

GA planes have more sensitivity etc. 

It "feels" better. 

[Bosco19 198]

The ARMA 3 experience I have described is a perfect example of the kind of "generic" treatment you can put into a flight model that suffices for a convincing control and response feel. It's definitely not a simulator in the vein of MSFS. But is it realistic? Absolutely. You can go into third-person view mode and watch every physical action and reaction that a helicopter makes in real life, down to how the rotor tilts to offset sideways translation induced by the turning rotor, how the rotor pitch changes as it rotates, advancing and retreating, how the helo tilts in response to the control inputs required to hold it in a balanced hover, and more. Even their twin rotor (Chinook) model is accurate.The flight mechanics of the aircraft are all there. Now, to get a sense that you're in a different bird you just have to play with the established parameters a bit. Every one of the half dozen or so models provided in ARMA successfully convinces you it's unique. The best aspect to me is that they have really nailed the sense of the physics of rotor blade energy. Even the dumbed down standard flight model ( the model most used in-game to get  troops to the landing zone without killing them) retains that sense of rotor physics. The advanced flight model requires so much skill most casual players never even activate it. It requires the same control inputs and thinking ahead that the MSFS models proclaim are necessary, but everything is smooth and predictable. A tiny control adjustment alters response without getting into the wild runaway reaction that suddenly shows up at some unexpected point in the collective adjustment in MSFS. If you don't plan and time your inputs perfectly in ARMA you can get into the same deadly oscillations exhibited in MSFS. One thing that's missing in ARMA is the sudden bizarre twitching that besets the MSFS helos periodically, for no apparent reason. Again, never having flown a real helicopter I can't say MSFS's characteristics are inaccurate, but the ARMA programmers' design feels legit, and though it still requires hours of practice to conquer, it's actually doable if you keep at it. And it's fun! There's a lot missing in ARMA from a simulator aspect. Though there are full panels with readouts, there is almost no interaction with the pilot. Most options are activated through popup menus. But there isn't any reason they couldn't extend this software to accommodate them. I think the bottom line is simply that MSFS attempts to do too much. The sim is just too massive to ever achieve perfection in all things. It's not even reasonable to expect it to.

 

[Noel 4,507]

7 minutes ago, Bosco19 said:

I think the bottom line is simply that MSFS attempts to do too much.

...and we can add the caveat, in too short a time.  XP has been around 26 years under the same dev...

[Bosco19 198]

Don't we have to ask what the expectations of the user are? I think a lot of users who have no actual flying experience want the involvement of a complicated cockpit because it seems more real to them. They don't mind an inaccurate flight model. They have no experience to judge by. I can see MSFS being a really good way to practice procedures, even for real world pilots. Then there are users like myself - old stick-and-rudder pilots who just want to get into a cockpit and yank-and-bank. Yeah, some switch throwing and panel involvement adds realism, but I want to FLY...not just OPERATE. I think the MSFS boys have a near impossible chance of satisfying both types - at least in the near future. Oddly, for me, a title that is actually a FPS with the addition of aircraft (it offers fixed wing as well) and isn't intended as a serious flight sim actually provides a better sensation of actual flying, as I remember it, than MSFS does. I haven't given up on MSFS. I'm just hoping for better.

 

[Noel 4,507]

38 minutes ago, Bosco19 said:

Don't we have to ask what the expectations of the user are? I think a lot of users who have no actual flying experience want the involvement of a complicated cockpit because it seems more real to them. They don't mind an inaccurate flight model. They have no experience to judge by. I can see MSFS being a really good way to practice procedures, even for real world pilots. Then there are users like myself - old stick-and-rudder pilots who just want to get into a cockpit and yank-and-bank. Yeah, some switch throwing and panel involvement adds realism, but I want to FLY...not just OPERATE. I think the MSFS boys have a near impossible chance of satisfying both types - at least in the near future. Oddly, for me, a title that is actually a FPS with the addition of aircraft (it offers fixed wing as well) and isn't intended as a serious flight sim actually provides a better sensation of actual flying, as I remember it, than MSFS does. I haven't given up on MSFS. I'm just hoping for better.

 

That's me, no actual flying experience and yes want the involvement that comes from a complex cockpit and managing all required to get from gate to gate plausibly, and have had that for a long time now.  Now here's what I find odd:  "I think the MSFS boys have a near impossible chance...." while at the same time an FPS with aircraft (as a minor component) has a superior sensation of flying".   I don't think they're at all mutually exclusive are even partially exclusive--they've just not dedicated to getting the FM better aligned w/ the physics involved, that's all.  No magic, no cryptic explanatons--they just haven't done it.  Not rocket science, it's aeronautical science so well developed now that even an FPS developer did a decent job of it per your testimony.  And here's the why which you've already alluded to:  it's a vast minority of total users that really focus on this aspect, the FM, so it's no wonder MS/A allocated it's talent towards what attracts the largest user base.   And also as you've pointed out--FM is only one component of the total model that makes up a FS.

Edited April 3, 2023 by Noel

[smurfboard 9]

I think Microsoft’s decision to market MSFS as an Xbox game has a lot to do with how Asobo is developing it. Follow the money.

[Bert Pieke 7,585]

4 hours ago, Bosco19 said:

I think the MSFS boys have a near impossible chance of satisfying both types - at least in the near future.

Not sure there is an inherent conflict... accurate cockpit interaction should not get in the way of convincing flight behavior..

I think that Asobo set out to create a better flight model than FSX and have struggled to make that better flight model work in all phases of flight.

Re-read what Rob Young posted above.. 😉

Edited April 3, 2023 by Bert Pieke

[Bosco19 198]

Mr. Young does make good points, and illustrates how difficult it is to make a myriad of control situations come together correctly at the appropriate times. Is this a problem that can even be solved at our level? Glad I'm not a programmer!

[Orlaam 312]

I don't think MSFS handles much differently than FSX/FS9 era.  The default aircraft are indeed too twitchy, but we've always had to tone down our controls to compensate.  The ground handling is the only aspect that I wish could improve.  You  should be able to turn better.  It's not like previous sims weren't awful in terms of sensitivity or crosswind handling.  There will always be limitations to the flight model, unless the engine driving the game allows for better physics.  I often wonder how many people complain about the sim flight dynamics but use a peripheral with terrible driver inputs.  I had a CH Yoke for years and it was terrible for banking without diving all over the place.  You  couldn't keep a plane level in a turn to save your life.  The RealAir Duke was one of the few that wouldn't pitch all over the place with the CH Yoke.  The A2A, as great as they were, were very hard to hand-fly IMO.  Yet, when I spend the money to buy the Thrustmaster Boeing Yoke, it was like night and day.  All of the sudden I could hold an attitude with greater ease in a turn and wasn't so annoying by the flight dynamics.

Even 737NG Driver said, in one of his videos, that cadets training in a full motion Level-D sim can't land well in real life because it's just not the same.

Edited April 3, 2023 by Orlaam