16 minutes ago, Chock said:

Let's not forget that for years, everyone said that FSX wasn't even capable of simulating an A320 realistically, then FSL went and did it.

Two concerns I have with the 'let the 3PD aircraft developers hack at the model to make their aeroplane work properly' course of (in)action are: that it makes it difficult for new 3PDs to create an aircraft from scratch to compete with the established names and that any subsequent version updates of the base sim breaks the aircraft.

It also appears that Asobo are making small changes in the background on a regular basis without making any announcements.
How many times has a new point release of P3D resulted in the FSL Airbus not working for some time until a patch is released?

I don't expect a Level-D full motion simulator experience for a home entertainment product.
However, I do expect to be able to take-off without needing the sim's take-off assistance feature, use of appropriate rudder to make coordinated turns, not be bounced around as if I'm flying a balsa wood plane kit in a thunderstorm whenever there's a bit of weather.

A little more effort on the flight modelling to address these issues would, IMO, make MSFS a lot more enjoyable and a lot less open to criticism. 
 

16 minutes ago, Alpine Scenery said:

I would say OOTB Xplane is the best ...

XPlane is the worst. Don't believe what Austin saying. XPlane not an "engineering tool", this is just a good marketing myth. 😄

Edited January 19, 20215 yr by OSM

9 minutes ago, OSM said:

XPlane the worst. Don't believe what Austin saying. XPlane not an "engineering tool", this is just a good marketing myth. 😄

Well I'm assuming his Xavion product works with the glide path and that if he had enough knowledge to do that, given he also has the degree, and hopefully he didn't release XAVION with bad enough physics to kill anyone. So I have to give him that. I am however not judging it purely on his credentials, but I am just judging based on how the landings feel to me as a SIM user, based on how the landings felt when I was in a real 737 or 787, but it's not scientific (I agree). I never thought XP 11 was perfect based on the data, but it seems closer. To be honest, it is nearly impossible for them to simulate real-world turbulence unless we are using a motion simulator, so that is not my primary concern.

Another hint I am using to try to reduce my own bias, is that some of the issue is that Asobo never checked the actual flight numbers and the glide ratios. I do not know if the flaps issue with several of the planes not having enough climb power was purely them not checking it, but seems like mostly it was. So I assume if they were this cavalier to NOT check stuff that could be fixed, then they were also too cavalier on the base flight model of stuff that is NOT yet fixed, and there is evidence that they were. I am not an authority on the topic however, but just my own opinion of what I've seen.

Edited January 19, 20215 yr by Alpine Scenery

4 hours ago, marsman2020 said:

I'm not over in the '500 hours completed' thread filling that thread with complaints about the flight model,

Patient, detailed analysis with corroborating data is welcome. Continual “Why don’t you all agree and accept this game is word not allowed?!” rants and repeated threats to abandon this sim and go and buy XPlane (without ever actually doing it), less so.

2 hours ago, Chock said:

Let's be realistic here, they weren't exactly promises. It's not like they signed a blood ritual in front of us and claimed we could strike them down in their stride if they didn't hold to it. As far as I'm aware, what they said on the matter was a combination of projections of what they were hoping/aiming for within the constraints of producing a nice flying game, combined with a bit of marketing speak.

At no point do I recall them saying anything resembling: 'we're gonna make something which is so close to the scientific aerodynamic data you can pull from of the Dryden Flight Research Center, that you'll not be able to tell the difference.' Now of course we'd all love it to be that realistic, but again I think we need to be aware that it's a fifty quid game where you fly toy aeroplanes around.

We could start saying that Laminar Research's usage of blade element theory is some perfect replication of flight modeling too, but we all know it isn't. These are first and foremost games; it's why you can buy them on Steam. don't get me wrong here, I love flying aeroplanes as much as anyone, but I don't expect a fifty quid game to perfect replicate it, and nor whould anyone else, because if it could, the airline training centres wouldn't be paying 15 million quid for their flight simulators.

Good point, especially when some folks complain about the high price of Pmdg aircrafts.

2 hours ago, F737NG said:

Two concerns I have with the 'let the 3PD aircraft developers hack at the model to make their aeroplane work properly' course of (in)action are: that it makes it difficult for new 3PDs to create an aircraft from scratch to compete with the established names and that any subsequent version updates of the base sim breaks the aircraft.

It also appears that Asobo are making small changes in the background on a regular basis without making any announcements.
How many times has a new point release of P3D resulted in the FSL Airbus not working for some time until a patch is released?

I don't expect a Level-D full motion simulator experience for a home entertainment product.
However, I do expect to be able to take-off without needing the sim's take-off assistance feature, use of appropriate rudder to make coordinated turns, not be bounced around as if I'm flying a balsa wood plane kit in a thunderstorm whenever there's a bit of weather.

A little more effort on the flight modelling to address these issues would, IMO, make MSFS a lot more enjoyable and a lot less open to criticism. 
 

It seems like more of the variables are required to be exposed without encryption, and people like PMDG don't want us to see what they are doing... Perhaps, because they are not making nearly as many changes to the flight model as they pretend to (I have no idea, so I don't want to blame or accuse). Could be partly the flight systems also (having to redo), but probably just a combination of having to do everything within the realms of how all the new stuff works. Could also be they know with all these "DECENT" planes in the default FS 2020, people just may not be as willing to pay the $80 anymore, that I have no idea on. I'm personally not as interested in what PMDG can do as I am in what Majestic can do with their Q400 - Dash 8, but then I don't think Majestic has said anything (do not recall), or maybe they have...

Edited January 20, 20215 yr by Alpine Scenery

Getting back to the original test data, it would be interesting to compare the graphs for the same data using the default C172 in X-plane and and a good payware addon such as the A2A 172 in FSX/P3D. Not that that data is representative of everything that makes up the flight model, but it would be interesting all the same.

6 hours ago, snglecoil said:

Meh...data without adequate context can say whatever you want.

The context: aircraft in MSFS fly like #%@$. Maybe that context is too basic for you to understand and you need more sophisticated context. I don't need anymore context.

And before the dogpile starts I might add that I really like MSFS...but the physics (lack of) is REALLY bad.

39 minutes ago, somiller said:

The context: aircraft in MSFS fly like #%@$. Maybe that context is too basic for you to understand and you need more sophisticated context. I don't need anymore context.

And before the dogpile starts I might add that I really like MSFS...but the physics (lack of) is REALLY bad.

Just curious, how much actual experience do you have at the controls of all these aircraft that fly like #%@$ in MSFS? 

Edited January 20, 20215 yr by snglecoil

1 hour ago, turnandbank said:

Getting back to the original test data, it would be interesting to compare the graphs for the same data using the default C172 in X-plane and and a good payware addon such as the A2A 172 in FSX/P3D. Not that that data is representative of everything that makes up the flight model, but it would be interesting all the same.

I don't have the fancy A2A C172, but I do plan to go collect data for at least FSX and XP11.  Aerobask has the free DR401 for XP11 that supposedly has some fancy flight model tuning going on, so maybe that would be a good datapoint.

Need to setup a script in Joystick Gremlin to produce a repeatable input disturbance every time.

1 hour ago, turnandbank said:

Getting back to the original test data, it would be interesting to compare the graphs for the same data using the default C172 in X-plane and and a good payware addon such as the A2A 172 in FSX/P3D. Not that that data is representative of everything that makes up the flight model, but it would be interesting all the same.

 

Interesting

Edited January 20, 20215 yr by Rouzbeh

30 minutes ago, Rouzbeh said:

Interesting

They are jumping straight into the more complex yaw and rudder issues, would have preferred if they'd have stuck to the lift and drag and glide ratios first on landing, as well as ground effect. These issues are much simpler to analyze.

I am not an expert in the area, but it seems like the type of analysis they did for yaw and rudder is sorely underserved and would have required some error plots, at the very least some type of distribution plot. Otherwise, a single plot would have to have some type of error quantization process summarizing the limitations of the analysis.

Edited January 20, 20215 yr by Alpine Scenery

All these graphs are interesting to check out, but in spite of this, I'm somewhat inclined to go off what it feels like and looks like with this stuff, and to me MSFS seems fairly convincing for normal regime flights. Not perfect, but on occasion it does make me recall moments from when I've been flying the real stuff.

In fact I did a flight earlier on in the Virtualcol Beech 99 and when it lifted off the runway and began to pick up some drift in the crosswind, I actually noticed that it was a lot like the kind of thing I've experienced many times when flying the real aeroplanes at that point when it came to the inputs necessary to keep the thing going the way I wanted it to. This is of course pretty subjective, but it certainly was a remarkable replication of my own experience of flying stuff on that occasion this evening, enough for me to actually take note of it.

I'm a lot less convinced by the flight model when you start throwing the thing around the sky, although MSFS is certainly not alone in that; I've never seen any flight sim do a really absolutely convincing spin for example, and if you try and do flick rolls or anything of that more extreme nature with a desktop flight sim, you're usually in for a bit of a disappointment.

But back with less extreme maneuvers, it's one of the reasons why I'll be happy when A2A get something into MSFS, because their Comanche 250 is really the only flight sim add-on which does a pretty convincing sideslip as far as I'm concerned. If how it flies in P3D can be shoehorned into MSFS, then they can't get that thing into MSFS soon enough for me.

Edited January 20, 20215 yr by Chock

11 minutes ago, Chock said:

All these graphs are interesting to check out, but in spite of this, I'm somewhat inclined to go off what it feels like and looks like with this stuff, and to me MSFS seems fairly convincing for normal regime flights. Not perfect, but on occasion it does make me recall moments from when I've been flying the real stuff.

In fact I did a flight earlier on in the Virtualcol Beech 99 and when it lifted off the runway and began to pick up some drift in the crosswind, I actually noticed that it was a lot like the kind of thing I've experienced many times when flying the real aeroplanes at that point when it came to the inputs necessary to keep the thing going the way I wanted it to. This is of course pretty subjective, but it certainly was a remarkable replication of my own experience of flying stuff on that occasion this evening, enough for me to actually take note of it.

There is no question in my mind that at times it feels more dynamic than the other sims, especially in transitions (taking off and landing) or going between buildings or mountains, but the issues seem to be apparent in the basic equations, like glide ratios and ground effect and effect the flaps have on lift, stuff that is relatively simple to model is why it concerns me. I believe this stuff is relatively accurate in the Xplane models, though I haven't verified it.

I agree that none of these sims have ever lived up to the hype of true accuracy, though the Majestic Q400 in P3D and FSX did have an amazingly unique feeling on landing, how accurate that was to real life is hard to say since I've never flown in one. I haven't seen this same level or felt this same type of thing in any MSFS plane. The physics feel very forced in MSFS to me, rather than actually responsive based on the speed of the plane or the drag/lift ratio. Who knows...

I think the larger planes can work fine even on landing with just a few adjustments, even some people have hacked up the settings to make the landing less "floaty", but the real question I have is what seems to be a total lack of stability for the smaller planes. I know smaller planes bounce around a bit in real life, but the control response seem absurdly far off from real life, and some of the angles I can 180 a plane seem like they should have resulted in a crash.

Edited January 20, 20215 yr by Alpine Scenery

That thread sort of reminded me of my younger days playing around in the engineering lab with differential equations for dampened systems and emulating them up on the old school analogue computers with op amps and resistors and capacitors etc  😄

It is clear now why Asobo introduced the delay option to the sensitivity curves and equally clear why it is just a stop gap that will never actually solve the real issue.