Absolutely agree! I’m not sure why some here are resistant at one of the best in the industry (A2A) sharing their knowledge!
3 Likes
I’m fairely sure they don’t use any of the sim’s CFD and all that stuff but have dragged the whole flight model calculation outside of the sim into their Accusim service. So they actually didn’t manage it somehow to do what others couldn’t do with the MSFS model but did it completely on their own, from scratch. That’s the difference. They did the same in P3D.
4 Likes
They don’t need to support it, it just needs to work. 
They went ahead, and did it anyway somehow. The first of many, I hope.
2 Likes
MS won’t learn much from someone who tailor made an FM for one plane that is injected into the sim then SLEWS the plane. MS is in the process of creating a native FM that will accommodate all sorts of aircraft based on physics.
I totally love the Comanche but I think the A2A way is not the right way for MSFS.
We already have a flight model in MSFS with CFD so ideally developers should use that flight model and have a shorter time to market and enjoy compatibility with Xbox, future updates, etc.
Some devs find that FM too limited so they go on their own path to circunvent the limitations to the point of only using MSFS for its scenery.
I think, and this is only my opinion, that Asobo and MS should really strive to make the FM as complete as possible to the point where no dev would even think about creating its own physics or aerodynamics, why would they want to do that if the sim has them right?
Having all these “outisders” like Fenix and A2A and some others is in my view not a good thing for the platform as a whole but today the outsiders are giving the higher fidelity in terms of simming and I think there is a lot to ponder about this.
5 Likes
Seb stated at FSExpo they were working on making the flight model more realistic. None of us have even seen how far they will progress yet. The new simulator is still a year away. But already here the forum is predicting that they will fail and imploring them to bring in help.
Will the new flight model rival A2As? I don’t know. They did say though that is will be more advanced. So it will be somewhere between what we have now, up to perhpas equatable to A2As. We don’t know yet. Lets wait and see before we start signing petitions to ask for help.
2 Likes
On the surface this OP seems logical enough, but it fails to consider a far bigger elephant in the room than custom flight modelling: Systems integration.
Copy/pasta from the other thread:
Also:
On the FM side Microsobo are doing fine by themselves, they’ve spoken many times about the strengths and the weaknesses of their approach, and they have been open enough about how they are working to evolve it.
A2A are also not the only ones doing custom external flight models. I’m sure if Microsobo need to they can reach out to any of them anytime, and in fact in view of the points raised above (not just by me it should be noted) they should be talking to all of them to ensure developers doing these sorts of integrations are conforming to standard sim interfaces to synch information to a common source of flight truth that other addons/functionality can rely on to be consistent and accurate.
Ditto for devs using base sim functionality, which is already an issue with some information not being synched. This problem has potential to get much worse, causing confusion and degrading the overall MSFS UX by not providing a consistent user experience across planes and addons.
Standards are important, so all users can have some common base level of expectation about performance, not just with the plane but how it integrates into the wider platform and ecosystem of addons/functionality.
4 Likes
So Stearmandriver is a real life 737 pilot and he also flies GA planes on his free time (as his name implies, he typically likes to fly the Stearman). He may also be a CFI but on this, I’m not as sure. Stearmandriver streams on Twitch and he primarily streams MSFS and sometimes, some DCS World. I have not seen Stearmandriver stream P3D or XP before. In addition, Stearmandriver often defends MSFS from some MSFS haters in the Avsim forum. Despite the fact that Stearmandriver is very pro-MSFS, this is what he had to say about MSFS’s flight model and how it compares to the A2A Comanche:
I’ll weigh in on this, after several flights in the Comanche (and as someone who greatly enjoys the 310).
First, the flight model. A2A has, basically, created a new simulator here. Sure, much of the flight regime in MSFS feels quite good - better than we’ve had before in a sim - but there are problem areas. Take crosswind effects, for example: in all other MSFS aircraft, a 10 knot crosswind component feels like about 25kts. You can land in a sideslip with the longitudinal axis aligned with centerline, upwind wing low to prevent drift etc… You can do this perfectly, and on touchdown weird things still happen in every plane I’ve previously used in MSFS. The plane lurches or slides sideways, the nose yaws around, there’s an extreme tendency to not only weathervane but drift INTO the crosswind etc.
The Comanche does none of this. 10kts of crosswind feels like 10kts, and if you touch down straight with no drift, nothing weird happens. You simply touch down on upwind main, downwind main, and nose gear, rolling straight all the while, and weathervaning is easily controllable with aileron into the wind as it should be.
It’s like you’re NOT flying MSFS; the quirks aren’t there. In THIS simulator, the plane behaves like it should!
Of course every other aspect of the flight model is great too. Stall / spin physics are phenomenal, and the fact that they spent time on something that the airplane isn’t even certified to do speaks volumes about attention to detail.
…
…Those are just a couple examples but the whole plane is like that. Oh, the instruments! Hand flying instrument approaches is an absolute pleasure; the needles are 100% smooth, there’s no ticking at ALL. That seems like it should be simpler, but other devs seem to struggle with it…
Ultimately, I’m coming to the conclusion that this plane, in this sim, represents nothing less than a new level of GA sim experience. Given that no true high-fidelity GA simulation devices exist (there are no level C or D Comanche boxes out there), I actually think the claim can be made that this is the most accurate, natural simulation of a general aviation aircraft that has ever been created. On any platform, not just limited to a desktop sim.
I get that this may sound fan-boyish but I’m objectively serious. It’s an entirely new level.
And for “casual” simmers? I mean we’re all here for fun so I’m not sure what the difference might be, but I think this plane is for everyone. I mean, it’s not hard to learn to fly a Comanche, and everyone will enjoy the flight dynamics from the get go. Interfacing with aircraft options is extremely simple via their tablet. There’s nothing to be intimidated about.
I mean, it’s really that good. I don’t understand what they’ve done. They should just build their own sim.
2 Likes
And to that I’d say, so what? (apart from, great, good job A2a).
The ability to do this is by the design of Microsobo. They allow that to be done. Think about that for a second. What does that imply? It implies Microsobo is OK with people developing alternative flight models++ to plug into MSFS.
If A2A hadn’t produced a great Accusim V2 flight model based on 3 years R&D built on the prior ~10 years of Accusim V1 I’d be wondering what was wrong with them.
None of that is the real issue.
The bigger issue is integration of aircraft using this method into the MSFS content ecosystem consistent with standards that don’t break things randomly left, right, and center, as pointed out directly above the post above: Please analyze the A2A Comanche for its flight model, and how it can help MSFS 2024's flight model - #27 by Sonicviz
Otherwise they may as well go and develop their own sim. Not just for A2A, but anyone who wants to do this.
Microsobo is redeveloping their own flight system, so who’s to say at this point how good/bad it will be be? You can’t, so best thing for now is chill out about it and see how it develops. Release of MSFS 2024 SDK should be illuminating here.
Whether it sucks or whether it’s a 10x or 20x improvement on the current internal FM SDK is also irrelevant to a degree, because you can still create your own if you wish, as A2A and others no doubt will. That’s actually pretty cool in itself. The skill level to do that is quite high, but if the new SDK is as flexible as they say it should allow for some interesting creative possibilities.
However, this integration issue is also a problem already with some other aircraft using the internal SDK and hybrid approaches.
This is not a good trend, and needs a spotlight on it before it really gets out of control.
There needs to be a single source of data truth that all aircraft update so addons and core functionality that rely on such data don’t break and the user has a seamless MSFS experience, whether a plane uses the internal SDK, an external custom system, or a hybrid approach.
If not, welcome to the balkanization of the MSFS 2020/2024 aircraft ecosystem.
Otoh, maybe that’s the plan, I don’t know. It would be sad to see that happen though: technically, content wise, and user experience wise. A case of a missed oppportunity to really level up flight simming.
1 Like
That was a good read, and pretty much mirrors my own thoughts.
My first flight, from taxi, takeoff roll, to landing again, the whole time I was thinking “This doesn’t even feel like MSFS!” They aren’t exaggerating, it feels like a new sim.
It’s quite unlike anything else, and is a real joy to fly. Approaches seem far more predictable, not as greatly affected by mechanical turbulence in a few places I went to known to throw me about like crazy on final.
1 Like
Hi,
Simply put, to me the Comanche feels like it should. A real airplane simulation.
I hope that Asobo will take a very serious look at it.
4 Likes
I’m not that impressed with their flight model. It did not have the trim characteristics that I look for. I have watched the flight models in the sim improve steadily, and the ones I appreciate most did not need a separate process to run.
I am not sure what is external in the module, but if they have created an external module that uses the MSFS terrain system, etc. how is that different from creating a module that uses Google maps instead of Bing maps?
I have heard it said that the module takes control of the sound system, and I could hear something going on with the sound as I exited my flight. If there are issues with sound, does MSFS have to first ask, “Are you running an A2A module?”
Finally, after reading the history of A2A and Accu-sim, I have to wonder why now? Why did it take 3 years to get a module ready for MSFS? They had a lot of success injecting their tech into sims that were end of life, and 2020 is about to go into maintenance mode. I have to wonder.
I have uninstalled their module, and I am not interested in their products anymore.
Ms and and asobo, if you cant beat them, join them.
We want accusim so bad in msfs 202x😀
1 Like
I agree very much with the points you have made.
As I think you say, the ball is very much in Asobo’s part of the court to get the flight model and such up to A2A’s standard.
Whether they have the ability or even the inclination to try to meet this high standard though is an entirely different question. It seems to me that since MSFS launched around 3 years ago that the vast majority of updates have been scenery and ‘fun gaming’ based. Maybe they will change tack a bit now they have taken on the IL 2 flight model guy but I am yet to remain convinced that the main focus will be to get the flight model and such up to A2A standards any time soon 
In the meantime, I am very happy to fly an authentic plane in the Accusim environment and enjoy the excellent scenery which MSFS provides
Scenery updates are done by the business partners Orbx, Gaya, and Perfect Flight. The “fun” updates are done by other business partners.
Sim Updates are what Asobo works on. They work on what is determined by the community what is the most important to work on. Sometimes though things aren’t strictly modeling based. Last year, one of those updates was support for DLSS. It took 9 months to get that update out of beta and released. While focusing on that, they didn’t get to work on flight models or season or any of the other things that people want. But the community as a whole wanted Asobo to focus on getting more FPS. And then once that was done, people are now able to fly a bit smoother now, and then they forget about that release, and they start complaining that the sim elements very little work has been done. Of course. Their focus was on DLSS to get people FPS like they requested.
Highly agree with all this. The quirks of MSFS that A2A have eliminated don’t make airplanes easier to fly in the sim… They make them harder.
2 Likes
Thanks for replying in this thread @Stearmandriver2! I hope you don’t mind that I used your comment as an example of the differences between the MSFS and A2A flight model, since I highly respect your real life experience not only with air liners, but also your real life experience with small GA planes.
This is all the more reason that Asobo work to make the flight model better for small GA planes, and hopefully closer to the flight model that A2A has produced.
1 Like
NO, they were “probably” moving across to be working on 2024.
2 Likes
Asobo started hiring in September, and they also hired Andrey in September. That seems to indicate that work on FS2024 didn’t happen until they wrapped up development on the 40th Anniversary Edition.
Yes, I think I am more inclined to agree with you on this one 