Welcome to
Microsoft Flight Simulator’s SDK Q&A Platform!
Please take a moment to read the platform’s guidelines before you get started!
Having done some high level study of the volocopter's game files, I can see that a WASM flight model completely replaces the in game flight model. How has this been done without breaking the in game aircraft editor. There is no material on this in the sdk
Hi Iceman,
That's interesting. How could you see a WASM file was used for the flight model?
I tried to study the game files like you did, and I only found .fsarchive files that are encrypted as far as I know.
Anyway, you are certainly right, I would just like to know more about this aircraft file structure.
Thanks.
I was curious about how it works and checked files on the beta stage. Some of my conclusions (none of them is answer, which I would like to hear myself):
1. aircraft folder calls "microsoft-aircraft-volocity". is it really made by MS developers, not ASOBO? you can expect different approach of the development from another team, especially if this developer worked on FSX years ago
2. currently it is possible to alter native flight model by pushing values of VELOCITY BODY, ROTATION VELOCITY BODY and ROTATION ACCELERATION BODY sim variables every frame (in fact every of 18 frames per second which makes simconnect not a best choice for flight model container). but still it will fight with native FM which became noticeable, usually, at airspeed higher than 100kn. I've even made request to add a way to suppress native FM but it left without attention https://devsupport.flightsimulator.com/idea/2053/suppress-aerodynamics-calculations.html
3. flightmodel.cfg is encrypted. for what reason is quite unclear. It is possible to read some variables by SimConnect but the list is really limited.
4. I was surprised that devs used the very same way as 3rd party developers used to create artificial lift for first helicopters. What was the problem with increasing the maximum number of engines and controlling them individually? That is no naked theory - for experiment purposes I have made NEXUS eVTOL fly vertically with 4 engines pitched 90 degrees up. Controls linked with engines thrust - it is not very stable as I spent maybe 4 hours on it, but it flies twice more realistic.
5. One more time, can we have an editable ACCELERATION BODY, please?
Increasing the max number of engines would be necessary to properly model fixed-wing aircraft with more than four engines, too. I can understand that back in FS98 days four might have been a suitable restriction and hard to extend within the silly resource limits back then, but couldn't it at least be doubled by now, if not more? (B-52, XB-70, Do X are obvious examples of aircraft that would need more for poper modelling.) And for some rare but historically interesting cases it would even be useful to have multiple kinds of engines (B-36D at least).
It is very interesting that the developers have now legitimized the simconnect way to create exotic types, clearly acknowledging the many limitations of their flight model. It is also very annoying that they haven't given any explanation of how to detach the in game flight model from the aircraft as all attempts I have made in doing this have broken the game. I would guess this will come later as an interface for a WASM module for making helicopters later on. If they could make this dynamic this would allow the creation of other types much easier. Either way, some clarity on this approach would be greatly appreciated, and it's nice to know I haven't wasted my time learning simconnect over the past 9 months
At the very least I hope the team includes some info in the SDK on using WASM for flight models. Although I totally disagree with using any external modules for the FM. I hope longer term the team addresses this.
It's the high speed "fight" that concerns us... that loss of control is a real problem.
@FlyingRaccoon any chance of an Asobo dev responding to this?
I've noticed that Asobo won't respond in any issue that they do not intend on "fixing"... is that the case?
Hi there,
"I've noticed that Asobo won't respond in any issue that they do not intend on "fixing"... is that the case?" - just in case this was a serious question, let me answer "no".
As for custom flight models, there is still no intention at this moment to allow third-party developers to create them, as has been officially stated in the past. Of course this may change at some point in the future.
I understand that some people have been using SimConnect as a workaround - while this can look like a viable option right now, please note that it might not be future-proof.
Best regards,
Eric / Asobo
Quite honestly, that's the most disappointing news I have heard since the release of this sim. I have never been an advocate of external flight models, always preferring to use the sim's internal workings (FSX, P3D and X-Plane) and I had hoped that would be the case here. However, having developed several flight models already for this sim which, despite my efforts, fail to demonstrate key aerodynamic features of the aircraft through accurate geometry and an almost total lack of ability to build behaviours around that, I think at this point that an external flight model is the only way forwards.
Aircraft I am currently working on have to have SimConnect and coding because the handling would instantly be noticeable as wrong, to the extent where the aircraft could even be described as unflyable. I have also pointed out that several upcoming projects (some which work has started on) will be fairly pointless unless we are able to circumvent the current situation. It's heading towards a state where the only viable aircraft will be regular airliners and GA aircraft, and even they will be more of a 'generic' flight model.
If the SDK for the flight model can become detailed enough that it actually makes sense, then the current core flight model will be ideal for bringing people into the sim. However, for those of us who actually want to create realism, deep-level flight modelling which many simmers really want, PLEASE give us back control. Keep the core as a core, and give us back the coefficients and tables we used to be able to build with to build around the core, rather than replacing it. Think of that as a sort of 'official external FM', if you will.
Best wishes,
Paul.
Please, reconsider this.
@EPellissier , "I understand that some people have been using SimConnect as a workaround - while this can look like a viable option right now, please note that it might not be future-proof." - Without wishing to sound impertinent, MS/Asobo have legitized this approach already as more than just a work around with the volocity, we know this aircraft is using a simconnect derived flight model and it is an officially supported aircraft.
The game with the possibility of default flight models/flight model configurations turned off would create such a sandbox of possibilities. I simply don't understand why the "flight" in flight sim only has to refer to fixed wing aircraft, in the current framework, helicopters, autogyros, tilt rotors, stovls, aerostats, blimps, rockets etc are unsupported. Given it is clear this has already been done, why can't the methods used be made available to developers as it is evident that it is not a game breaker as was my initial thought before I saw the volocity.
The alternative, something I think should be in the development pathway regardless of potential/demonstratable workarounds, is official support for multiple flight model types natively within the game architecture, something that was never truly available in MS ESP, which I think this game has too much carryover from.
Best regards,
Eric / Asobo
Hi Eric,
The Volocity is using SimConnect through a WASM module - my comment about not being future-proof was referring to "out-of-process" (external EXEs) flight models that use SimConnect.
Best regards,
Eric / Asobo
I don't understand this statement and I believe I'm not alone in wondering what this is referring to. In effect, regardless of being solicitated from an external app, or an in-game WASM gauge, SimConnect API doesn't change isn't it?
The way I read your statement and I believe I must be 100% wrong is that in the future, external apps using SimConnect might no longer be able to do what they can do now (i.e. writing a few simvars) whereas this functionality won't be changed for in-game WASM gauges?!?
Would you mind clarifying what this could be in the future, or can you give an hypothetical example of what you're thinking about?
@EPellissier, I have to ask... did you turn "it" off for the Volocopter? Because, we do know for sure that this product is using an external flight profile.
Our concern is that it would appear that Asobo is locking us out from methodologies used by your team....
Note, I'm just asking here... not trying to start a war...
We, quite literally, do not understand why you won't allow this. It makes the simmers happy, makes all of us money and keeps things new, innovative and interesting...
Interesting discussion... I have no knowledge in this field (Scenery designer only) but I tweaked Air files back in the late 90s for FS98 and FS2000...
I was also wondering if the initial architecture of MSFS with Legacy vs Modern flight model (we toggle in the GUI), were a prelude to more native flight models or flight models developed by 3rd party based on SDK...
Now that the other simulator is showcasing its Alpha Videos, of course, the big hype there is the flight model yet, the immersion as far as the videos show is 15 years old...
So I hope MSFS stakeholders will establish a strategy to make the flight model open (or modular) and with the actual top notch immersive world in the sim, it would put the competition way behind I believe.
Also supporting opening up the flight model to customization on a per-plane basis - there are simply so many different possibilities with aircraft configurations that having a locked-down flight model implementation is lacking. It's one of the few areas which developers cannot work around in the sim, yet a good flight model is probably one of the most defining aspects of a solid third party aircraft. I do hope this is taken into serious consideration.
Trying to be constructive about this topic again, it is evident from the replies here that 3rd party devs are still struggling with the core flight model, a year and a half after release. From background discussions with other development groups I have not worked with, the comments are all along the same lines - inability to actually control directly what is happening (over-reliance on the core FM to get it right when frequently it doesn't) and a lack of clarity in the SDK.
Looking at the various forums and social media posts over the last few weeks, the SU31 team have posted that the core flight model is not compatible with high-end aerobatic flight. Ant's Airplanes noted the inability to directly create stall / spin behaviour and that accurate dimensions have had to be changed to inaccurate ones to try and get something like the correct behaviour. I have real-world hours in the Tiger Moth, and his P3D version was one of the best I have seen, only lacking interaction with the environment (hence my comments upthread about the core as a core and the FSX FM on top).
There has been a mass of interaction with developers about so many aspects of the sim, but the flight model just seems to be an area where there is a mass of dissatisfaction on our side and relative silence on the other side. We all want the same thing in the end - the ability to create high end, detailed aircraft to make the best use of the incredible architecture that has been built for us but we still cannot build basic aerodynamic behaviours which developers in the other sims can.
@GrimPhoenix9349 agreed, the FM and rigidity of the current framework is the primary blocker to pretty much all of the aircraft coming out being realistic, it's alright making a beautiful aircraft (the JU52 I purchased recently is a standout for instance) but if it doesn't fly right then what is the point
The point I want to drive home here is that all we are asking for is the ability to start from scratch with regard to flight dynamics, the community will be able to create so much variety on its own, even if the developers aren't interested. This game is worth doing this for, it really is worth it, and I think it is such a shame it is held up by something that conceivably would be a simple fix. That isn't even mentioning the possibilities beyond flight in such a rich world that I don't think have even been considered, though that may be wishful thinking
I know with my own project, which is a simconnect fudge fighting with the in-game FM, when it is smooth it is a far cry from anything I could get with the in-built systems. Of course the alternative to a method of taking off all of the locks in the game would be to expand the flight model system natively to alternative aircraft/vehicle types.
I have a feeling that all this at the end of the day comes down to PC vs Xbox stuff. I understand that if the total unification of these platforms will continue, SimConnect could be ditched altogether in future, just to satisfy the Xbox crowd who otherwise would never get add-ons of PC quality due to closed architecture. I so much hope that won’t happen and MS some day decides that Xbox is Xbox, PC is PC. Unification of these won’t fly if we are talking in terms of a great simulator, rather than an average flying game. On the other hand, if Asobo manages to make the default flight model completely self-sufficient, then I’d be fine with that.
15 People are following this question.
