It seems like an intelligent compromise is still the best approach.  Working outside of FS to get around the limitations is a good thing, but doing too much on the outside breaks compatibility with other things, as Tabs said.  I like the PMDG, and maybe A2A, approach better, and I'm starting to lean towards giving the Q400 a pass for this very reason.  I hope I don't have to, but I am hearing some very mixed messages about incompatibilities with a few programs I'm not sure I can do without.

I'm guessing it will become more common as FSX continues to age and everyone keeps bumping into its limits. FSUIPC already injects code into the sim to fix bugs and enhance functionality. Many are using shaders that insert themselves into the FSX rendering pipeline to modify the graphics. Various weather programs replace parts of the FSX weather system. Some programs replace FSX ATC. It's only natural that some programs also replace the flight dynamics.

One advantage of external programs is that they should be able to better take advantage of multiple CPU cores. With the physics, flight dynamics, weather and ATC running in their own process rather than as part of the main FSX process, the Windows task scheduler can move them to another core.

The danger is that the sim just becomes a cobbled-together mess of various third-party hacks. How will the external flight dynamics interact with the external weather engine, and will the external ATC interpret the weather correctly?

Forward-compatibility is another issue. FS itself is obviously dead, but Prepar3D might become more accessible to the hobbyist market in the future. It will probably remain compatible with add-ons that conform to the FSX SDK's, but might not work with third party external "hacks" that rely on injecting code or modifying specific memory locations within the FSX process.

FSX is all about addons and those addons have to play together. External FDE, at this time, makes a lot of other addons unusable and this is unacceptable for me. They have to find a solution for that or this external FDE business won't fly far. I agree, more realism is better and FDE is an important part of the whole FS thing but... if people bought 100's or even 1000's of $$$ worth of addons which they use on every flight I don't think they would exchange all that just for a better FDE. Well, we'll see..

I think the main reason a completely external flight model hasnt been embraced by the mainstream is because only the top few actually have the skill/knowledge necessary to do so. Making a external flight model is not a simple task. As for wind/turbulence it IS possible to create a EFM that takes these into account. It is something i am working on myself. But again, its still not a simple task in any way.

 

And the negatives (like only locked spot view) can be overcome. It is just dependent on the skill of the programmer.

 

FWIW there is another external flight model in FSX that has been around for some time. Helicopter Total Realism by Fred Naar.

I used an external FDE, and a very neat application + fsuipc to fly the best 747-400 simulation ever produced for a PC, as far back as 2006. I'm talking about Aerowinx PS1 :-)

 

Used it with fs9 and latter fsx. The sims were there only for the purpose of rendering scenery and weather...

 

ELITE also uses ESP in some of their FNPTs.

It seems like an intelligent compromise is still the best approach.  Working outside of FS to get around the limitations is a good thing, but doing too much on the outside breaks compatibility with other things, as Tabs said.  I like the PMDG, and maybe A2A, approach better, and I'm starting to lean towards giving the Q400 a pass for this very reason.  I hope I don't have to, but I am hearing some very mixed messages about incompatibilities with a few programs I'm not sure I can do without.

The "intelligent compromise" works for jets and piston props, but with turboprops FSX's modelling of them is so fundamentally broken I don't think that compromise can deliver acceptable results. I don't think using the FSX FDE was ever really an option when modelling the Q400 at this level.

 

Flight1 tried to get a half-decent result using the FS9 FDE for the ATR 72, but it's not capable of accurately simulating beta range and prop pitch drag effects like the Majestic Q400 can.

 

PMDG managed to get the compromise working reasonably for the Jetstream 41 but that's a different kettle of fish because the TPE-331 is a geared, direct-drive engine, which is very much unlike the PW150 free power turbine engines fitted on the Q400.

If a plane gets externalized to the point that you can fly it through a violent thunderstorm that would otherwise swat if from the sky, but in the simulation it flies like it is on rails, is that acceptable?  You can have the best theoretical FDE in the world, but if it is too detached from the simulated environment, what's the point?  Developers are simulating an airplane, but I think many flight simmers (customers) are also trying to simulate the larger experience of the captain on the flight deck in that particular airplane, on top of that.  I would hope that developers don't lose sight of that, however they decide to go about it.

Totally agree. But since P3D is a hopeful but a big IF because what the agenda will be on licensing, XPX still quite a ways away from being complete like FSX, and no announcement of any other platform to be the next big thing other than the future is bright, I think this might have to be a forced issue. 

IF one was to dream, the AI model and scenery generation takes up a good amount of the precious 4gb limit we are up against not to mention steals FPS for flying, and if these were external somehow, then the life would be extended yet again while keeping the FDE internal with the enviroment. 

Im no programmer but there must be a way to call up textures according to the scenery externally and loading them in memory outside fsx. It all needs a mediator client to talk to everything. 

 

Before someone says no that is impossible, just remember what was impossible when FSX was released and where we are today. 

 

 

FWIW there is another external flight model in FSX that has been around for some time. Helicopter Total Realism by Fred Naar.

 

I think the Dodosim 206 also has an external FDE?

 

I think the numerous examples have shown that devs have been doing it in some form or another for quite some time.  Nothing to the extent of MJC with the Q400 of course.

 

When you think about it, perhaps the future is going to simply relegate FSX to role of providing the scenery; everything else could be done outside of it.  I think the trick in providing a complete environment to the user (turbulance, atc, AI traffic, etc) would be better served in establishing module to module communication.  Wouldn't it be great if the effects from the weather were derived purely from say, the Q400 talking directly to Opus...

 

 

If a plane gets externalized to the point that you can fly it through a violent thunderstorm that would otherwise swat if from the sky, but in the simulation it flies like it is on rails, is that acceptable?

 

If you've visited the Majestic forum today you would see a comment from "the boss" indicating something addressing turbulence is being worked on.  Looking forward to that since I miss the DHM in Opus.  My question would be can they keep the frame rates high?  http://majesticsoftware.com/mjc8q400/support.html

If you've visited the Majestic forum today you would see a comment from "the boss" indicating something addressing turbulence is being worked on.  Looking forward to that since I miss the DHM in Opus.  My question would be can they keep the frame rates high?  http://majesticsoftware.com/mjc8q400/support.html

 

Anything is possible when multi threading is used.....

I think its explained pretty much many times, that FSX is able to handle FDE very realistically. I really don't see how it really improve flight that much (aside for turboprops), so dont expect it to go mainstream.

Other than the turbine engine parameters that can be really screwed, FSX is able to simulate realistic to 95% on the book numbers.

IMO, The reason why you might see that MJC Q400 is smoother, is that the External FDE (JSBSim is the External FDE Majestic uses) controls (note not calculated) the aircraft performance via a frame by frame basis. So that means even when your FPS drop very low, it will handle it very smoothly. On the other hand, FSX FDE is calculated on a frame by frame basis, so any low FPS can cause weird FDE movements, or unsmooth movements. Thats why people with constant FPS get a much more enjoyable ride than those people who have 50+ fps but variable. 

I could be wrong on what I said above as I am not at expert at it, but this are my findings and conclusion. Just my 2 cents.

There are many things, specially regarding Yaw in Single Prop airplanes, that I think would be much better with an external FDE than using FS standard code. Side Slips are something that takes many compromises in other areas of the FDE to make realistic in FSX, as told my the FDE Guru, Rob Young. And as you said, just the smoothness of the external FDE simulation is worth alone the upgrade IMO, it feels so realistic, probably because it's running at 120fps instead of the usual FSX's 30fps.

 

That's my opinion, of course many aren't going to agree with me.

As long as they can get this turbulence/winds thing fixed and ensure compatibility with 3rd party like AES, GSX or FSPAX, I see no problem in using an external FDE if it brings superior aircraft handling and enhances the overall realism with better "fly-by-numbers".

 

That requierement of stable FPS is also present in X-Plane, which reduces the sim speed to render the aerodynamics correctly, if you are having low FPS.

 

For high quality add-ons, "external" systems and FDE with FSX just as graphics platform seem to be the future to me.

While the evidence supports that it benefits turboprops only to be external, I would hope that finding a solution to making it work for jets continues on for the benefit of such an increase in FPS. We have optimized machines to run it, yet FSX while seemingly runs fairly well, is still bumping up on so many limits thus making it very volatile not just in a CTD, but the flying experience overall or lack thereof. 

 

Of course all this would be conjecture if we really knew what the next best big sim will be to overtake, thus spending time and resources in it would be futile.