Since you're responding to @MattNischan's quotes above, he can respond best if he has the time.

I fully subnscribe @blingthinger notes above.

The ASOBO documentation mentions a worth of totally wrong assumptions about X-Plane's fdm that only reveal how short from really understanding it they really are, or maybe more worried with promoting their approach as opposed to X-Plane's... 

This being said, I have been defending the positive aspects of ASOBO's work in MFS, and they really didn't have to use such a bad marketing-oriented approach....

I love MFS, but when it comes to FDM, I  still find X-Plane's approach and concretizations way ahead...

18 minutes ago, jcomm said:

The ASOBO documentation mentions a worth of totally wrong assumptions about X-Plane's fdm that only reveal how short from really understanding it they really are, or maybe more worried with promoting their approach as opposed to X-Plane's... 

Can you point out where in the Asobo docs they currently talk about XP's FDE? Or are you talking about some initial mentions there in the early days after release that were since removed? (I never got to see that).

1 hour ago, jcomm said:

I love MFS, but when it comes to FDM, I  still find X-Plane's approach and concretizations way ahead...

Without a mature set of aircrafts fully taking advantage of MSFS's FDE, that statement remains to be seen :) We have certainly seen some initial high fidelity implementations that start to show what MSFS's FDE is capable of, and I'm looking forward to what iniBuilds can showcase with their A310 and A2A with their Commanche. Always good to hear from aircraft developers who have developed for multiple sims to speak to the FDE differences and positives/negatives of each. So when a team experienced on the XP FDE like iniBuilds go out of their way to say: https://discord.com/channels/535246634448191499/784467650612822016/972850172067459103 "Just to be fully clear, our flight model in MSFS is on par to XP. So those still wanting to negate MSFS as a viable sim solution based on those reasons - don’t speak too soon", I'm gonna pay close attention and note that down as interesting insight/feedback :)

Edited August 21, 20223 yr by lwt1971

57 minutes ago, lwt1971 said:

Can you point out where in the Asobo docs they currently talk about XP's FDE? Or are you talking about some initial mentions there in the early days after release that were since removed? (I never got to see that).

"Compared to other flight simulators out there, our geometrical discretisation is more precise - our model uses 640 surfaces where other simulators have only ten elements per side per wing or stabilizer."

"This means, for instance, that additional drag and pitch due to gears for  cannot be accounted for in their model, while it can be in ours."

https://docs.flightsimulator.com/html/Samples_And_Tutorials/Primers/Flight_Model_Physics.htm

I never saw the explicit LR mentions either and I'm wondering why they were removed. Did LR send lawyers after them? And even still, it's obvious and somewhat slanderous. Extending landing gear in XP most certainly does change the drag characteristics and their claim of spatial resolution is not only incorrect but is comparing apples to oranges. In terms of overall market cap, this is like the 18 yr old throwing punches in the elementary school. Maybe XP12 will change that some?

 

1 hour ago, lwt1971 said:

say: https://discord.com/channels/535246634448191499/784467650612822016/972850172067459103 "Just to be fully clear, our flight model in MSFS is on par to XP. So those still wanting to negate MSFS as a viable sim solution based on those reasons - don’t speak too soon", I'm gonna pay close attention and note that down as interesting insight/feedback :)

100% agree with you and I wish devs like these (who have spent blood sweat and tears in both models) would expound on the hoops they've had to jump through to get each model working. Unfortunately, talking about the fudge factors would reveal their trade secrets, so I totally understand the silence.

Airliners and autopilots. If you stay in the safe, stable, center region of these performance curves, you'll be fine. That FSX/P3D can hit numbers, is proof of that. Hitting numbers outside of that is going to take lots of tweaking. Given how much closer XP can get to the actual 3D geometry, there's likely more tweaking (unrealistic fudge factors) needed in FS2020 than in XP. Will it get there? In many cases, yes, but at what cost to the mental health of the developer! :)

1 hour ago, lwt1971 said:

Can you point out where in the Asobo docs they currently talk about XP's FDE? Or are you talking about some initial mentions there in the early days after release that were since removed? (I never got to see that).

Woulkd have to check  again, and I will report back if I still find those mentions...

There are aspects of the aircraft geommetry which are simply absent from MFS's approach. Tails and wings with multiple surfaces are an example.

Of course there are "ways around" and that's probably what developers explore when they fine-tune their models.

For instance the recently released - which I bought day one and am really enjoying! - V36 B is a good example of an aircraft whose specific tail geometry cannot be represented in MFS, and XP allows to properly model, at least from a geometric pov.

When it comes to propeller dynamics, and while MFS is still working on their modelrn FM and further refining it, the way we can directly translate rw data from a prop manufacturer specification into XP vs MFS is a big difference, and in MFS the moder prop model is still a headache for developers seeking as close as possible to realistic performance.

This doesn't mean I don't like MFS's FM and approach -. quite on the contrary !

Edited August 21, 20223 yr by jcomm

24 minutes ago, jcomm said:

When it comes to propeller dynamics, and while MFS is still working on their modelrn FM and further refining it

And now that they've broken it further with the CFD, they have a while yet to go. At least they're up front about the experimental nature of this one!

"For example, with the CFD simulation enabled, the airflow generated by the aircraft propeller will hit the fuselage and change the propeller effects such as the p-factor."

https://docs.flightsimulator.com/flighting/html/Developer_Mode/Aircraft_Editor/Debug/Debug_Aircraft_CFD.htm

The CFD is a novel idea in a model like this and might be a decent way of adding dynamic responses to the FSX theme, but I think it will be difficult to get realism with the spatial resolution they're using (20x20x20) as well as complications with the orientation of the airframe elements. In the propeller scenario, they now have prop wash impacting the fuselage. In the model, the fuselage is treated as a flat plate with only side forces: the lift force conservation equation in the SDK has no fuselage component (F14??) and the drag equation isn't even shown (show us the fudge factors!). IRL, the air is swirling around a "tube" shape, not just slamming into the side of a flat plate, as is happening here.

As for spatial resolution, the force elements are over-refined, but the CFD is under-refined. And the latter always wants more cells! They call them "voxels", but that's a computer graphics term, not aerospace. What's cool is that the CFD solution runs at 100 fps! It's no wonder they caution against increasing the cell count, but the hoop-jumping associated with that coarse grid is probably a pain.

Some effects on propwash to fuselage interaction like Coanda effect are also missing. 

Their visualization of the propwash stream under x-wind shows a detachment of the flow from the fuselage immediately as it leaves the prop disk area.

In the demonstrations that they made sometime ago showing a b200 stopped prop that was feathered and kept still as the airctaft was made to roll along it's corrected for sideslip longitudinal axis also contrasted with a total absence of movement of the prop of the stopped engine if the aircraft starts flying in a sideslip...

Still a long way to go, but lots of potential and a promising / interesting approach.

Edited August 21, 20223 yr by jcomm

1 hour ago, jcomm said:

in MFS the moder prop model is still a headache for developers seeking as close as possible to realistic performance.

Is it? According to that list on the MSFS forums, some of the better planes (free- and payware) use it. Who said that?

Edited August 21, 20223 yr by rka
list

12 minutes ago, rka said:

Is it? According to that list on the MSFS forums, some of the better planes (free- and payware) use it. Who said that?

Can't quote from memory, but if you poll the dev forums you'll find reports of how strange incidence values have to be used in order to get closer to realistic outputs...

Edited August 21, 20223 yr by jcomm