Requirements for realistic flight simulation!

Fellow flight simmers,

The wish-lists, bug reports and Dev Q&A questions are overpopulated with requests for improved visuals, better night lighting, new world series. In short, more eye-candy. Although graphics are important, this simulator already looks 100x better than anything else on the market while lacking in other (in my opinion more important) areas.

It seems to me that the broad audience for MSFS has a downside, some people are interested in a realistic flight simulator while most are in it for the stunning visuals. A lot of bugs and feature requests important for hardcore flight simmers require some level of knowledge and are therefore not receiving enough attention. Also, the forums quickly become a mess sometimes with many posts regarding the same subject which individually never get enough votes.

So under the slogan “no pilot left behind” I hope we can make a comprehensive list together of items in need of fixing for the hard-core simmer, engage in healthy discussion and make sure those items get the needed attention. Please, no discussions about gamers vs simmers, VFR vs IFR pilots, X-plane vs MSFS, real world vs sim pilots etc. I already made a start with some items:

Propeller Effects

The TBM and Kingair just won’t slow down when you slam the power to idle, I haven’t flown the TBM myself but I did fly the Kingair in real life. I can tell you, as soon as you kill the power its like throwing out the anchors. It must be a piece of cake to fly 160 kts until 4 nm with any turboprop and be stabilished at 1000 ft, not so in MSFS. Same is true for basically all propeller driven aircraft in FS2020, deceleration is not accurate.

I did recently test the TBM rate of descent with the engine OFF and throttle idle versus throttle feather with no change in rate of descent to maintain 120 kts. Even when selecting the throttle into full reverse in flight (again with the engine OFF) the RPM goes up significantly but remains within limits (in real life it would overspeed and get destroyed by centrifugal forces) but absolutely no change in rate of descent to maintain glide speed. This for me is the ultimate proof that no propeller drag is simulated.

During take-off roll, some right rudder is needed to stay on centerline, when airborne however these effects seem to disappear completely. On any single engine piston aircraft continuous rudder input (or rudder trim) is need to compensate for the propeller slipstream effect and P-factor.

  • Torque Effect - According Newtons 3rd law, every action has an equal and opposite reaction. If the propeller turns clockwise seen in the direction of flight (as is the case for most aircraft) the whole aircraft tends to rotate anti-clockwise. I’m not convinced the torque effects are simulated at all, this effect is most noticeable when applying full power at low speed. It is definitely not accurate, on the Kingair for example a lot of right rudder is required at the beginning of the take-off roll to compensate for torque effect in the real world.
  • Propeller Slipstream Effect - The air leaving the propeller picks up a swirl, rotating around the aircraft and creating a slightly different angle of attack on the wings roots causing a difference in lift, this effect is very small. The air also creates a different angle of attack on the vertical stabilizer, creating a yaw to the left (right rudder required). This effect is 100% missing in FS2020 which is a shame really as this is a very important characteristic of propeller driven aircraft. During flight no rudder input is required at all for coordinated flight is FS2020.
  • P-factor - This is not a very strong effect on single engine aircraft and is more a concern on multi-engine aircraft in determining the critical engine. This effect is only present at positive angle of attacks creating different AOAs on the up and down going propeller blades. The result is when pitching up the uneven thrust built-up creates a yaw to the left (again right rudder required). I’m not convinced this effect is present in FS2020, its not very noticeable at least.
  • Propeller Drag - A windmilling propeller (for example power selected to idle in flight) is a condition where the angle of attack on the propeller is negative due to low rotational speed and / or blad angle (constant speed propeller). A windmilling condition creates a ton of drag and therefore a lot of deceleration. In FS2020, unfortunately this is completely absent. Just try to reduce speed with the TBM or Kingair while on approach, the deceleration is even slower than a jet. This does not only affect turboprops though. The deceleration on all aircraft I’m able to compare to real life is considerably slower than the real deal, also glide ratios are unrealistically high because of this.
  • Propeller feathering - Non of the multi-engine piston aircraft and single / multi-engine turboprop aircraft in FS2020 have the ability to feather the propeller in flight. On the Kingair and TBM the ITT gets as hot as the sun when selecting feather on the ground indicating a problem in the turboprop engine logic but I guess that is a different subject. The Cessna caravan just doesn’t react to selecting the prop to feather at all.
  • Gyroscopic Precession - Mainly a problem on taildraggers where the aircraft yaws strongly to the left when lowering the nose during take-off roll. I haven’t noticed this in FS2020 so far, if it is there its definitely not realistically represented. Taildraggers also don’t have the tendency to ground loop in FS2020 but that is is not just because of missing propeller effects, different topic.

Lateral Flight Director Mode

  • When flying with Flight Director ON but Autopilot OFF and any lateral mode selected (HDG - FMS - VOR - LOC doesn’t matter), the Flight Director commands wings level and does only respond to the active vertical mode.
  • When selecting autopilot ON the flight director starts responding to the active lateral mode and the autopilot starts following it.
  • Of all the aircraft I’ve tested so far only the Airbus A320 seems to be unaffected but that might be due to the A32X mod.
  • This is not how the Flight Director system works on the real aircraft.
  • A big annoyance for me as well is that the VS mode, when selected commands the last commanded vertical speed instead of the current vertical speed.

In a real aircraft, the only difference between autopilot “on”, and autopilot “off”, is that when it is selected “on”, clutches in the electric aileron and elevator servos engage, so that the flight director steering commands are applied to the control surfaces. The flight director is part of the overall autopilot system, and the command bars should reflect any adjustments the pilot makes to the heading or pitch modes, whether the autopilot is engaged or not.

Incorrect ATC Phraseology

The ATC phraseology is completely made-up with some US influences (“Flight Following” for example). It is however not US phraseology nor European or ICAO. It’s pure fiction and has no basis in reality. Some issues summed up below:

  • “Ready for take-off” is an absolute nono in aviation, the word take-off is only used in relation to the actual take-off clearance i.e. “cleared for take-off”.
  • Climbs and descends to an altitude or height shall be spoken as: “climb / descent to altitude / height … ft” in relation with flight levels the word TO shall be omitted.
  • There is a lot of plain language used in the default MSFS phraseology, things like “I will contact you when leaving my airspace” should be “report leaving … airspace”, radiotelephony shall be concise and precise and plain language should be avoided where possible. “going to” and “with you” causes itches for everyone on the frequency.
  • “Is going missed”, “missed-approach” and all diferent forms used in MSFS is really simple in real life -> “going around!”.
  • “Flight following” is called “Flight Information Service” elsewhere than US (its not 100% the same but still).
  • Type callsigns are abbreviated as follows: first letter of the registration + last two letters of the registration, not the last three as used in the default MSFS RTF. Example: PH-ABC = PBC. The use of aircraft type or model in the callsign is rarely used in European airspace at least, only if there is chance of callsign confusion may ATC temporarily change the callsign to something like this.
  • IFR clearance request shall consists of: callsign, type, position, ATIS, QNH, flight rules, destination, “request enroute clearance”.
  • The IFR clearance normally consists of a SID and a squawk. The altitude and departure frequency are part of the SID.
  • “Copy”, “copy that”, “ready to copy”, “would like” are meaningless, it is “roger” or “request”
  • On most controlled aerodromes, VFR traffic shall request start-up as follows: callsign, type, position, ATIS, QNH, flight rules, destination, “request start-up” + VFR departure where applicable.
  • VFR departures are not to the “north”, “south” etc. they are via VFR routes and reporting points, although I guess this is difficult to implement.
  • Every initial call after frequency change on ground should include the position of the aircraft, taxiway, parking position, gate, holding point etc.
  • ATC won’t yell at you the moment you touch to vacate when able etc. they should first let you roll out to taxi speed before passing taxi instructions.
  • All the pulp at the end of the ATIS like, “inform you have information …”, “readback instructions” are all meaningless, just cut it out.
  • Although not strictly required, it would be easier to understand if the phrases “degrees” and “knots” where included in the surface wind reports.
  • ATC won’t tell you where the airport is and how many miles in the approach clearance. They assume you know where you are in relation to the airport. They might pass you the track miles to go to plan descent.
  • On uncontrolled VFR airports there normally (in Europe) is a FISO answering the radio, given traffic advise, departure arrival information etc. Not so in MSFS as all uncontrolled aerdromes are completely silent with nobody answering the radio.

And I’m sure there are many more issues…

The ground services communication via ground or tower frequency is also hideous. Its unrealistic, the only call which needs to be done on the ground or tower frequency is the request of pushback. All other communication is done via intercomm to the ground crew. Other service requests are done via the ground handler on their designated frequency.

There are loads of threads on this issue, I did fix most phraseology issues within the limits of the ATC system in this mod, there is a RTF manual included as well. Asobo, feel free to take it over.

Other ATC Issues

There are loads of threads on this one, let me know if someone knows a good one to include on this topic. In general, no radar vectoring, less functionality compared to FSX (eventhough SIDS / STARS and transitions are now included).

  • No vectors are ever given.
  • People often complain descent clearance is given too late, so far I haven’t noticed this myself. Descent clearance has been passed to me at or before TOD so far.
  • The approach clearance is given at a very late stage 20 / 30 nm out usually.
  • For some reason the ATC system prioritises RNAV and VOR approaches over ILS.
  • Active runway is not always wind related or the actual runway in use in the real world.
  • The altimeter system is US based with the transition altitude set to 18000 ft globally and inHg instead of hPa. It would be really nice to see this fixed including Transition Levels given by ATC and in ATIS broadcasts.

Flight Model

2D Panels & Multi-monitor support

The current 3D cockpits are nice but not very functional when not being used in combination with head track and VR. People who want to use multiple screens and home cockpit builders find 2D panels more practical for use. Read all about it here:

Offset Approaches

Two problems here:

  • Currently there is no support for offset approaches in MSFS.
  • Bearing pointers (RMI) give bearings towards localiser antennas, this is impossible in real life.

If we take Innsbruck - LOWI airport as an example, there are two main navaids. One of which is located on the airport for the LOC / DME East approach to runway 26 which is offset in real life. The other is located off airport and is used for the LOC / DME West approach to either runway 08 or 26 and for departures to the East. The navaid is located 10 miles from Innsbruck and doesn’t end up at any runway. When following this LOC you will pass the airport on the right at 7500 ft, you’ll then need to circle for either runway. The runway heading is 078 / 258 degrees.

  • LOC / DME OEV is located at the airport and should be aligned with the Innsbruck valley with a front course of 255 degrees, in MSFS however the localizer is aligned with the runway (258 degrees).
  • LOC / DME OEJ is located off-airport and should have a front course of 066 degrees and a back course of 064 degrees. Although this navaid is not related to any runway, both the front and back courses in MSFS are 078 degrees and following it puts you in a dangerous location…

This blows my mind really, this is such an iconic airport and it is one of the “hand-crafted” ones. Its unbelievable that they got the primary navaids so wrong. Note the blue bearing pointers in the screenshot below pointing to both the localizer antennas. This is impossible in real life.

Turns out that if you fly the LOC DME east approach you’ll end up flying parallel to the runway into the grass, as can be seen in the picture below:

This means that the geographic location of the localiser antenna is correct in MSFS. In real life the antenna is indeed located to the left of the runway into the grass but at an angle. The offset will cause you to end up over the runway threshold in real life before going into the grass when tracking the localiser. I have flown this very approach in real life down to pretty much minima, and you’ll definitely end up at the threshold eventually. You will see the lead-in lights on the left when breaking out of the clouds at minima, from that point you normally disconnect and visually line-up with the runway.

Line-of-sight Reception

VHF signals, radio, VOR, ATIS etc. should be able to receive within line of sight, this can be approximated with the following formula:

Rang in nm = 1.25 x (square root altitude transmitter in ft + square root altitude receiver in ft)

Its an approximation, in real life range is often shorter but still the range at which the ATIS for example can be received in MSFS is set way to low.

Camera System

  • Even with camera shake set to OFF, the view is never 100% steady and the viewpoints keeps moving in turns and turbulence making it hard to use the scroll wheel to adjust something, as soon as the viewpoint shifts you are zooming in or out instead.
  • On the subject of turns, when flown coordinated there is no reason why someone head would tilt in or out of a turn as is the case in MSFS.
  • Switching between different view modes is a mess. Why do you need to switch to outside view first before you can switch to fixed views or drone view and vice versa?
  • I have set up custom views to be able to quickly peak and change switch positions, I often need to select the wanted custom view twice before I have the correct view point and zoom.
  • Why can’t the drone cam be made more intuitive. For example, the translation controls with arrows, WSAD or whatever (like it already is now) and orientation with the mouse. It would be more intuitive to change orientation (pitch & yaw) with the mouse instead of keyboard assignments.

Turboprop Engine Model

The PT-6 turboprop engine model is completely wrong:

  • Torque and ITT with changing altitude are wrong in MSFS. It works like this, with increasing altitude the air becomes less dense the drag on the internal engine components (compressor primarily) decreases with reducing density. The engine Fuel Control Unit maintains the selected engine speed for which less fuel is required. Result is: same engine RPM, lower ITT, lower torque.
  • To compensate for the reduced torque the throttle lever(s) need to be moved forward, increasing the fuel flow. Torque is restored but the low air density (less cooling) combined with an increased fuel flow increases the ITT. At some point the ITT becomes more limiting than the torque and 100% torque cannot be achieved without exceeding the ITT limit.
  • In MSFS the torque increases with increasing altitude and vice versa and the ITT limit is never reached not even at the aircrafts ceiling.
  • The PT-6 is a free-turbine engine, this means that the compressor spool and the power spool (which drives the propeller) are in no way mechanically connected. This means that they can run at different RPMs without affecting eachother.
  • In real life you could therefore feather the propeller and have the powerspool running at a very low speed without affecting the compressor spool. Unfortunately in MSFS the ITT spikes if you do this and gets hotter than the surface of the sun (well almost). This is behaviour you would expect with a fixed shaft turboprop engine. There feathering the prop would cause the whole engine to spool down and the Fuel Control Unit (FCU) would increase fuel flow in an attempt to maintain engine speed with a high ITT as a result.
  • I have flown the ATR 72 which is not powered by the PT-6 engine but its also a free turbine so the principles are the same. On the ATR we had a propeller brake which we could use to run the engine for electrical supply and airconditioning with the propeller stopped, like an APU. Feathering the prop on a free turbine engine on ground is completely normal and should not cause a high ITT.



In addition to all the bugs:

  • No correctly functioning go-around mode on most aircraft, on the TBM it goes into ROLL mode (correct) when pressed but no active vertical mode.
  • When deselecting all vertical modes the flight director reverts to the basic PITCH mode, the pitch can however not be adjusted with the thumbwheel or nose up / down pushbuttons.
  • Autopilots SHALL DISENGAGE in case of stall warning activation. The MSFS autopilots remain engaged and when in pitch or vertical speed modes trim all the way nose up!
  • Autopilots in MSFS don’t disengage when control input is given by the pilot.


The GPWS mode 3 “don’t sink” alert sounds often at incorrect times. Weirdly this is the only GPWS mode implememented. What about the other GWPS modes? And what about the nuisance “don’t sink” alerts? This mode 3 alert should ONLY be active during take-off and go-around from 50 to 700 ft radio altitude and trigger when the loss in barometric altitude equals 10% of the current radio altitude. (example: 500 ft RA, triggered when 50 ft barometric altitude is lost). This mode does not arm on approach until below 200 ft RA.

What about further GPWS implementation? I remember default FSX planes had GPWS:

FMS / GPS Pre-flight Preparation

Currently the route is automatically loaded into the FMS or GPS before flight, what about people who want to practise setting up the FMS / GPS themselves? Why is there no option to disable this automatic loading into the FMS / GPS?

Control Bindings

Being able to keybind all controls especially the different GPS controls is important to anyone who wants to avoid using the clunky mouse controls for knobs.

Live Weather & ATIS

There are loads of threads regarding this issue. In my case the real world weather, the weather in the simulator and that given in the ATIS are three completely different things. I’m assuming these will be ironed out eventually.

The ATIS always gives 3 clouds layers, one around 700 ft, one around 4000 ft and one around 11000 ft. Also the temperature and QNH don’t always match. Visibility is given in miles and is either 3 or 6. Would be nice if this would change with the selected units of measurement.

The current weather system does not allow a visibility below 3000 m more or less and therefore no low visibility take-offs and or landings possible. It is possible to put the cloud ceiling on ground level, but precise control of visibility / RVR is unavailable currently.

Icing Effects

Icing effects are hard to predict in real life, icing conditions are pretty easy to predict, for icing we need three things:

  • Visible moisture, can be clouds, visibility < 1600 m (1 statute mile), precipitation or taxiways / runways contaminated with water, slush, snow, ice etc.
  • A Static Air Temperature below 5 degrees on ground or below 7 to 10 degrees Total Air Temperature in the air (depending on the aircraft type).
  • The lack of freezing nuclei, this causes droplets to become “supercooled”, in this case they stay in liquid form although the temperature being below 0 degrees C.

The difference between TAT and SAT (the Ram Air Temperature or RAT) is caused by the “ramrise” due to forward motion of the aircraft and the resulting friction with the air. The ramrise can be approximated by the following formula:

(TAS / 100) squared

Slightly positive temperatures won’t protect against icing. This is because the airframe itself is still cold from cruise at higher altitude, due to cold fuel in the tanks (cold soaking) and because certain areas on the aircraft, primarily wings and engine inlets, lower the static pressure (for the engines only at low speed, before ram rise takes effect) with a resulting lower local temperature.

What kind of icing, rime ice, clear ice or frost depend on the relative humidity, the size of the droplets and the temperature of the droplets with the larger droplets causing rime and clear ice. It is hard to predict what the kind and severity of the icing conditions will be, even two clouds with the same conditions may vary.

Supercooled Large Droplets (SLD) is what causes severe icing conditions. They are primarily found in Cumulus and Numbostratus type clouds. These severe icing conditions are rare, and the extend is usually not very large. In cumulus clouds the icing will be intermittend, so the horizontal extend is relatively small. In Nimbostratus clouds the horizontal extend can be large but the vertical extend usually isn’t. In other words, change of course, climb (if at all possible) or descent just a few 1000 ft are usually sufficient to escape.

The icing effects in MSFS are unrealisitic in a couple of ways:

  • The severity is too high. Almost every clouds immediately causes icing and at a extreme rate.
  • In the real world ice builds-up only on the frontal surfaces of the aircraft, i.e. leading edges, nose, spinners and forward windshields. Not on the side of the fuselage or cockpit side windows.

I noticed on the TBM for example that the side windows freeze when selecting the windshield heat OFF, first off all I don’t thing those windows are heated in the first place, second those areas won’t pick-up any ice in the real world.

Flight Planning

Although the world map looks very fancy, the functionality is less in terms of flight planning compared to FS9 or FSX. There are no airways, the automatically selected routes won’t pass the Eurocontrol IFPS server when used in the real world.

I remember the FS9 / FSX route planner actually using real world airways from A to B etc. The introduction of SIDs and STARs is an improvement though. However, it is unfortunate that the departure route doesn’t appear in the ATC IFR clearance for example.

As said before an option to disable auto loading of the flight into the FMS / GPS would be nice for people who want to learn how to program the FMS / GPS. It would also be helpfull if a flight can be planned or modified without having to go back to the main menu.

Mass & Balance

There still is a bug in the mass & balance section when using metric or hybrid as units of measurement. I assume this is a conversion error and easy to fix but like with everything not related to improved visual effects it didn’t got any attention so far.

This bug only occurs when using Metric or Hybrid units and only affects the payload slider. When moving the slider, lets say to 50% and release, it jumps up to a certain percentage. I tried to load fuel in such a way there was only 1000 kgs available as max payload and selected 50% payload on the slider (should be 500 kgs), when released it jumped up to 56.9% and 570 kgs.

Again, only when selecting hybrid or metric: When selecting fuel in kgs you can’t type in fuel per tank, it keeps changing whatever you’ve typed. Same for the different compartments, seats etc.

Ground Handling for GA Aircraft

What is the deal with the pushback tugs for all the tiny GA aircraft? First of all its not realistic, nobody uses a towbarless tug to pushback a small Cessna from a parking spot while the aircrafts turning radius allows it to turn around on a post stamp. On top of this they are getting run over all the time as you can’t get rid of them if you just want to taxi out yourself.

If you have anything to add or correct, please drop it in the comments.




Obviously Multi-Monitor support is a big one. Also, Autoplilot bugs (and missing buttons/knobs on certain aircraft).
You are certainly correct in the lack of urgency in fixing features for the serious simmers and MS/ASOBO do respond to the masses, which puts us in the minority. Just looking at the schedule of fixes- visual vs flight-sim-specific, confirms that. Thanks for the new thread.


Let me know should you have any good threads to add, preferebly those with the most votes already. I’ll hope we can keep updating this list with as much background information etc. as possible, get threads the attention and votes they need and get Asobo’s attention :muscle:

1 Like

This has been done to death right from Alpha and is already on the Zendesk ticket system. I submit that this topic is, once again, unnecessary and a waste of time.


Maybe, maybe not… hopefully we can at least bridge the gap in knowledge here on the forum and within the Asobo team. There are enough knowledgeable people out here which know what there are talking about.


I bought three monitors already in prep for this. How hard is it to add? Would make it a professional level and even casual gamers would love it if they have three monitors.

1 Like

patience, everything will

I would also add ITT which is not correct. Like for example in the TBM, the ITT rises when ascending. Also the torque rises when ascending. Torque should decrease when ascending which implies you push the throttle forward to keep speed. Now the torque goes up when ascending, and you have to lower throttle. When going to higher altitudes ITT temp should be the limiting factor when your throttle is all maxed out.

Being able to keybind all controls, especially the different GPS controls is important to anyone who wants to avoid using the clunky mouse controls for knobs.

1 Like

Jup good one

There are some issues on twin pistons that need to be addressed. For instance, on the baron, and I believe even on the da62, you cannot currently fully shutdown and secure an engine in flight. Rudder trim also appears to be INOP, at least on the baron.

Under ‘Flight Model’ (which is a blunt enough point in itself - the whole thing needs rebuilding), I’d like to add the removal of all of the ‘pre-determined effects’ which affect the handling of the aircraft such as the lateral wobble which seems to be used to suggest instability.

It would be nice to understand how the flight model actually works. I’m not a programmer. The impression I got from the feature discovery serie was that it is physics based with “1000 simulated surfaced” with their own aerodynamics and what not. But I’m getting the impression its mainly entries in the aircrafts .cfg file which is doing the job. I don’t understand how those two things work together.

Neither do we, really - the SDK for the flight model is woefully inadequate and much is currently left to guesswork.

1 Like

Whom is this sim made for? Simmers? Pilots? Gamers?
Maybe the answer for this, is what make the devs take decisions.

Hopefully for everybody but listening to the masses will not fix fundemental flaws in the sim unfortunately as not everybody has the knowledge to recognise something is off and understand the underlying problem. They have already said last Q&A that up / down draughts in thunderstorms were dumbed down to prevent bug reports because people “wouldn’t understand the reason”…

I just hope the same thing which applies to aircraft does not apply to flight simulators. Something which is supposed to be multi-role always turns out flawed, more expensive, with extra complexity it wouldn’t have otherwise and far worse at the job at hand compared to something particularly designed for the job. I hope a broad audience at least means that there is sufficient revenue to keep developing the title.

1 Like

Apparently quite challenging as the multimonitor support request is currently slated to be released between 2021 and 2022

No way! Really?! 2021 / 2022? So VR before multi-monitor support?

Where did you got this info from?

The update today

See #4 on the list

Wow, I did miss that. Well 2021 is soon but 2021 / 2022 doesn’t sound very hopefull.