When FS2004 came out, there was a lot of head scratching and worrying over the Flight Dynamics of the new program. As I remember, several of the variables that had been used in previous versions of FS had been tweaked or deleted in their entirety...Some prominent community figures such as Lou Betti basically said there was "nothing to worry about" followed by a long drought without any pay or freewarewere.But, with the release of the SF260 and the Decathalon, it seems like the flight models may actually be better than what was available for FS2002???For real pilots that have flown both FS2002, and FS2004, can you comment here? Also, there don't seem to be any "heavies" that have been released with companies that draw special attention to the flight dynamics of their planes...

Hi Caveman,One thing that was immediatly apparent when FS2004 was released was that the default aircraft had improved flight dynamics. One example is that the default Extra 300S is now spinnable.I do remember the problems developers seemed to face but there were some who believed that the FS2004 physics engine is in fact improved over the FS2002 physics engine, and that fewer variables are in fact needed in the air files because the new physics engine infers those missing parameters based on existing parameter information.The RealAir SF.260 and the Decathlon do exhibit better flight dynamics than they did for FS2002, so my guess is that the flight dynamics dilemma is solved.James

Not better, just different, I reckon. Some problems from the previous version WERE addressed, but they left others in their wake. And in turn it took a while for those small changes to be identified and addressed as they had the usual side- and knock-on effects. The problem remains that there are 100 eye candy specialists for every flight model specialist, and many of those flight model specialists are now allied to payware teams and have rather turned their attention away from the rest of the freeware community. Law of supply and demand, I guess.With a few notable exceptions, if you want realistic flight modelling, you have to pay. THAT'S the difference.Allcott

Greetings 'Caveman',You said,<>No, we were grappling with changes to the Flight Model (FM). The changes to the Flight Dynamics (FD) were relatively easy to understand and resolve. I will try to clarify the difference below.We now understand what Microsoft changed in the FS9 FM, however since MS never tell us what they have changed, every time a new FM comes out, we have to waste months testing things that have not changed, just to be sure that they really have not changed.No one wants to install update after update because the early updates do not work. Look what irate consumers have said about that process in this forum, and in the producer forums. Equally no designer wants to have an aircraft nearly ready for release and then realise that they have to go back to square one. It is a good idea to work out how a new software version really works before releasing add on products, or updates, for it. A new flight model therefore causes months of delay in production of both freeware and payware.It may help if I comment on three quotes from different posters.<><><>Real pilots are not well placed to understand the issues involved. Most consumers, including pilots, confuse the Flight Model (FM) with the Flight Dynamics (FD).Both the FS8 and FS9 FM support spin. The producer of a particular aircraft decides whether to include spin in its FD. This takes extra time, which costs money. In FS8 Microsoft decided that only the default F4U Corsair would have spin encoded in its FD. In FS9 they deleted the F4U altogether and encoded spin within the FD of the Extra 300S instead. This tells us nothing about the FM changes that caused the problems and delays. It only tells us about the time spent writing the FD of a particular aircraft release. The FM and the FD are not the same software. They have different declared variable lists. The FM is always released by Microsoft. They alone decide what variables are declared within the product code. The FD may be released by Microsoft, or by somebody else, but whoever creates the FD they can only write to FM variables declared valid by Microsoft. Improvements in freeware or payware FD in FS9 are due to the FD author. They are not due to changes in the FM. When Microsoft change the declared FM variables between versions FD updates have to be rewritten to work with the new FM declared variable list, whether it is longer or shorter, better or worse. Microsoft decided that a number of software variables within the FS8 FM would be deleted or else replaced by hard coded constants in the FS9 FM. Some new variables were added, but there was a net loss. Conversely the FS9 FD have a net gain in additional declared variables. They can now write to software variables that were always defined in the FM, but were not previously defined in the FD. Optional FD upgrades for some aircraft in FS9 therefore became a possibility and some producers decided to delay the necessary FD update needed to match the revised FM until an FD upgrade had also been written by the FD author.Collectively we have now devised generic FD work arounds for almost all of the changes which Microsoft made to the FM declared variable list, but some problems require individually coded solutions which cannot be supplied by FD authors. A specific example follows only for those who would like more detail. Aircraft with constant speed airscrews need manifold pressure gauges to allow the consumer to set climb and cruise power in accordance with the handling notes. The inlet manifold pressure drop in different normally aspirated piston engines differs by up to 7%, even if vintage engines are ignored. The FS8 FM contained a declared inlet manifold pressure drop variable that the FD could write to. The FS9 FM has lost this variable even though it is still in the FS9 FD declared variable list. Before the release of FS9 it was possible to produce a default MAP gauge for use in any panel. The project FD author then wrote the driver code to match the generic gauge to a specific engine. Now in FS9 the gauge author has to write the driver code instead and each normally aspirated piston engine fitted with a constant speed airscrew needs unique gauges containing their own unique driver code. This is far from trivial.Panel producers, are having some trouble coming to terms with this change. Some still think that FD authors can match the engines to the gauges. Consequently some freeware updates and new releases said to be FS9 compliant are not. Hopefully all payware releases and updates are fully compliant. Only those consumers who have the real POH, or flight manual, will be able to tell whether the MAP gauge in an afflicted aircraft is reading correctly. To put it another way, whether the correct power is (still) being produced when you use the throttle to set the specified manifold pressure (and rpm) for climb and cruise configurations. If a particular aircraft, with normally aspirated engines and constant speed screws, released as freeware for FS8 depended on the use of default or generic gauges, and the project did not involve a gauge author, the problem of updating the release becomes one of finding a gauge author willing to write unique freeware gauge code for each afflicted aircraft. Consequently many may never be updated for use in FS9. The shortage of FD authors is real enough, but in any event FD authors cannot resolve all the problems that Microsoft have created by deleting variables from the FM. FSAviator

Flying real cessna 172s I have always found that the FS FDE for their 172s are highly touchy and unreal.

I agree with respect to the default ones. There seems to be too little inertia. Being able pop a 172 into 1500ft/min climb by twitching at the wrong moment is silly. If I were to do that in RL it would just stall. I think if you plug real numbers in you don't get real performance out. To get a more realistic 152 in 2002 I had to double its weight, halve the power, incread parasite drag and reduce induced drag.

(deleted before available)

One thing that has always bugged me about the flight dynamics in MSFS is the way an aircraft can be almost mechanical in its movements. This is most noticeable during replays of landings. When control inputs are made the plane just sort of clanks over from one attitude to next with no overshoot or visible ramping of the rates. Physical objects, especially aircraft just don't move that way. There's always some ramping up and ramping down of the movement where as the rates in MSFS almost look constant. The effect seems to be intrinsic in the FM as it's not an effect of frame rate and is seen in all types of aircraft, including the best commercial offerings.TonyDigital-Flight

I think replays may be sub-sampled to reduce storage space so this makes movements more jerky that the original.Ian

I've always thought of MSFS, at least from MS's CFS2 as being one of the better sims when it comes to having the look of boyancy, especially from spot view. I long complained of simulations in which the tail seemed to be tied to the joystick as with a puppet. Very mechanical!MSFS & IL-2 Sturmovik have been two favorites for NOT looking as if the aircraft is in a direct link with the joystick. Oh well.... different opinion I guess.But.............. I should also mention, that as with real aircraft, the amount of "overshoot" as described will greatly vary. I just watched a video of a Van's RV4 with camera's facing for and aft during a 4 point roll. The plane actually followed the aileron deflections exactly, at least as close as the eye can tell. When the aileron neutralized, so did the aircraft. That's pretty precise.................and could be the reason I bought a RV6 :) L.Adamson

>I think replays may be sub-sampled to reduce storage space so>this makes movements more jerky that the original.>>IanCould be. As I've mentioned in my other reply, the movements don't seem very mechanical to me. But then I do use spot mode a lot, and replays very little.L.Adamson

Hi Tony,The replay modes in FS9 (and other versions) reduce the resolution of movement by quite a few factors and do not in the least represent the actual animation of the sim in real time.As an experiment, try loading any decent aircraft which is not deluged with polygons, hi-spec gauges and eye candy, and with a decent flight model. If you can get 30 fps, or preferably more for generous overhead, and fly stutter free, I think you'll see a mechanical-free movement.Regards,Rob Young

No, my replays are smooth and it's there in realtime as well......TonyDigital-Flight

What I'm seeing is mostly noticable during landings. Movement looks more natural at me at normal flying speeds.TonyDigital-Flight

Hi Rob,I've spent a lot of time looking at this and I can tell you that the effect doesn't seem to have any relationship with frame rate. My replays in general run stutter free at the same frame rate that I see in real time. I can pull down all my rendering settings to minimum and let the FPS run wild and there's absolutely no effect on the kind of mechanical movement I'm talking about. The planes move smoothly, they just don't move from one attitude to another in a realistic way. I can see the effect in realtime from within the cockpit so I'm pretty convinced it's not a replay issue.TonyDigital-Flight