ZaltysZ

Is this a custom mission? Each mission can have its own difficulty settings, which override global ones, unless you force usage of global settings for every mission (not sure if responsible checkbox is fault free completely, I had some inconsistencies with it in the past).

You need to keep in mind that XP is a civil aviation sim with possibility to throw some military machines in, but it doesn't even remotely make it close to combat sim. Now, you can take a liner jet and abuse it to extent of imitating RC plane, but most fans of XP won't really get you, because they don't normally even approach RL control limits (you know, imaginary passengers love smooth ride and so on). Different use case, different priorities, different sandboxes to play in.

If you talk about cannon, then just trim for stable flight with K, T, H channels on, lock the target with Shkval and use your cannon. Cannon is not fully fixed, so targeting computer can aim it for you and give focused fire even if you and/or target is moving. If you talk about gun pods, then the procedure is pretty similar to unguided rockets employment.

I use 3 methods: 1) FD on 2) FD off, but trim button pressed. Effect is same like with FD on, but I don't need to fiddle with turning FD on/off. 3) FD off, directional (H) AP channel off. Aircraft trimmed for keeping piper on target. Final corrections feels like stick is super insensitive, so precision rises. You probably need rudder pedals for this method to work correctly.

Basically it would mean less of such anomalies:

Tail provides stability and better control at high speed forward flight. In case of low speed, tail does not do much in coaxial design like Ka-50.

It is so weird sometimes, when AI is doing plausibly more or less, but then it suddenly screws its position by doing some unnecessary barrel roll. :(

Do you have rudder assist turned on?

As far as I remember, pilots used to be obsessed about 2 things: 1) engine failure while hovering results in not so successful landing 2) launching weapons while hovering has greater risk for stalling one of engines due to exhausts of missiles/rockets. After that comes all other dangers of hovering: artillery strikes, flanking guy with RPG in the bushes, higher heat signature due to more power used in hover, easy of aiming at static target, delay between being static and moving (i.e. how quickly you can get out of here). In addition, some helicopter like Mi24 have design issues, which make hovering attacks problematic. I.e. Mi24 wings interfere with downwash and make hovering difficult while carrying combat load.

Could it be that elevator reversal had contributed to that feeling?

I don't trim at all as it is not hard to correct yawing tendency with rudder pedals. Most people are having problems with take off because they are too pushy with the plane: use too much power on runway roll and/or lift the tail too soon or too quickly. If it is done in gentle way, P51 will take off like on rails.

If I recall correctly A2A Spitfires are MK.I and MK.II. They (and their carburetors) are kinda ancient when compared to P51D. It would be more sense to compare DCS P51 to A2A P51, and according to videos on youtube A2A P51 startup seems easy too. P.S. I remember comments about P51 tested by germans. 2 things were highlighted: high speed and easy startup in cold weather.

Not necessary. There are pitfalls. If you want performance benefit from calculations on GPU, you will need highly paralleled/vectorized computations with infrequent need to transfer data between CPU/RAM and GPU, or else access latency will eat most of performance gains.

PhysX is a recurring topic for flight sims, however there are issues, because of which PhysX is still not used in them: 1) It is tied to NVIDIA. CPU implementation is poorly optimized and falls behind in performance too much. This means that devs have to limit its GPU assisted usage, or they will loose user base with ATI cards. 2) Mostly only effects (smoke, debris, explosions and etc.) can be run on GPU. Physics calculations like collision detection, ragdolls and so on always run on CPU. 3) PhysX is just physics engine middleware, which allows devs to quickly write code for trivial/general case physics interactions and offload some stuff to GPU. If you want general rigid doll dynamics, vehicle dynamics, particle effects, then it is ok as it saves time and gives good performance. However, if your game/sim requires very specialized calculations (like detailed FM, DM), then most of PhysX features becomes too crude to get wanted fidelity. PhysX is designed for games (whose players do not moan that zombie head bounces few centimeters too high) and not for sims or general computations. 4) Pluging PhysX into existing engine is not an easy task. Even little rag doll soldier with RPGs might require big architectural changes.

I wonder what does "object" really mean? Quickly browsing through EDM docs, I see that EDM can contain multiple objects: renderable geometry, bounding box, collision shell and etc. That could explain variation in object count. Can these 1 object models be destroyed in game? Or are they only for rendering?

You forgot to fix the "biscuits" part. :pilotfly: http://www.colourbox.com/preview/1723650-849612-fat-with-vodka-on-a-newspaper.jpg

Do you have huge drop when heat blur is rendered or even when it is just supposed to be present (i.e. you fly slow and look at HUD)?

1) Harbor scene is very dense. Lots of objects close to each other, what means that they are all rendered at the same LOD. Looking at city at angle might be less heavy on resources even if more objects were rendered, because most of these objects would be rendered at distance and so at lower LODs. 2) Animation = CPU hog, always. I wouldn't mind if there was an option turning off scenery animations.

It seems like ordinary programmer cheat-trick to workaround some kind of problem with crude FM. Remember that FM in FC2 has such pearls as hard speed limits of missiles and so on. If you fix FM parameters in one part of envelope, you may mess the missile in other parts. I think it is better to invest time in tweaking AFM in FC3, as it is more likely to be possible to bring missiles closer to real life using AFM.

http://lockon.co.uk/dev_journal/characteristics/ First 2 charts should give a clue for Hog pilots to not play heroes while Su25 is around and aware.

Tearing usually means absence of VSYNC. Did it start after driver update? Sometimes vsync gets broken by bugs in drivers (i.e. checkbox is on, but it simply does not work). ATI is kinda well known for having reappearing vsync bugs. If you were using older driver version when it still worked fine, try downgrading.

Holding trim button on non-FFB stick is like flying with FD on.

I think you read too much between the lines. :) Article speaks about B/C not requiring hardware changes for simple software updates, and how it is possible to update software just by reprogramming missiles inside containers via cable.

I have it reversed too, but it has nothing to do with idle detent. If you have it done according to the manual, then you should have smooth travel between idle and max and remaining detent between off and idle. Bringing throttles to off position should still register as holding DX29 and DX30 buttons pressed (check windows control panel, game controllers applet). Maybe you have corrupted DCS controller profile?