ZaltysZ

If AOA is maintained... Are you sure it is maintained? Multiple people have already said in multiple ways, that current control stiffening model makes it hard to maintain the same AoA without changing stick position, but you just brush it off everytime by interpreting us in bizarre ways. Get Tacview, make it export telemetry data and do your test. Then watch how well that AoA is maintained by you. You can get AoA, Gs and other readings at every point of Tacview track.

Theoretically, if FFB was done right and you set FFB forces high enough, you should have more natural handling going through wide speed range in turns/loops. Strong FFB force should make it very obvious when to stop pulling, if you didn't want to pull more when virtual pilot, and that would solve overcontrol issues while slowing down.

Where?

It is getting weird. :) Virtual stick is the stick inside the cockpit. It is directly linked to aircraft control surfaces and its position is shown in red square (ctrl+enter). When you move your real physical stick (the one on the table), your input is transformed to movement of virtual stick by application of response curves and stiffening model. There is no autopilot, instructor, mouse aim, fly by wire or similar assistance.

Virtual joystick limits deflection angles of control surfaces. You don't need fully deflected elevator to reach critical aoa.

Why? Are implying that stall is impossible unless stick if pulled fully back?

How? Yes, but if you can't keep 400km/h or higher, virtual joystick won't stop at 50% anymore. Lower speed = higher limit. If your speed drops to 350km/h, virtual joystick will stop at, let's say, 60%, and if 60% is too much for aircraft, it will stall.

This K4 begins to stiffen from 300km/h. The faster you go, the more elevator deflection you loose, that is why I am giving you examples with high speed, because at high speed you can overpull the physical stick without danger that virtual one will be overpulled too, and then see how virtual one is being gradually overpulled while you loose the speed. Due to large difference between positions of virtual and physical sticks at high speed it should be easier to see how stiffness modeling works and how it causes overpull with decreasing speed. Exactly the same principles apply when you go from 500km/h to 380-420km/h, but due to smaller speed change and larger max deflection angle, it is easier to dismiss the modelling of stiffness as culprit and put blame on other things. I have attached a picture from your 2nd video. Left one looks like a settled condition, and right one is taken at start of shudder. Right one has virtual stick pulled more. The difference is small, but important if you already were near critical aoa in the left picture. Was it due to overpull caused by stiffness modelling? Maybe. Was it pilot error? Maybe. That is why such lower speed case isn't suitable for explanatory examples.

It is a long video, so description of key points of timeline would be great. Does stall at around 1:00 illustrate your issue well?

No. I said "relation between them is 1:1 only at lower speeds. As your speed increases, virtual stick will deflect less and less, despite you pulling your real stick fully." Max deflection % is not constant, but is a function of speed, or in other words, max deflection % depends on speed. It might be 30% at high speed, 60% at medium speeds and 100% at low speeds. If you pull beyond max deflection % of virtual stick, game will ignore that, but if you loose some speed virtual stick will deflect towards your physical stick until new higher max deflection % is reached. This way you can overpull by loosing speed even if you don't pull your physical stick more. Check with ctrl+enter. Begin turning at high speed with constant 80% deflection of physical stick and observe how virtual stick position changes as you loose speed.

Deflection of what? Your real stick or virtual stick? Even if you use linear response, relation between them is 1:1 only at lower speeds. As your speed increases, virtual stick will deflect less and less, despite you pulling your real stick fully. Example. Let's say you are turning at high speed at which modeled pilot only has enough strength to deflect stick no more than 30%. If you pull 10%, virtual stick will deflect 10%; if you pull 30%, virtual stick will deflect 30%, but if you pull 40%, 70% or 100%, virtual stick will still deflect around 30% only. This is how heaviness of controls at high speed is simulated. The problem begins when you pull your real stick beyond limit of virtual stick and allow the speed to bleed. As speed bleeds, heaviness of controls decreases, and virtual stick catches up with overdeflected real stick. Remember, it is ok to have real stick at 70-100% at high speed, because virtual one deflects only by 30% and that keeps aoa within allowable range, but if you keep real stick at same 70-100% at lower speeds, virtual one will be near 70-100% too, and that will cause a stall. P.S don't use these % for FM reference. They are 'fake' and chosen for explanatory purpose.

It is very likely he did not increase the pull and kept the physical stick at same angle which settled at mid of turn. If it was static ratio between input and output, he would not have stalled, however this ratio is dynamic (game is trying to substitute a "force required" with "deflection required" depending on speed), so if you settled on the edge of high speed pull and keep that deflection angle as speed drops, virtual pilot will be the one which increases the pull despite you not pulling the physical stick more. You need a gamer mind to overcome this mind f*ck. :pilotfly:

It seems these axes command rate and direction, but not the absolute position. I am puzzled why they are implemented this way. Such action may be useful for things like canopy handles, but certainly not for trim.

CloD should not be included in his post, as it allows 2 axis adjustment for each gun separately allowing not only single point harmonization, but custom patterns too.

Is AN/APS-13 volume knob supposed to affect rear warning radar buzzer volume? It does not affect it now.

http://http1.files.eagle.ru/dcs/manuals/DCS-P-51D_Flight_Manual_EN.pdf

Check page 116 of the manual.

There is usually no "Stall AoA", but sometimes "Critical AoA" is given, or more often, even whole chart of AoA vs CL (lift coefficient) is presented. Lift depends on both speed and CL, and CL depends on AoA. Stall speed is just lowest speed at which aircraft can stay at flight while being at critical AoA. Increase AoA past it, and you won't stay at level flight even if your speed indicator shows stall speed. You can compensate low CL with high speed, and low speed with high CL (high AoA), but within limits (CL past Critical AoA begins to decrease sharply). You can't have one of them at zero, or near zero without making other impossibly high.

Something like 65-90 degrees.

In fact you can avoid using CDU for calculating distance between 2 mark points if you want. As it will take some time until you reach point "C", you can use that time for selecting the first mark point as steer point. Then you will see ever increasing ground distance to it on the HUD while approaching point C, and just note the final reading while being over the point "C". This way you don't have to check time, calculate the distance or keep constant speed/altitude.

Look at diagram. Imagine that you are have radar straight ahead of you on the path which coincides with edge "c". Let's say, you turn 45 degrees to right from your initial heading and now you are flying on edge "a". Your goal is to measure length of "a" by flying straight until you reach point "C", which will create a nice 45(ac)-45(bc)-90(ab) isosceles triangle and make a=b. It is very easy to know when you reach that point, because radar will slip exactly to your 9 hours. That exact moment radar will be at same distance from you as you have covered from initial turn to the moment radar slipped to your 9 hours, because edges opposite to equal angles are equal (ac=bc, thus a=b). By the way, you can forget flying at constant speed and watching clock. Just create mark points: one at initial turn and one at the moment radar slips to 9 hours, then check the distance between these points.

His triangle is an isosceles triangle (ab=90 degrees, ac=45 degrees, bc=45 degrees), so a=b, while a<>c, nor b<>c.

Point is simple: limiting how many users can use single license. Ideally it should be 1, but not all forms of DRM are capable of that without seriously hampering the user. Method DCS uses generates a hardware id and ties license to it upon initial startup. This can be done automatically by inputing key and allowing DRM to register hardware id and key pair via internet, or it can be done manually by inputing hardware id and key on website and getting activation key which is correct only for computer with specific hardware id. The later feature allows DCS to run on system which is never connected to internet. However, exactly the same feature means that validity of license is checked only once (excluding hardware id changes) upon activation. This means that subsequent activation on different computer has no ability to remove, invalidate license on previous computers. Unless, some means are taken, such DRM would offer protection comparable to oldschool cd key, but just with much more complex inputing method. One of possible remedies for inability to remove license on previous computers is limited activations, because they prevent you from using the same key on unlimited number of computers.

One buys a campaign, completes it, then passes it to a friend. Friend completes it, passes to other friend, and so on. 3 activations gives less such spread than 10, but still allows some maneuvering around screw up cases.

Pre 1944 doctrine told escorts to stay close to bombers. Germans could attack with altitude advantage, cherry pick the targets, conserve the ammo and simply safely disengage because escorts would not follow them. Things changed in 1944: escorts were given more positional freedom and even were allowed to chase attacking fighters. Think about how the situation changed for Germans in 1944. They could intercept bombers and attack with altitude advantage, however bombers likely had escorts above them. That leaved Germans choosing between making a single pass on bombers and then having to shake the escorts off (read: hard to make repeated attacks on bombers), or preoccupying escorts from the start and almost forgetting the bombers. No ability to overwhelm in numbers meant that whatever role Luftwaffe pilot got, he likely ended mangled with escorts while still having lots of 30mm. So, yes - once engaged with fighters, Germans did not conserve 30mm for bombers, they just hit whatever they could with everything they had trying to seize opportunities while minimizing the time they were tied to single target, because that target usually had a friend rushing to save him.