Mr Whippy

Thanks for the detailed reply. I know what you mean wrt IAS vs TAS. It just seemed a logical approach that with IAS there would be some linearity in the control input vs expected aircraft output vs altitude. But I suppose the IAS vs TAS curve is different for every aircraft, since all aircraft have different aerodynamics and thus performance varies non-linearly vs air density. I suppose I need to fly with an IAS and TAS readout to get an idea of what is really going on and build up the relationship mentally between the two values in different conditions. It's ironic really that IAS is used with such an aircraft, where the pilot is deciding to fly so much with knowledge and consideration and fineness, to then rely on an automated value like IAS. Given the mantra of the flight system design, you'd think TAS would be used entirely, and the pilots just understand what the capabilities are at varying air densities and TAS. I'm going to have to find a book about the Su27 that has lots of information from the pilots and engineers. I'm intrigued by some of these choices because they seem illogical, but there must be some logic there which suggests I'm missing a lot about these aircraft! Thanks Dave

I don't expect the FBW to change physics. But I'm still curious why you'd desire such overly sensitive controls in given flight conditions when you could instead have them not overly sensitive. Ie, if the aircraft wobbled about somewhat and the not unsubstantial forces on the arm would move the stick enough to propagate a feedback loop. Unless again this is a limitation of our non-FF sticks operating in 1G downwards conditions while sat at a computer and stationary. So to summarise. At 30.000ft and TAS of say 400kts, the Su27 controls input range is about 10% useful, with the other 90% just range that will mean you get too much AoA and lose control? So the pilot has to increasingly finely balance inputs under very high g forces, with a big waggly stick that flaps around, to the point it may feel like it's on a hair trigger, vs at 10,000ft at the same TAS? I'm more than happy to accept this is the way it is, if it really is. I just can't see what pilot or engineer would see virtue in that behaviour if they can get rid of it. Dave

In the training mission for guns the SU27 flies like a dream. I can completely appreciate here that you'd have to fly like a complete idiot to do something wrong, and the stability control feels perfectly measured to keep the aircraft stable within the reasonable flight envelope any person would use. But then in the quick intercept mission it seems really hard to fly hard without slipping into feedback loop levels of AoA that just leave you unable to correct and you just end up sat dead in the water at about 250kts. But like we keep saying, it's apparently a beta, so to really say it's right or wrong right now seems fruitless I suppose. I hope they finish it up soon. It's a bit gutting that it's unfinished as it's the main reason I bought the FC3 after seeing it on Steam... it didn't say it was a beta version :( edit: Hmmm, I think I spotted my problem. Replay view shows feet and kts, cockpit is showing kph and metres? I was probably flying higher and slower than I thought??? My brain is conditioned for imperial units for flying ;) As much as I bemoan imperial, for flying I've just got used to them over many years of flying Falcon and Tornado etc. But I still think that at 20,000ft (6000m or so), and 400kts, even with my controller curve more S shaped (so about 15% input is probably about 10% request to the FBW), is ample to have my AoA be exceeded after a short while (ie, transient manoeuvre of a few seconds), and feedback loop ensues and I lose all airspeed and AoA runs up to some limiter. Why wouldn't the controls be re-ranged for such flight envelopes? You essentially just have about 70% of your stick range made useless. Is this how it really is in such flight conditions? Dave

Do you fly with the stability control off? Or are you relying on the stability control already? If you are using stability control already, by definition you are letting the game intervene on your inputs. To a level determined by the setup of the stability control system. The fact is the aircraft wouldn't even be safely flyable without input intervention. OK my specific point is different about cutting input, but the spirit of the point is that you're already giving up all of your inputs to be processed. So how is the stability control simulated? It sounds like you know it's simulated fine, so describe even in the simplest terms how it functions. LUT's, parametric type systems, single value limiters? I get where you are coming from, but if you believe the current SU27 stability control is a perfect copy of reality then that is fine. But I think many don't, ie, auto-trim. Why wouldn't you let the aircraft auto-trim? It has gyros and FBW, why not have a dead stick fly zero pitch/roll/yaw? But as people have said it's beta so I'm happy to wait for it to get better to fly. This aircraft should be easy to fly well, and hard to fly very well. Right now it's just hard to fly well. Why make an aircraft hard to fly when it could be easy? What purpose does it serve a pilot, who over the years have been increasingly provided a lighter workload via technology. PS, I flew one of the SU27 training missions (guns training) last night and the aircraft flew like a dream vs the quick-start missions. Is the physics different, or is this just the difference between an empty vs loaded AC making it so much more unstable? Thanks Dave

For something like DCS then you could probably get away with the cockpit only after about 250m in altitude... and even on the ground maybe the airfield is ample to give you parallax in renderings. I'm not sure how the DCS renderer works, or how it'll work in DCS 2, but I'm guessing with DCS 2 being a ground up renderer with stereo rendering in mind at least a bit, then they'll have optimised it quite a bit for rendering. It's always gonna cost more to do 3D, but I bet there is a lot you can do to cheapen the cost of double rendering if you design with it in mind. I'm not even sure what the new DCS 2 renderer is, is it deferred or forward rendering? Given the potential need for lots of lights/flares and night time explosions and all that stuff, I'd expect deferred, but you never know, given 99.9% of the time you just have a sun as the only light source/shadow caster. Either way, given the cost of a VR system vs say loads of screens and/or cockpit screens on your desk or trackIR etc, then I'd happily buy a VR and a monster GPU as it's still cheaper than a load of screens and a similarly high powered GPU! Dave

Just to clarify a few points. I'm using the SU27 in FC3. If I load up with just the joystick plugged in, then all is great. If I load up with both plugged in, then the controls seem double entered for both controllers, so I need to tidy them up lots (is this a sign of a bug?) Both controllers work fine in the menus. But then in flight neither work ok. All modes that I could find were set to simulation/realistic or equivalents. Ie, loading up instant action, choosing SU27, and diving straight in. Even going into edit controls in-game, they set up and look fine, buttons work etc. Then back into the game and no buttons work. Hmmmm

OK so I have a cheapo Logitech Attack 3 and my Logitech G25 plugged into the PC. Joystick works fine without the G25. But if I keep the G25 plugged in I can use the brake/throttle on combined axis for the rudder, I also have loads of other buttons on the G25 gear box thingy too. This config also works fine in Falcon BMS. But in DCS, even after clearing all duplicate entries for both controllers (why does it do this, and/or not highlight conflicts?), the joystick buttons just don't seem to want to work in-game, neither for the G25 or the Attack 3. In the controls setup screen they work fine. So all axis work fine, but no buttons work. Am I missing something really obvious? DCS sees both controllers, and sees them individually in the controls screen. Everything 'looks' right there, but in game no go on buttons?! Thanks in advance for any help! Dave

Yes I don't generally have a problem with the 'realism' of the flight model, it's the level to which the stability control lets you still enter that feedback loop zone which you'd probably never want to enter into... especially in a negative g case! Stick kickback is a really valid real life way to stop pilots over-doing the inputs without really knowing about it, along with real g-forces. So what do we use in a simulation? I'd say if the real stick kicks back, then filter the I/O for a kickback. Make it happen in the visuals, make it happen to the simulation input. OK there will be an inconsistency between user input and the aircraft input, but I'd prefer that temporarily than have the aircraft dead in the air. But I think if this is a beta then it'll improve. If we simulate the instability super accurately, but don't simulate the stability controls effectively too, then we're left with something that still isn't a complete simulation of the 27. I bet the latter is just as hard, if not harder, than the former, to get right!

Just flying the SU33 and suddenly it does feel quite numb in comparison to the new flight model 27. I can imagine the 27 getting lots better if this is still a beta!

Good point, when 'direct' control is on, the FBW is probably a raw control system rather than a corrective one. So stuff like asymmetric load while doing a cobra could prove quite interesting. Technically though, if the flight model is 'right' for the SU27, an asymmetric load cobra right now should be as real as it gets! I was watching a documentary about the Mig25 yesterday and in early testing when firing missiles, it'd roll upside down and out of control at high altitude due to the sudden CofG movement. They had some weird automated system to apply corrective roll just after release to 'catch' the aircraft. I bet it's quite amazing how complex the FBW systems are these days to make combat aircraft safe to fly :D Dave

Yeah I think flying around the problems are somewhat reasonable to do. Kinda fun sometimes too. A lot of what makes the SU27 great is it's instability, and if you overly control it's pitching then it's losing what makes it great. But there is a point currently where you get so much AOA dialled in, and the feedback loop takes you all the way to a total loss of airspeed and control. Once you hit that point, which is easy to hit without being cautious, then it's too late. I honestly can't see any purpose for that, and if you want it, that is what the stability off switch is for. With stability control on, I'd assume the aircraft wouldn't let you get unstable to the point of being dead in the air. It's not really stability control otherwise, it's anti-death by insta-g control only :D It'd be interesting to get clarification on the current system. I assume currently inputs just go direct to aircraft outputs, so there is no system looking at inputs and adjusting them based on other variables... except the simple controller curves. Or if there is a properly populated limiter system it'd be interesting to know if it's still a work in progress, or was based on real data or something. As for trimming, I'm trying to get it to work. I assume trim to stick on button 4, or whatever it is, lets you set the trim with the joystick once the button is pressed? Or do you just set the level flight with the stick, press the button, and that is set as the trim? Either way it doesn't seem to be working for me right now... I think maybe my G25 pedals are confusing things though... hmmm. Thanks Dave

I just decided to 'upgrade' to FC3 in DCS and I'm finding the SU27 a bit of a pain to fly too. What seems to be missing is what you mention. It's all very well to adjust curves for the controller, but this is processed at all speeds and flight conditions, and that just isn't appropriate. It's entirely reasonable to say that the SU27 probably has a lot of calibration tables that have curve adjustments based on all kinds of inputs, so the controls always feel 'just right' at all times. Curves based on AOA, g force, speed, etc etc. Yes you can fly around it, but why would you? Why wouldn't you design the FBW to do all the hard work? From almost everything I've heard and read about this aircraft, it's meant to be a pleasure to fly. Easy to get the best from it, not hard to get the best from it. Why wouldn't you design the stability control and FBW systems to offer the best of all worlds? Why design a stability system that is leaving you still able to do really wrong things, when you could program it not to? I can see that with the correct approach you can get stunning performance, but why make it so hard to access, needing perfectly balanced inputs, when you have total control over desired input from the pilot, and can figure out from that, the desired output... as if taking away the risks of the unstable design, but adding the benefits of it. Does anyone know how the current DCS/FC3 stability control system works? I've looked in the data folders but can't see anything specific about the stability control vs other aircraft without it. I'd assume it'd use a bunch of calibration tables somewhere? Cheers Dave