The leftward lurch
Results 1 to 20 of 20

Thread: The leftward lurch

  1. #1

    The leftward lurch

    There's been some discussion recently in a couple of threads about the sudden leftward lurch at 40 knots that affects all MSFS aircraft but is most notable in taildraggers.

    Developers have pointed out that it's caused by the way MSFS handles the rudder - it has zero effectiveness until 40 knots, and then suddenly resets to 100 percent.

    But there are another couple of possible issues - the extreme sensitivity of the rudder once it's on, and the excessive weathervaning tendency in any kind of crosswind.

    I was recently pointed to this thread on the MSFS forums, where a user reports on systematic research he did using the wind tunnel feature in MSFS developer mode. Members of the MSFS team seemed to take note.

    I don't know if any of this was addressed in sim update 4 - the aircraft seem a little tamer but the issues haven't yet been fixed.

    Would be interested to hear what others - especially developer friends - have to say after reading the report.
    "Ah, Paula, they are firing at me..."

    -- Saint-Exupery

  2. #2
    Side force is just one of the many parameters the developers have "over-cooked". Any dev who has tried entering real world values for the major specifications for an aeroplane will know what I mean. Engines are another case in point. I do not understand why the figures that appear in final files are so wildly different to RW. No longer are we able to enter the book values as a good starting point. If we do enter them, the outputs have to be dialed right back to make the results anywhere near sensible.

    Side force and rudder authority are poor and need to be fixed. It should not be up to devs (or end-users for that matter) to find the fixes. Altering the figures right down just makes a nonsense of code of the original flight dynamics engine. Yes, it may be a treatment but it is not the cure.

    I for one, will not believe anything I read about "aerodynamics" in this game until we get the basics like a castoring tail-wheel and opening doors.

    And why, pray, does the "new"SDK documentation focus on legacy modeldef entries when all along we have been told that they are no longer recommended?
    If devs had been allowed to use legacy modeldef, you would have far more new add-ons out there by now. Perhaps they have realised that.

    Left kick on takeoff roll is one symptom of a poor code. It needs to be fixed.

  3. #3
    From a user's perspective, the unfortunate thing is that some of the addon publishers have, very understandably, taken matters into their own hands by adjusting their flight models for the MSFS bugs, while others have not. Thus, for example, I have flown a couple of mods recently where the rudder authority has clearly been dialed way down to compensate for the game's over-sensitive rudder. But they seem to assume that I have my pedals on 100% sensitivity and no dead zones, whereas in fact I have dialed my extremity dead zones way up so that I only have 25-30% of stock rudder travel as my own way of compensating for the same thing. When I fly one of the dialed-down addons, combined with my control profile, I have hardly any rudder at all, so I have to remember to load a profile with a more sensitive rudder before I fly those planes. Usually, I remember to do that just before impact with the trees beside the runway. The longer these bugs go unsquashed, the more of a mess things will get, with different levels of compensation for the issues. And on that magical day when the issues are fixed, everyone will have to patch and un-mod their planes to make things the way they should have been on day 1.

    We're learning bad flying technique, as well. Many times on this board and others, I read someone saying, "It's easy to prevent the swing, you just ..." and then describe a way of flying that is totally non-authentic. Well, yes, you can train yourself to beat the game, but you're actually getting worse at flying a decent simulator. If I want to learn and beat a game's quirks, I'll play Portal or something. For simulation -- well, let's just say I'm still spending a lot of quality time with P3Dv4.

    August

  4. #4
    I don't know that much about the physics, but ever since the last update, I've had an easier time taxiing with a crosswind. Something has definitely changed, I just don't know what...
    Thermaltake H570 TG Tower
    X670 Aorus Elite AX motherboard
    AMD Ryzen 9 7900X 12-Core Processor
    NVIDIA GeForce RTX 3070
    NZXT Kraken X cooler
    32GB DDR5 RAM
    750 Watt PS
    Windows 11 Home

  5. #5
    Very well said, August !

    Also particularly:

    The longer these bugs go unsquashed, the more of a mess things will get, with different levels of compensation for the issues. And on that magical day when the issues are fixed, everyone will have to patch and un-mod their planes to make things the way they should have been on day 1.
    We can only come to the conclusion that we're still flying a more or less Alpha here... Tremendous beautiful one though...

    Btw, i remember 'Lurch' as the big tall butler looking like the Frankenstein monster in the 1960's "Adams Family" TV series. Hence we could think about the dreaded 'leftward lurch' as the 'Butler Effect'...

  6. #6
    @Baz - thanks for that - one of the best (and also most depressing) descriptions I've read about the current state of the sim.

    @August - agree. I've been on the receiving end of too many of those discussions: "You really just need to learn to handle a taildragger." As if all those P3D hours didn't count, and as if a Waco is indistinguishable from a P-51. And you're right about the wide variability in the way flight models interact with controllers. My current solution is to keep my hangar as small as possible, and stick with the handful of aircraft that deliver something resembling realistic performance. The JustFlight Arrows, the Working Title Cj4 and the Aerosoft CRJ currently deliver. The last two are Alexander Metzger productions. I'm wondering if he has a hand in the DC-6 - he did the FSX/P3D version and he seems to be able, somehow, to deliver a sense of inertia and momentum in MSFS. Enjoying the Islander, too, though it likes to skate a bit on rollout.

    @Tom - it feels as though there's been some improvement, but with a ways to go yet. At least the worst tendencies have been damped out a bit, though they're still there.

    @Jan - You rang?
    "Ah, Paula, they are firing at me..."

    -- Saint-Exupery

  7. #7
    I can also see it here, the behavior of rudder during takeoff and landing it's very badly done here. P3D is a super realistic sim if we look at it from the MSFS perspective .
    Webmaster of yoyosims.pl.

    Win 10 64, i9 13900 KF, RTX 4090 24Gb, RAM64Gb, SSD M.2 NVMe, Predator XB271HU res.2560x1440 27'' G-sync, Sound Blaster Z + 5.1, TiR5 [MSFS, P3Dv5, DCS, RoF, Condor, IL-2 CoD/BoX] VR fly only: Meta Quest Pro

  8. #8
    Alan, didn't mean to sound too depressing, the game will get fixed, I am certain. But it won't get fixed properly if we don't stop trying to do it for them.

  9. #9
    Hiya,

    Has anyone tried the mentioned settings in this topic:
    https://a2asimulations.com/forum/vie...p?f=23&t=72535

    Still practising so myself not sure it works 100%, however have the feeling it helps a bit.

    Marcel

  10. #10
    I respect Dudley as a knower of things about planes, and I used to use a rudder profile something like the one he posted. But, personally, I find it more realistic to keep my control setups as linear as possible, and his is very nonlinear. Curves are admittedly the only way that you can have a reasonable response at low to mid deflections while still having access to full authority at max deflection. My own compromise is to give up that maximum authority, since I only need it for 2 seconds on takeoff and then only because of a game bug, so that I can have a linear profile that feels more natural to me when flying. I have learned not to be concerned about ugly takeoffs and landing rollouts in MSFS. Those things are just bugged up in the game, and until they're fixed, the only purpose of a takeoff is to get flying, not to enjoy a realistic takeoff as such.

    One workaround that has been successful for me, but that requires you to have an extra axis handy on your controller, is to map a second control to the rudder, but at full authority. Maybe you aren't using your stick twist axis for anything now that you have rudder pedals, or, if you have an X56 like I do, your controller is encrusted with more buttons and axes than you know what to do with. I have my pedals set up linear but with 60% extremity dead zones for normal flying, but a little joystick axis on my throttle handle also mapped to the rudder with no dead zones. I use the mini-stick on takeoff when I briefly need full rudder authority to fight the lurch, and also for sharp turns on the taxiway. Mapping two different controls to one surface is usually a no-no, but I find that they don't interfere with each other.

    August
    Last edited by K5083; June 9th, 2021 at 05:07.

  11. #11
    SOH-CM-2021 warchild's Avatar
    Join Date
    Feb 2008
    Location
    Bend, Oregon
    Age
    72
    Posts
    5,466
    Blog Entries
    3
    Quote Originally Posted by mgr View Post
    Hiya,

    Has anyone tried the mentioned settings in this topic:
    https://a2asimulations.com/forum/vie...p?f=23&t=72535

    Still practising so myself not sure it works 100%, however have the feeling it helps a bit.

    Marcel

    It;'s not your settings Marcel.. On purpose or accidently there were some changes made in SU4 that threw us all for a loop.. My team was working on a floatplane before SU4 hit and our sister team was working on another floatplane, Now we're having to focus completely on what Asobo did to change the water and rudders.. It's by far and away, not your settings sir.. Its a fundamental change in the sim that Asobo may or may not know anything about..

  12. #12
    If it helps anyone struggling with ground handling in cross-winds, here's a formula I use on problem aircraft:

    Open your chosen aircraft’s flight_model.cfg file with Notepad and scroll down to the [AERODYNAMICS] section.
    Look for the side_force group of entries, there will be four in total but we're only editing three of them:

    side_force_slip_angle = -2.12726 <-- figures are for example only, they vary for each aircraft
    side_force_roll_rate = 0.25450
    side_force_yaw_rate = 3.09566

    Copy the value for the first entry, e.g. -2.12726, and paste it into Calculator.

    Now divide by 4. Copy the result and paste it back into your flight_model.cfg like so:
    side_force_slip_angle = -0.531815 ; -2.12726

    The ; between the two values is there to retain the original value while making it inactive.
    Repeat this process for the other two values.

    Now look for the yaw_moment group of entries, there will be eight in total but we’re only editing two of them:

    yaw_moment_yaw_damping = -16.27597
    yaw_moment_delta_rudder = 0.64763

    Copy and paste the yaw_damping value into Calculator and divide by 2:
    yaw_moment_yaw_damping = -8.137985 ; -16.27597

    Copy and paste the delta_rudder value into Calculator and multiply by 2:
    yaw_moment_delta_rudder = 1.29526 ; 0.64763

    This formula has improved cross-wind ground handling on every aircraft I've applied it to so far, and improved adverse yaw and sideslip in the air too.

  13. #13
    Quote Originally Posted by mgr View Post
    Hiya,

    Has anyone tried the mentioned settings in this topic:
    https://a2asimulations.com/forum/vie...p?f=23&t=72535

    Still practising so myself not sure it works 100%, however have the feeling it helps a bit.

    Marcel
    The problem with those settings - with anyone's recommended settings, really - is that the results are highly hardware-dependent. With short-throw joysticks like the Thrustmaster Warthog and less-resistant pedals like the Thrustmaster TPM, I've had to reduce sensitivity to somewhere in the -10% to -30% range (-30% for the rudders). Users of shorter-travel hardware apparently have to go even lower. I'm currently using a Virtual Fly setup, including a full-travel yoke and very stiff pedals. Those respond better at a linear setting - but better still if I run them through the proprietary Virtual Fly software, which seems to add refinement in some sort of black-box way that I don't understand. I assume FSUIPC or Axis and Ohs would have a similar impact, though I haven't tried them.

    I respect everybody who's trying to come up with settings and to tweak .cfg files, but it's a bit ridiculous that we have to keep doing this to resolve basic problems that seem to linger through multiple revisions, or get pushed sideways by updates but not really solved. And I'm wary of global setting changes given the very different way individual aircraft flight dynamics seem to be tuned.

    @August - that's interesting about using a second rudder axis for full travel. Was going to say it seems like a kludge, but I guess it's more like a wartime field mod. I've got a couple of buttons on a switch panel that might work - I'll give it a try.
    "Ah, Paula, they are firing at me..."

    -- Saint-Exupery

  14. #14
    Quote Originally Posted by Alan_A View Post
    I respect everybody who's trying to come up with settings and to tweak .cfg files, but it's a bit ridiculous that we have to keep doing this to resolve basic problems that seem to linger through multiple revisions, or get pushed sideways by updates but not really solved. And I'm wary of global setting changes given the very different way individual aircraft flight dynamics seem to be tuned.
    I agree with you. Tiger's flight model tweaks are interesting, but I'm not all that inclined to try them. It's trying to fix one screwed-up thing by screwing up something else in a calculated way, and I'd be concerned about unwanted broader consequences. For example the yaw_moment settings seem designed to boost general yaw instability and rudder effect, neither of which I think MSFS needs.

    @August - that's interesting about using a second rudder axis for full travel. Was going to say it seems like a kludge, but I guess it's more like a wartime field mod. I've got a couple of buttons on a switch panel that might work - I'll give it a try.
    You may find that buttons don't work as well as an axis. I tried it myself. If you program a pair of buttons to be left and right rudder, a press just increases the rudder deflection by a small increment, and if you hold it down, you get that stupid Windows keypress sequence of a small increment, then a pause, then a rapidly accelerating series of increments. It will probably take you about a second to get up to full rudder deflection, far too slow to catch the lurch. With an axis - and the X56's little thumb joysticks are perfect for this - you get full analog control and you can have full rudder instantly if you want it. It is totally a kludge, and when the lurch bug is fixed, I'll happily un-map that control and do something else with it.

    August

  15. #15
    Quote Originally Posted by K5083 View Post
    For example the yaw_moment settings seem designed to boost general yaw instability and rudder effect, neither of which I think MSFS needs.
    The yaw_moment settings reduce the overdone weathervane effect on some aircraft, and increase the rudder effect to give more control on the ground in particular. All of these edits depend on correct controller sensitivity settings of course.

  16. #16
    For a sim that is far too often disparaged as being designed for Xbox users (which has been proven and shown to be false, time and time again), a good set of rudder pedals goes a long way with this sim.

  17. #17
    Although I have not personally disparaged the sim as designed for Xbox users, that is a perfectly valid and reasonable opinion held by some, and has not been proven false, nor can it be, being an opinion.

    August

  18. #18
    All of the statements and actions made by Asobo throughout development, and to this day, in addition to the heavy hitters/companies that are backing it, prove that it is meant to be a great flight simulator. More work than ever continues to be invested in that direction. I just find it ironic that, although it has been so-often referred to as a sim designed for Xbox gamers, it really cannot be enjoyed without proper flight sim controllers. For anyone only using a twist grip for rudder control is going to have a heck of a time, let alone a rocker switch or trigger keys on an Xbox controller. It all becomes quite 'easy' with good controllers, especially rudder pedals. I also don't understand the opposition I see sometimes to having to reduce the sensitivity settings for the controllers, when one of the most respected flight sims, DCS, has required exactly the same for many years, and it is just acknowledged as a simple requisite within that community.

  19. #19
    I think any flight sim, no matter how crude, cannot be enjoyed properly without flight sim controllers - at least, by me. The kids can do amazing things with those game pads, and I will not be surprised to hear, once MSFS gains some penetration into the consoles, that a lot of gamers are very happy flying it on game pads. On the other hand, I own "Birds of Steel" and "IL2 Sturmovik: Birds of Prey" for Xbox 360, and they are definitely not great flight sims, but I might still be enjoying them if I had proper flight controllers for that unit. So the need for flight controllers doesn't correspond with whether the flight sim is much good, for me; it's just a question of who the player is. As for Asobo claiming they want to make a great flight sim, OK sure, but I don't take that at face value, or even assume that Asobo would know a great flight sim if they saw one. What I do see is a game that has great potential but, almost a year after release, is still struggling to meet FSX standards in every aspect except the graphics. And, to bring this back on-topic, that is the issue we are talking about here. Can the community ever teach Asobo what a good flight sim is, or is it on us to try to fix issues like the leftward lurch with a patchwork of makeshift solutions?

    August

  20. #20
    Quote Originally Posted by Bomber_12th View Post
    It all becomes quite 'easy' with good controllers, especially rudder pedals.
    I'd say it becomes easier, but not necessarily easy. Even with the Virtual Fly hardware, there's a fair amount of tuning required, and even after that's done, it never feels quite right. Roll axis is fine, but rudder never seems to strike the right balance, and pitch is overly twitchy - which makes me suspect that there's something wrong with the modeling of airflow over the tail, beyond the rudder sensitivity and side force issues we've been talking about. Reducing sensitivity can make things more stable, but at the cost of control authority in sudden sharp movements on the rare occasions when those are required. I don't recall things being similarly unsettled either in FSX/P3D or X-Plane for that matter. Hardware helps but can't totally mask the "backstage" problems.
    "Ah, Paula, they are firing at me..."

    -- Saint-Exupery

Members who have read this thread: 0

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •