X-Plane 11.10 went final Thursday night – you’ll be notified to update. Jennifer has a good write-up of the high level features here, or you can read the release notes.

11.10 was a big patch for us – new airports, new autogen, new lego bricks, the G1000, joystick profiles, London landmarks, and significant engine enhancements to both the physics and rendering engines.

We may do an 11.11 patch next week – we have a few bug fixes that didn’t make the RC. Jennifer gets a lot of bug reports over the weekend, so we’ll evaluate Monday.

VR Is Coming!

The rendering engine work in 11.10 gets the sim “VR-ready”, and Chris has been working on the actual VR features of the app in parallel to the 11.10 release.

The next major patch of X-Plane will be 11.20; it will feature native VR support for the Oculus Rift and Vive on Windows, and the beta is coming very soon.

About Ben Supnik

Ben is a software engineer who works on X-Plane; he spends most of his days drinking coffee and swearing at the computer -- sometimes at the same time.

188 comments on “X-Plane 11.10 Is Here!

  1. Great news, but…
    waiting for 11.10 I felt the dynamic of X-Plane development is gone (I remember Austin’s hacking nights with tecno music and some exciting news about every day).
    OK, Austin got older and he has a family to care for, also 😉
    While I understand that every release, idependent of the amount of changes, is some fixed work to do, so you don’t wat to do that too frequently.
    I also understand that developing bug fixes and features independently causes some extra work and hassle when integrating thos into the next version, but OTOH if one such features takes much longer than expected, it’s easier to postpone it until the lext release (IMHO).
    What I’m trying to say: I’d prefer bug fixes to arrive sooner, being more patient for new features.

    1. There are some bug fixes that we could have (in hindsight) put in an 11.06 that waited due to disorganization on my part — the 11.05 vs 11.10 decision for bugs came right out of 11.00 final – it was a hectic time! But a lot of 11.10 fixes weren’t done until…11.10.

    2. I’m going to disagree here that Austin not producing development as much as in the past. X-Plane11 has had a lot of performance worked on and delivered, in fact I can think of no version in the past in that has had as much changed. The point maybe that the scope and the size is far more different than in the earlier X-Plane days so it doesn’t stand out as much with all the other huge features delivered.

  2. Is this (“11.05”) a typo in the release notes?: “New airports with 3-D scenery since the 11.05 release: 534”

    1. That’s not a typo – over 500 new 3-d airports from 11.05 to 11.10. We put up about 500 new 3-d airports on every patch — the pace of community work has been amazing.

      Now…you’ll see 75 new airports and over 600 updates. If I understand Tyler’s stats, the 75 airports are totally new – e.g. we didn’t have them – hence the predominance of non-real-world airport IDs (Xnnnn) – these are airports where users requested a new airport ID from us.

      But the 600 updates can include updates that took an existing 2-d airport (and we have a LOT of existing 2-d airports – tens of thousands left over from v9 and before) and upgrade them to 3-d.

      So…most user work is updating an existing airport, not creating a new one, but a lot of that update work is updating from 2-d to 3-d.

      Finally, note that the 3-d label just means there is at least ONE “3-d” thing on the airport – the stat is automated, so an OBJ counts as 3-d even if it’s a pavement marking. This is not a stat on how detailed the airports are – some are amazingly detailed and some just have a few small annotations.

      But you can step back and just look at all of the great airports flowing in and see that the community is doing a tremendous job.

  3. When it works, this this update offers some great glimpses into the future of XP; but is two steps forward, and one step back for AMD users. Loading times are much slower, and so far too many hangs and crashes, and rendering glitches. I know that you warned that things might not great for Radeonistas, but until you’ve out the bugs, it’s back to 11.05.

    1. Hi Dave,

      There shouldn’t be any change in loading time, AMD or otherwise. Our perf tests indicated neutral performance for AMD.

      Note that we perf test the old and new code on the SAME art assets. If you go to a gateway airport that got a big upgrade in 11.10, for example, you’ll see a “fps hit”…because there’s more stuff there.

      1. Hi Ben, well, to be specific, I can report that several AMD users, myself includes, can only get XP 11.10 to load by deleting shadercache prior to starting the sim. Not sure if it’s the new Radion 17.11.4. driver, but will work on it today to see if I can isolate the cause further

        1. Me too..I’ve had to delete the shadercache to get the Sim running. Really hope in some improvements for us AMD users.

        2. Hi David;
          This may be of some help.
          I have a Rx 460. I purchased the xp reality gtn750. Upon installation it would not work. I figured out it was the radeon drivers. I think it was one of the 17’s that I had loaded at the time. I located an archive and downloaded version 16.12.1. Gtn worked great. My xp version was 11.05. I have tried most of the 11.1 betas. All crashed after a couple of starts. So I gave up on 11.10 and stayed with 11.05. When the current official release showed up I tried it. Low and behold works great, no crashes. I have been using 11.01 for several days now without issues. You might want to try 16.12.1.

          1. Hi Brad,

            Thanks very much for the info. I’ve at least managed to get things going in 11.10 by a flush of the shadercache before each flight; but unless that happens, the sim hangs and crashes.
            I”ll look into that 16.12.1 driver, but, to be honest, I’m a bit unsure about installing, because (i) it precedes my RX580, and (ii) I’m getting a bit performance / smoothness boost with the FSL A320 in P3Dv4 with this latest Radeon driver, the 17.11.4. At some point I might have to make a choice, though ….

    2. On the .org forum there are many AMD users who reported crashing at load time. Most of them said, XP started once after the 11.10 update and then continuously crashed when loading a flight.

      AFAIK, all of them could fix it by deleting the contents of the Output/shadercache folder.

      Maybe this also helps with glitches and performance?

      1. Hi Y’all,

        Interesting. For what it’s worth, I _do_ see a huge surge in AMD-related crashes on the crash reporter. Since the crash is in the driver, we have _no_ visibility into what went wrong (we literally cannot see what the sim is doing, just that the driver crashed).

        Does running with

        X-Plane.exe –no_shadercache

        On the command line fix the problem?

        cheers
        ben

        1. Hi Ben, I also had crashes on startup until I deleted the shader cache, it’s been smooth sailing since.

          Not sure if it was after driver update or what, I haven’t upgraded drivers since just in case.

          I’ve an rx480 8gb.

        2. Hi Ben,

          I am not sure it is driver related. I did not update my drivers, 11.05 was running stable and framerates where acceptable. 11.10 keeps crashing, When I delete the shader cache. I can run X-plane but framerates are lower then before. I do not see it being a driver issue, maybe X-Plane asks something incorrect from the driver.

          1. Hi Ben, just in case it could come in handy…my 2 cents:
            I can confirm that the update went fine from 11.05 to 11.10..no issue at all. I got in trouble in next attempts to start the sim…no way.
            Right now it seems to be fixed after deleted the shadercache folder.
            I regularly update my AMD card driver to latest version.
            Sent several bugs report on this

  4. Talking about bugs…after 11.10 update I get lots of crashes as soon as started the Sim..and I have a plain vanilla installation..just IVAO and X-Camera plugin..removed to avoid any issue…I even deleted all the preferences files…no wau to fix it.

    1. I have the same issue, even after reinstall without any addons.
      I filled a bug report, and almost everytime it crashes it offers to send a bug report.

      Is there a way to downgrade to previous stable version?

      1. I have the same problem, after the update X-Plane 11.10
        so not start!
        I have send the crash report.

  5. In Jennifer’s writeup: are “Detailed Custom Scenery for KPHX & KBOS” lego brick airports, or are there custom objs buried in there?

  6. please bring back chemtrails and dirt/dust clouds on ground!!!

    realistic chemtrails were introduced and looked great in MS FS 2004 more than 10 years ago, i can’t believe we are being stripped out of visual and immersive effects in such an ‘advanced’ flight simulator the X plane v11 is prentending to be

      1. In all seriousness, the particle system sounded very far along back when you guys did the live stream about half a year ago.

        No news on it at all? The contrails really are very jarring.

        1. We made some progress – Max did some work with it – but we found the existing tech for contrails still isn’t that great. We should probably pack and ship what we have so far..

          1. Visually, XP contrails are not that bad, their only problem is their dark smoky color whereas they should be white. Even using a white texture for them doesn’t help. Couldn’t you make a quick fix just for this?

    1. Please no Chemtrails in X-plane. . But please don’t poison our beautiful X-plane world.. Contrails yes

      1. I’ve found some very interesting videos on YT that clearly prove without doubt, that for 11.20 LR secretly plan to mix substances to the fuel that make us users will-less and supple (to stop all that complaining about fps, ATC and weather engine).
        DO NOT use the fuel provided by the default fuel trucks anymore!!!

  7. Hi Ben, it’s been some time, can you comment now on on when the Chicago landmarks are coming, and or Mac VR support? thanks

    1. Not sure on Chicago – we’ll do another landmark pack in 11.20, then in 11.30.

      Mac VR support will not come until after we’ve ported to Apples’ Metal API. The Mac OpenGL driver (even the new ‘core’ one) doesn’t support multiple viewports from vertex shaders via hardware instancing, which is mandatory for us to run a stereo render without murdering frame-rate.

      A port to Metal is in the v11 road map, but we’re not going to hold up on Windows VR to wait for that. So Mac VR may happen sometime in 2018. (I can’t promise anything, as we have -never- seen X-Plane run on a Mac with a VR headset, even in the lab.)

      1. Hi Ben,

        I know nothing beats raw hardware specs – like me being bottlenecked by my 2 GB VRAM– but for us Mac users, do you anticipate the port to Metal to make a significant performance improvement? Thanks.

  8. It may sound silly … but one of the “killer features” of 11.10 for me is the lack of ground shaking :-).

    I am very exited for VR. X-Plane was in fact the major reason I ordered a Rift recently. As far as I can remember you stated that right now the x-pad is “only” able to display the map. It would be very cool if it could do one or more of the following:

    – show an integrated web browser
    – run an integrated android emulator (so we could play X-Plane mobile while flying X-Plane desktop – just kidding ;-))
    – display the current computer monitor content (similar to what Bigscreen VR / //bigscreenvr.com does)

    I know that any of those items would probably mean a tremendous amount of work to get it running adequately but still – it would be very cool ;-).

    Another entirely unrelated thing that would be handy to have but has nothing to do with VR would be a mute button in the menu bar (maybe next to the pause button).

    Nevertheless – great work – and I am very much looking forward to that VR update.

  9. “VR is coming…the beta is coming very soon!” <– My Oculus will be here Tuesday. I'll be waiting (im)patiently 🙂

    1. I can’t express how excited I am, I think I might set a screenshot of Ben’s text saying “VR is coming…the beta is coming very soon!” as my windows wallpaper! =D

  10. What are your recommendations for a VR setup? I saw that you can manipulate controls via the touch controllers. I was thinking a physical yoke and rudders and using 1 touch controller to manipulate the other controls. Thanks-

    1. We can support both ‘all virtual’ (two controllers, no throttles or yoke) or part virtual (one controller that you keep all the time and a yoke, or one controller and a yoke and throttles). I’m not sure how hard it will be to re-find your physical hardware if you let go of them, but finding the controller is easy- you can see it sitting there in VR space. You can use rudder pedals – they’re optional just like without VR.

      I would suggest two controllers – even if you use only one some of the time, there are modes of flight (using autopilot, starting the plane) where two virtual hands is a win.

        1. We’re not looking at them right now – we support the standard controllers for the Rift and Vive, which provide 6-d location and a bunch of buttons. We may look at gloves at some point but we gotta walk before we run, and hit the hardware that’s going to be ubiquitous first.

          1. Makes sense, I’ll adopt gloves in a heartbeat if they get X-plane support and don’t interfere with being able to feel my TPM etc.

  11. I said back in September when VR was promised before the end of the year that we would most likely not get it till next year and it looks more and more likely that will be the case.
    There is only 2 weeks remaining before the xmas holidays can’t see it happening before then.
    Can’t enjoy X-plane in 2D and it is meant to be enjoyed in 3D and Flyinside is crap, so haven’t flown now for months waiting for native support.
    It will be interesting to see if Vulkan makes 2018 as promised or it blows out to 2019.

    1. Wait a few days – Ben stated that 11.20 has been worked on in parallel to 11.10… I can really imagine that the 11.20 beta will become a LR style (“very soon”) christmas present if they make to release 11.15 next week 🙂

    2. Ben Said about 3 weeks ago that it is still coming out this year. They can still release it in beta easily this year. I do have to agree with you on Flyinside, although a cool product it is garbage and extremely bulky. I’m sure Laminar’s VR will be 200x better.

    3. I think you are being a little hard on Flyinside. I don’t work for them, just a customer, but am happily using XP11 and FI. It is a good VR experience and with reasonable performance (I am using Ortho4XP scenery at ZL17/18). P3D v4/4.1 has native VR but it does not offer as good or practical an experience as FI and P3D v4…. yet. Sure, I want native XP11 VR to be good (and would hope better than FI). However, we might have to go through the learning pains. Also, remember that FI has to keep updating after each XP11 release just like any other developer, so in one sense they are always playing catch-up. If you have dropped FI and not used it for a while, give it another go. You might surprised.

      1. I haven’t looked at FI for awhile it’s true, but it is still in beta for XP and there has been only one upgrade of FI and I find it difficult to believe it is any better.
        Maybe the improvements in XP have made it a little better.
        I gave up on FI after weeks of testing not just a whim.
        It is a single threaded application only using one core or thread and the remaining cores or threads are at idle.
        I tried every conceivable combination of settings, I tried every tweak, searched every forum, watched every you tube video I could find looking to get it bearable to play without it looking like you were trying to drive your car without your windscreen wipers in the rain.
        It took me 2 weeks of testing before finally looking at the numbers in MSI Afterburner with it’s graphs to see what was going on.
        Some people may simply be able to put up with the limitations of FI and the poor visuals like Wayne on Luna’s world seems to think it is ok but for me I tried all I could and it was abysmal.

        BTW Wayne on Luna’s world just released a vid on Youtube and he also believes as I do that VR won’t be here this year it will be a 2018 release despite all the promises and hype.

  12. I presume Linux VR will be with the Vive and Vulkan? Should give me time to upgrade my hardware first.

    1. I don’t know if/when we will have Linux integration – basically if SteamVR works on Linux, we’ll ship it, if it doesn’t, we won’t. Since I don’t know what back-end texture formats SteamVR supports on Linux (Vulkan image objects, OpenGL texture objects, et.c) I don’t know if this will work now or requires a Vulkan port first.

      To be blunt, it’s not a priority…the total population of Linux users for X-Plane is tiny, and the number of headset users is low too; so the intersection is small.

      If Linux users want to try to get this to work, we’re happy to leave the functionality in — that is, we haven’t used #if to kill the functionality, but I can’t send Chris off on a multi-week odyssey to try to figure out why some distros work and some don’t and the driver has to be just right, etc.

      1. Hi ben,

        Not a linux user but I guess that if you would have to choose one distro for that, it would probably be Solus. For what I know they have the best Steam integration of them all…

      2. SteamVR works on Linux today (e.g. Serious SAM VR) but it still has some rough edges which would be sorted out with time. SteamVR on Linux _requires_ Vulkan, it cannot work with OpenGL.

        I saw you comment above about Mac VR support. If XPlane11 is ported to Vulkan, why do you need to port it to Apple Metal AP and Windows DX12 for that matter? Isn’t the purpose of Vulkan is cross-platform support (including VR).

        1. Our porting targets are Vulkan for Windows, Linux and Android, and Metal for iOS and Mac OS X. Apple doesn’t provide a Vulkan stack and based on their investment in their own API, we don’t expect them to.

          In terms of APIs and porting:
          – On Windows, we can and do use OpenGL with VR now – we can get stereo hw rendering and use OpenGL to render to the compositor.
          – On Linux, apparently we have to port Vulkan Metal first – no GL SteamVR integration. I’m taking your word on this, I haven’t investigated for myself.
          – On OS SX, we have to port to port to Metal first. While we can talk to the compositor in GL or Metal, we can’t do hardware stereo rendering — not even Apple’s ‘core’ drivers support it. Without stereo rendering, the GL-based framerate is unusably slow on even the fastest hardware.

          We have no plans to port to DX12 – there’s no need since we can get Vulkan there.

  13. What does it mean when X-Plane is utilizing only 50% of the CPU and 50% of the GPU? What is causing the bottleneck?

    For background, my system has a 6-Core i5-8400 and runs stable at ~3.80 GHz per core at max load. Resource Monitor shows each core being ~50% utilized with X-Plane.
    Task Manager shows my GTX 1070 also being ~50% utilized.
    I also have 16 GB of 2666 MHz RAM and an NVMe SSD with 3.2 GB/s reads.
    This is with vanilla X-Plane 11.10r3 on a clean Windows 10 1709 install.
    I have ruled out any other potential causes such as not enough wattage, heat throttling, old drivers, VRAM overuse, third-party plugins, etc.

    Also interesting is that with exterior views (free camera) the CPU/GPU stays at 50%/50% but in interior views (virtual cockpit) it changes to 50%/100%. All this is with HDR+SSAO, Global Shadows on, Dynamic Reflections off.

    1. I’m surprised about the CPU utilization – it probably means the OS is ‘migrating’ the rendering thread across cores (utilizing two cores at 50%) instead of using one at 100% and some others partially. Either way it’s not surprising – not only do we not have enough rendering work to saturate background threads during regular frame render, but if we did you’d see frame rate fall apart when we went to do background loading. So we want to load background threads more in the future, but NOT at 100% – we need to leave some reserve.

      Re: the GPU, lots of things can cause non-100% loading:
      – Not having enough rendering work.
      – Non-ideal dispatch (E.g. small triangles) – this is actually pretty pervasive on the bigger GPUs – they have thousands of compute units, and restrictions on how they can dispatch work to them, so getting them to saturate is NOT trivial.
      – Bottlenecks on memory bandwidth on the GPU, stalling the GPU cores. ROP bottlenecks are sort of a form of this.
      – Bubbles in the pipeline due to how we dispatch.

      1. Hi Ben,
        For me (and some others), SSAO kills FPS in internal view.
        Example : In the default 737, I look at the levers and enable SSAO with DRE
        – FPS instantly drop from a stable 45 to a stable 11 !!!
        – GPU time increases from 0.022 to 0.089 and GPU use increases from 87% to 100%
        – CPU time doesn’t change
        If I disable SSAO, FPS come back.
        Is this normal ?

        I have a 6700K with a GTX1080 and 32GB of RAM.
        I did a bug report about a year ago.

      2. I can confirm this observation of good CPU utilization:

        On i5 7300HQ (2,5 / 3,5 GHz), GTX1050ti, 8GB 2400 DDR4, Win10 I mostly have CPU usage (total usage!) of 60% to 80% by XP (no other applications running). The load is nearly evenly balanced between the 4 cores.

        But I think it needs fullscreen and gaming mode activated.

        (BTW XP11.10/11.05 run like hell on this mid-range(?) Acer gaming laptop, even with most graphics sliders far to the right.)

      3. you guys need to take a look at water reflections “sim/private/controls/caps/use_reflective_water” they seem to be MAJOR source of slow down by disabling them i see a 20% improvement in frame rate EVEN IN AREAS WITH NO WATER

  14. I hope in the next beta some of the smaller areas could also be addressed. The translucent power lines bug has been on the fix list now for years but it is still there, it shows up badly on third party helicopter and propeller blades as well? better traffic management would be nice as well (I know this is a problem), so looking after some of the small stuff and it can make a far better overall improvement to everything else… or are they in the “too hard basket”.

  15. thanks for all the hard work on new airports guys, any chance we can get airports for Southern Africa sometime? FHVA, FHBU, FVFA, FAOR and FIMP+FMEE etc?

    1. Motivate the South African community to start working on 3D airports. Not that difficult to do, all the tools are there and 100% freeware.

      1. Dashing through the sky
        In a hundreds horses plane
        O’er the clouds we go
        Laughing all the way
        Bells and warnings ring
        My HMD’s so bright
        What fun this is that VR brings
        to my flightsim game tonight!

  16. Hey Ben – Looking forward to trying out 11.10! Now that your done with 11.10 and with 11.20 coming out “very soon”, is your focus now shifting to Vulkan?

        1. Absolutely. With Vulkan, many cross-platform woes are at least theoretically less of a headache.

    1. No. It will really depend on (1) how mature or broken actual Linux support is and (2) whether the community is able to do some of the leg-work. We’re not going to intentionally sink Linux VR, but we can’t go down a rabbit-hole.

      1. Sound and fair answer. Many Linux users are developers (myself included) and are willing to help Laminar along with way to get this done.

        Linux VR today works (few games published) but has rough edges and requires very recent NVidia drivers + Vulkan. I bet by Q1 or Q2 2018 you will see a more mature support from Steam.

        1. Right – the Steam API provides a number of ‘surface’ options to the compositor: GL texture, Mac IO Surface, Vulkan texture, DX12 texture, etc. But only some of those are actually wired up to do something, e.g. you can’t use an IOSurface anywhere but on OS X. So if GL surfaces “just worked” our code would work – if the compositor can’t handle them…we have to wait.

  17. What can us AMD users expect from VR performance? X-plane perfomance vs nVidia is pretty poor for us (on my Vega 64 the clocks aren’t even kicking in to max, and if I force the clocks GPU utilisation drops), would be a shame if we can’t enjoy VR to the same level.

    1. I don’t know. It’s too soon to make a call on hardware, but what I can say is that, as of now, using the latest code for AMD (and the code does run without crashes in our VR code), I see 25% better FPS on my 1070 than on my 580, and this margin is the same for both VR and non-VR in _CPU_ limited cases. In other words, the same “driver speed gap” exists in VR as in the regular sim.

      We had hoped that modernization for VR would narrow the margin, but it has not; both AMD and NV got faster, and thus NV is still 25% quicker. Our new hope is that Vulkan will narrow the gap.

      Note that I have an old i5 4600 – there’s a good chance this will be considered below minima for VR – and thus my machine is VERY sensitive to CPU speed. Chris and Sidney have i5 or i7 7600-like things that are quite a bit zippier.

      1. Well, 1070 is a fair bit faster than a 580, possibly even up to 25% faster. Shame to hear the gap hasn’t narrowed though, but as you say: that CPU could well play part in that.

        I take it this is a different AMD rendering path than what is available currently in 11.10r3, and we’ll get that same rendering path in the next beta? And can you say anything about performance of current 11.10r3 vs that new code? It would be really nice to see some of that low GPU utilization go away and performance go up, I’ve been considering filing my findings as a bug report, but it would be a sizable amount of work with loads of pictures and such, and not sure how useful it would really be (as you’re probably well aware).

        1. Let me be clear: the 25% difference in frame time we see between NV and AMD cards is _entirely_ due to CPU side driver efficiency. (This may partly be induced by how the driver _has_ to talk to its hardware, or it might just be the driver code.) I run tests to look at CPU efficiency in a minimal 1280×720 non-MSAA window, and in this condition the 1070 and 580 are both totally bored as GPUs.

          The GPU’s actual hardware performance matters once you start to push more pixels, e.g. 2k, 4k, FSAA, HDR, etc. But I bring this up because today’s HMDs are not particularly high res, so HMD fill rate is a lot less important than CPU time. When HMDs jump res, that may change, and having a real 1080 and not a 1070 may matter.

  18. The new objects for airports are amazing, I still think that few more generic GA hangars or GA hangar facade would help greatly as that is always an issue when making GA airports, there never seems to be any appropriate hangar, but still the new objects are indeed really useful. However I can’t seem to be able to access the new autogen buildings in WED, was that intended? I can only see mid and suburban for europe, not the new high and industrial. It would be a shame if that were the case as they are really nice and would make creating an authentic and good looking sceneries of cities really easy.

        1. What is the drop-down set to above the libraries?
          Make sure its not ‘new scenery only’ (or similar).

  19. Hi Ben,
    Loving your work. Thank you so much.

    When will the new API Vulcan implementation take place?

    Thank you for your time.

      1. Will the Vulcan work be released in stages – like, first a move to Vulcan, followed by optimisation releases.
        I expect many people will think there will be all the magic on the first release, but as far as I understand, Vulcan allows new approaches to interfacing to the graphics card – it in itself doesn’t really optimise anything?

        1. Yes, but no. We will surely optimize even more after we release – we never stop tuning the code – but we have to get a HUGE amount of work done before we can show Vulkan at all.

          Vulkan allows new approaches, particularly in the multi-core area – writing a multi-core GL renderer is possible (see all the AZDO talks) but miserably difficult – the GL threading model is exactly not what you’d want.

          Vulkan also _requires_ new approaches. In OpenGL you can just mix and match pieces of your graphics pipelines willy-nilly, and the driver will, mid-frame, go “oh, wait, you want that with onions, not garlic, let me re-cook your meal” and build a whole new pipeline. This is lousy for performance in OpenGL, so Vulkan is fast in that the slow thing doesn’t exist.

          We could do a faster Vulkan port by simply implementing OpenGL on top of Vulkan as a shim layer and running X-Plane, emulating all of the goofy things that GL drivers have to do. We are NOT doing that – doing so would result in an X-Plane that runs at a worse speed than it does now — we’d have all the slow-down from GL non-optimal abstractions, and our code wouldn’t have been tuned by a team of 100+ engineers.

          Instead we are reshaping X-plane to talk to OpenGL the way a modern card wants to hear – when this is done, Vulkan and Metal will be relatively painless drop-ins. So the work is incrementally released in that X-Plane’s use of GL is becoming “all modern”, but the actual Vulkan switch has to be hit at the end.

          These ‘modernizations’ are usually good for FPS in their own right. In 11.10 and 11.20 we’re moving X-Plane’s material model from loose constants to constant buffers (dynamically streamed in 11.10, static in 11.20) and it’s good for fps, as one would expect – it’s an abstraction that’s better for X-Plane and better for the driver. It will be better for Vulkan too.

  20. I´m one of those VR waiting people. When new news of X-Plane comes out, I don´t bother reading whats new regarding propwash, new cockpit Instruments, engines etc, etc (all important too, I know…), but scan the report for the mention of the 2 holy alphabetic characters: “VR”! My Vive and I just cannot wait… :-))
    (But if the holy 2 don´t appear… well then I do go back and read whats been going on under the hood elsewhere…. Life isn´t only virtual,… but it is so damm addictive! :-))…).

  21. It would be fantastic if VR were a toggle and we could change from VR to flat screen without having to load a new flight and scenery. That way on long flights we could get out of the headset for a while.

  22. I’m very happy to see the g1000 finally make it !
    What could be a solution in a home-cockpit context to have the g1000 running on a different machine than the X-plane main application?
    Running another (networked) copy of x-plane and then poping out the window from this second instance…. or is it somehow conceivable to see a network enabled stand-alone Version of the G1000 (and / or the 430/530 devices )…..?

    1. Use one machine and mulitple monitors to run the G1000 off a second display on the master machine.

      We may have networked support for this at some point but it’s totally overkill – you can easily get a graphics card with lots of video outputs, and external monitor GPS devices don’t hit FPS.

  23. Hello Ben,
    I want to say thank you for this blog updates, technical explanations and your answers to the comments. I want to say thank you for your time.

    I usually fly over some Islands and coast, and always i see the ocean. I see that the sea always has the same color, including the coast too. So I was wondering if there is any plan for implement barometryc and have realistic sea colors deppending on the depth and light intensity.
    I think X-Plane sea color system can be easily improved and would give a lot realism without a lot of effort.

    ¿Is there any plan regarding this?

    Thanks and greetings.

    PS: Excuse me my English.

      1. Bathymetric data would be nice and I guess there is no need to have the full coverage, just what’s near coastlines down to 1000 ft or so and than level flat for the rest of the “sea” world…

        And if you are on the lazy side there is also the Proland Ocean Plugin to help you with 3D water… 😉

  24. Ben, sorry about this question,
    but just to feed my curiosity.. we are in cycle xp11…
    are you working in parallel on xp12 ?? if not , when it` started?
    when you have a plan for the content of the new version?
    I mean whole team of course.
    Martin

  25. Ben,

    You held out some AMD optimization code that NV was able to take advantage of in 11.10 and had mentioned a few weeks back that it would be in beta after 11.10. Is that still the plan?

    Thanks for all the hard work!
    Jeff

    1. Yeah, AMD gets back on the fully optimized path in 11.20. We just had a bug in 11.10 that was destabilizing, so we had to hold off on enabling it.

    1. It’s _already_ in the sim…I think the real question is: when will we actually ship some art assets and docs. I don’t actually know…Max has done a lot of work with it, but we’ve been too side-tracked with VR (which requires artists to upgrade planes – more on this during the week) to finish it off.

  26. What happened to shaky ground feature? It was there for many-many years, different solutions were suggested (like buying more powerful computer or increasing models per frame) and nothing worked. Now its gone, was option shake_ground==1 just commented out?

    1. The XPLM camera plugin API provides a way for third parties to re-introduce ground shake if you miss it. 🙂

      The solution was to rewrite a portion of our transform stack in double precision and to add an extra subtraction to the single precision vertex shaders – I’ll write it up in a hacksoflife blog post one of these days.

      1. For me the ground shaking is not fixed in 11.10final.

        It was fixed from 11.10b1-b4, but starting with b5 the ground shaking is back. Already reported during the betas and assigned XPD-8461, but a little bit sad it’s still in the final. According to some posts on forums.x-plane.org I’m not alone with this regression.

        Clean install of X-Plane, no plugins, Win10, Intel I7 6700K, Nvidia GTX1080 (driver version 388.59)

        1. The bug you reported is _not_ a regression, it’s a different bug.

          The bug that was fixed was that _everything_ would jitter in most points in the DSF due to inadequate floating point precision in the camera code.

          The bugs you see are:
          – The texture mapping of projected .ter layers can jitter at some points in the DSF. The underlying mesh is holding still but the texture mapping is moving.
          – The water reflections are unstable. This is totally unrelated to anything iwth jitter.

          It’s not a regression from b4 either – these things _never_ were more stable than they are now.

          EDIT: actually, it is possible that the texture jitter was more stable in one of the betas … but only by being in the wrong place. At one point the texture mapping was mis-calibrated and thus was very stable but not of the right sections of texture. I _thought_ we fixed that bug before public beta thuogh.

          1. Thank you for clarifying!

            I thought the texture jitter was the ground shake people where talking about. It didn’t occur to me those are actually two bugs, so when the jitter returned for me, I thought it was a regression.

            Up to 11.10b4 the graphics seemed so perfect and super stable to me. No jitters, no flickers and perfectly stable reflections.

            If there’s anything I can do to help to fix those bugs, like running some tests, sending logs or anythings else, don’t hesitate to contact me and I’ll be more than happy to help!

  27. Hi Ben,

    I still don’t completely understand how the Draw Parked Aircraft setting places aircraft.

    For example, if I assign American Airlines to a Category C Gate in WED, X-Plane will sometimes place a Category B Air France jet even though the MisterX Static Aircraft Library contains a lot of jets that fit the actual requirements assigned to the gate in WED.

    Why does it ignore the requirements sometimes? Am I missing something?

    1. Does the MisterX static library meet the _library_ requirements of your gate?

      If someone has a scenery pack that exports an airline-correct library path, creates an airline specific spot, and that airline is ignored, file a bug containing ALL of this small scenery pack and I’ll take a look.

      1. After some experimentation, I believe I’ve found the two root causes of the problem. However, I haven’t filed a bug report yet as I think these might be intentional.

        Cause 1: X-Plane places aircraft of the specified size category AND one size lower.

        Cause 2: X-Plane randomly chooses 1 airline from the list of airlines assigned to a gate. If it can’t find that airline in the library, it doesn’t check the other airlines that are assigned. Instead, it just ‘gives up’ and randomly chooses an airline from all that are available in the library, including ones not assigned to the gate.

        IMHO, this strategy of choosing static aircraft should be reconsidered. X-Plane should not allow aircraft that are one size smaller. For example, I can often find category-A Cirrus Vision GA planes sitting at terminal gates assigned to category-B airliners.

        Also, X-Plane should check ALL airlines assigned to a gate, not just one. If none of the assigned airlines are available in the library, it shouldn’t place any aircraft at all. I think seeing no airline at all would be more realistic than seeing a totally incorrect airline.

        1. Case 2 is a design mistake – please file it. If you list 8 airplanes and we have one, we should do the one we have, not pick one we don’t have and then go random.

          Case 1, I’m not sold on…if we don’t go a size lower, the traffic’s going to be way more homogenous…

          1. Bugs filed!

            I understand your reason for Case 1, but the one-size drop should be disabled in certain situations. For example, X-Plane should know that applying the one-size drop to a size-B airliner gate would result in a size-A GA jet being placed at an airline terminal (which looks ridiculous)

  28. Hi Ben,
    Pretty sure you’ve been worked to death with these questions. With the advent of Vulkan, do you think we may see improved atmospheric and planet rendering? Current the sun phases over the earth, the earth_orbit textures are low resolution, red and draw in a weird manner. Local tweaks in temperature and skycolour? Also the return of any gamma controls or HDR brightness sliders? (coming in with more the mindset of a photographer so apologies!)

    *hands over a bottle of scotch*

    1. There is NO relationship between Vulkan and any of the ‘how the frame looks’ issues…Vulkan doesn’t make our shaders look better, it doesn’t change the scattering model, it doesn’t change the way we render the planet, etc.

      I would like to some day provide photography-style rendering controls, e.g. white balance, exposure, etc. Gamma control is _never_ coming back…it’s the wrong way to manipulate an image for aesthetic reasons; having it was an artifact of long-ago times when Mac and Windows weren’t in the same gamma space. Now everything is in sRGB and often stays in sRGB through a lot of processing — calibrate your monitor! 🙂

      (Austin was complaining that the 737 was too dark inside…he has a third party 4K monitor on his trash-can Mac. When we were at the last meeting, we couldn’t believe how much darker it was than what we saw on our iMacs and laptops. Chris looked up calibration from Tom’s Hardware and it made a _huge_ difference…)

      1. Sounds most interesting, looking forward to it. (And I know that Vulkan has no correlation on frames or features, I probably should have worded that better!) Would have been interesting to see how dark Austins setups looked!

        Thanks for the reply!

      2. A dedicated ‘Photo Mode’ would be really cool, with control over things as motion blur, depth of field, etc.

  29. Great job as always guys, I do have one question though regarding leading-edge slats.
    Those Xplane support leading-edge slats that are gravity actuated? Such as those found on the A4 Skyhawk and some other aircraft. The slats or leading-edge extensions are gravity actuated but retracted based on airspeed. In addition to dynamic actuation as a result of g loading.? I can simulate this in the animations quite easily and very effectively. But it does not affect the flight model. Is there a way that flight model can be calculated from these third-party graphics?

  30. Hi Ben, thanks for taking the time to respond to so many comments.

    Are there any plans to fix the “vibrating shadows” phenomenon? With the 11.10, while the ground shaking is fixed, I still continue to have building shadows “vibrate” or “shake”.

    Is this a known bug? If not, happy to file one.

  31. Got some noticable fps drop after 11.10 update. Had smooth flying with high setting in my GT740M card in 11.05. Not can’t even takeoff due to stutters. Expecting hot patch from laminar for this fps drop bug. Regards.

  32. A greeting.
    Good luck with the implementation of VR.
    As a user from xplane 7 to 11 I would like to have photo-style rendering controls someday.
    Would that be possible in the near future?

  33. Hi Ben, I want to use C172 to make home cockpit, can you tell me how to remove the glass reflective effect.

  34. I’m sure you’ve been asked this before, so sorry for asking again. Have you guys considered simply buying off a game-engine, like Unigine, and focusing yourselves on the physics aspect?

    If my understanding is correct, the license is around 100k usd, which is roughly the cost of one senior developer for one year… and it seems to me that it would be worth it, when I look at the visuals…

    Cheers, thanks again for all the good work.

      1. Well, I guess Laminar’s lead graphic developer wouldn’t be too happy about that suggestion. I hope he doesn’t ever read this blog!

        1. I think he’s off drinking scotch or something. I’d drink if I saw how weird shader binaries are acting on the AMD drivers. What voodoo!

          This is an actual snippet from an email Austin sent Chris and I many years ago, based on customer feedback.

          austin meyer wrote:
          > hi ben and chris!
          >
          > well, it looks like i have to fire you guys, and quit x-plane and go on welfare myself.
          >
          > once of the customers says i need to ‘farm the software dev out to an experienced software-dev team’
          >
          > austin
          >
          >> (user name removed)
          > :
          >> , I guess in summary I just wish Laminar could farm the software dev out to an experienced software-dev team,

          In all seriousness, using an off-the-shelf engine wasn’t really possible 5 years ago – general flight sim requirements were too far off from what the engines had. This _may_ be possible now – Unigine has features UE4 and Unity (for example) don’t have that make it more appropriate for flight sim – you don’t get planetary scale in every engine these days.

          It’s not as simple as sitting on top of an off-the-shelf engine; the engine has a huge impact on what kind of content you can create and how it looks. So moving an existing in-house app to an off-the-shelf engine is years of engineering .. it might be easier to enhance the existing engine.

          1. > you don’t get planetary scale in every engine these days.

            Side note: look at the countless millions Chris Roberts and the Star Citizen team are burning through trying to tackle something like this.

        2. Marco,

          The lead developer might be happy if they would simply buy the code, and the lead developer would have freedom on what to change in the engine.

          I don’t know if this is possible for full-fledged game-engines, but this is possible for trueSKY, and trueSKY looks awesome in FSW.

          (Keep in mind that X-Plane is by far my favorite simulator, but I’m unbiased enough to recognize some aspects that are better in other sims, and would love to see this improved in X-Plane).

          Cheers!

    1. Recovering 2.8 ms of frame time by killing the FFT water is _not_ a bug. You get a FPS boost by pausing the sim (and thus killing the physics) too.

        1. Elios, if that’s true that could be interesting! Generally, areas with a lot of water have better performance due to the associated “lack” of autogen to draw… so if in areas without water we could have that fps boost, it could be a win-win situation!!

  35. Hi Ben,

    more and more often I observe people tuning X-Plane to look like watching through the lense of a camera. Some plug-ins come with lense flares or even turn red at night to simulate a CCD chip without proper white balance. In __my__ opinion all this is what X-Plane should NOT do. This is a flight simulator and it should simulate the view from a human eye in the cockpit.

    CCD camera chips tend to turn to red in the dark while the human eye looses the capability to see colors and therefore turns to a grey color view.

    I think the demands for those camera effects come from people having no real world cockpit experience. The only reference these people have are video sources – often taken by smart phones and shown on you tube. The worse the sim looks like by matching the poor camera views, the more they feel X-Plane looks real.

    It could be argued to be a matter of taste – but this would apply for screenshot hunters – not for those people looking for simulation showing the human eye view.

    However: Imagine someone would like to simulate a UAV – then all these camera/photographic filter setups would be very useful.

    Maybe some time there could be a button to switch from human eye view to camera view and back…? This also would meat all the different user demands.

    Just my 2 cents…

    Regards,
    Marc

  36. Hi guys,

    I just wonder, if something with rendering cars and FFT water was rolled back in the release. I was impressed, that car/water rendering does not cost any FPS since 11.05?

    But now, it costs me about 5-7 fps. Once I disable both with lua, I got my full fps from 11.05 back.

    Otherwise, great work guys…

    Cheers,
    Michael

    1. Whether or not they’re free depends on how fast the sim is running and what the other cores are doing. For the other user who is already running at 75 fps, yeah, he got some latency back by killing them. What happened is the other threads were running so quick they got done first.

      1. Well, I tried different scenery and the problem is still here, from 30-33 to 38-40 just killing FFT (in windowed mode). It definitely worked much faster in 11.05.

        I must add that it is not that noticeable in full screen mode, only windowed. Disbling FFT in full screen doesn’t seem to affect FPS

        1. When you are concerned about FPS, you always should use full screen, to make sure the OS gives all resources and processor cores to X-Plane. (At least on Windows, don’t know about other OS.)

          If disabling a feature increases FPS, I would not call that a problem. Your numbers say that the FFT calculations take about 5ms per frame. Assuming some non-optimal handling of X-Plane by Windows due to windowed, this seems ok to me.

          If disabling FFT did not make a difference in previous versions, I guess that could have a lot of reasons. For example other tasks on another core are so optimised now, that FFT matters in some conditions.

          BTW: The big difference between windowed and full screen shows – as far as I understand – that X-Plane is not at all that “single core”, as we still think.

  37. Are you guys still on track for 2017 for VR , I mean there are less than 3 weeks left.

    crosses fingers.

    1. If you look at their upgrade schedule Mary Ann you will see that 11.10 first beta was released October 14 and it took till December 9th before it was final.
      So VR will be released in 11.20 which is the next major release so therefore based on their release schedule there is no way it will be a 2017 release.
      My guess would be the first or second week of Jan for the first beta (maybe)
      Only time will tell.

      1. You are correct that there is no way VR will be _final_ in 2017. We are still aiming to get the public beta going this year. Our time frame for the public beta is basically ASAP.

  38. Hey guys, I just wanted to say I’m a recent convert from FSX. I’ve used both sims in VR, but X-Plane just blows away FSX in terms of realism, immersion, and scale. It makes me extra happy to see that native VR support is just around the corner!

    1. Same here (P3D &DCS actually), I’m super happy with moving to X-Plane. I’m just starting to play with it but my understanding is that X-Plane has the best telemetry for my 6DOF motion simulator as well. =D

  39. Are you guys considering more bumpy runways and taxiways? I ask because the wings bounce as we move, whereas the wings get quite jumpy in real life.

  40. Hi Ben,
    Sorry if I repeat myself, I already asked this question a bunch of updates ago, so I apologize in advance. But given the relevant graphics engine tweaks I wonder if the “disable threaded optimization” for Nvidia hardware owners still holds.

    Thanks a lot for everything!
    Filippo

  41. Hi Ben, will X-Plane 11 ever work `offline` or does it need to check every start-up of the program…

    1. If you have the DVD set, you can operate 100% offline. For digital download users, we require you to at least periodically be online in order to re-authenticate your key. Right now the requirement is once every two weeks. (This protects us from someone sharing their key online, and having thousands of users download the sim at our expense.) The latter isn’t likely to change.

      1. Checking online every two weeks doesn’t make any sense. I may happen to be operating offline on the very day the check needs to be made – and then I’m stuck!!!

        A much better solution would be to check every time I’m online and if valid then reset a counter to let’s say 10. Every offline startup decrements the counter by 1. At 0, force an online check.

        This would make checking seamless to the user!

        An even better hybrid solution could include a time period plus a counter!

        1. Sorry, I should have been more clear: we do indeed re-authenticate every time you’re online. In the event that you’re offline, though, you have up to 2 weeks to connect to the internet before we require it.

          1. Hi Tyler,
            a very fair and user friendly solution! Same as the whole installation and “portable style” of X-Plane. Makes everything very easy for us. : )
            Not very common for a software of this size and quality…
            Thank you very much!

  42. Are the new airports downloaded automatically with the new releases or do you have to download each one separately?

      1. I am a bit confused as to what is “new;” what is the date by which the Gateway airport became the Recommended one for it to be included in 11.10rc3? Is it still 20 September 2017, as in the NOTAM on the Gateway?

  43. Please can we have the old little and White sight for those who fly with mouse yet???
    Please that black box and cross are horrible specially in blue skyies!!! I will be very gratefull!!!

  44. Hi Ben,

    FMOD question: is there a launch milestone for the event system you mentioned a while ago which would make easier to override the native sounds triggered by the flight model and use FMOD instead?

    Thanks!

    1. No. I’d like to see it in the release after VR, but we’re so “get VR done!!” right now that the release after that is sort of a fuzzy “things happen here” release right now.

  45. Great work as usual from the whole team, constantly improving XP. 11.10 has been runny very well for me. I must say I really love the new UI features for saving different Joystick profiles. As someone who fly’s several different aircraft on a regular basis this has been a welcomed improvement.

    UI Question:
    In the future, do you think you will add a “Custom FOV” setter per profile. So, the main one in the graphic setting would be default but it would be nice to be able to set the custom FOV per aircraft?

    I also enjoy the airport search and sorting function and the ability to mark airports as a favorite. As someone who enjoys always trying new airports and all are not my favorites. It would be nice to have some indication when looking at the list of thousands to distinguish between airports I’ve been too; verses my favorites. Maybe sort by unlogged “not in log book” or something.

    Just thought I’d mention it because I know you have nothing better to do and always appreciate dumb ideas from you users.

    Keep up the awesome work LR Team and friends.

  46. Ben, I read on the questions.x-plane.org site that someone “heard” that the intentional limit of 10 USB devices to the sim would be uprated to 20 USB devices with 11.10. Did this occur? I checked the release notes and saw no mention. This is a limitation more for cockpit builders than casual enthusiasts. But with touchscreen input becoming more and more useful, the device counts will go up. I saw one reference to someone with 13. Must be quite the cockpit!

          1. Will modes (allowing you to reprogram each button differently for each mode) like you see with the ChPro Throttle ever be supported?

          2. Tyler – “What’s the use case for that compared to per-aircraft joystick assignments like 11.10 supports? What’s the benefit?”

            CH Pro Throttle
            Benefit is you get 3 times the buttons at your fingertips if you can use the modes switch.
            You can only use one mode at a time.
            So, while the buttons per single USB device is less than the max limit, it would mean allowing each button to be programmed to 3 different commands, each command assigned to a specific mode.

            So it would be nice if I switched to a mode using the switch on the throttle, in the UI it would indicate the mode and I could program the button for that mode.

            Or

            Each mode was treated as a separate USB device meaning the Throttle would show up 3 times in the drop-down menu for devices as a separate mode so the buttons/hats/etc. could be programmed separately for each mode.

            3 axes and 24 buttons
            • X and Y axes on miniature joystick
            • Z axis on accurate throttle sliding motion
            • 3 traditional push buttons
            • 1 mode switch button
            You switch to a different mode and program the same buttons to something different.
            • 3 four direction HAT switches
            1 eight direction POV HAT switch
            Total of 176 programmable functions with Control Manager software )
            Three color mode indicator LEDs on base

            Hope this makes sense.

            Thanks,
            Dan

  47. I’m trying to figure out when and how I can trigger the “low fuel pressure” warning.
    By doing this I had a look for “sim/cockpit2/engine/indicators/fuel_pressure_psi” at the Cessna 172 (high wing) and the C90 (low wing) and at the Carenado B200.

    Both King Airs have a pressure of about 6 to 7 PSI, without a running fuel pump, the Cessna shows about 0.9 PSI.

    Where does that “high” pressure of the King Airs does come from?
    If it is a bug I know what I have to do 🙂

      1. Engines were not running and stand by pumps were off.
        With fuel pump on the pressure is 30 PSI.

  48. What works to the render engine have been done? With the upgrade it now feels like I’m flying drunk – everything has a double vision style blur to it. I’m running GTX950 and I7 and previous to this update was smooth and visually OK. Now however, it’s too hard to play.

  49. Hello, dev Team.

    Can I help you to translate the X-Plane for Portuguese.
    Please, there is a lot of errors and incomplete translation.

    So, if you have interest, pls contact me.

    Regards,
    Anderson Lucas @simulanderson

  50. I was curious if there were any plans to add more features to the ATC so that it can be used for tower instructions/directions to active runway for takeoffs and landings while not having to file a flight plan?

    1. At some point we will add this. We call this feature “VFR flight” because we need the entire package of VFR pattern ops, access to the tower without an IFR flight plan, and a little bit of glue for getting into and out of the system without an IFR plan in-air.

Comments are closed.