Beta 15: A Quick Fix

Just  a quick note: beta 15 is up and fixes a few key bugs:

  • No more Plane-Maker crashes opening files.
  • ACF files aren't locked while the sim is running.
  • Cessna takeoff should be a lot better.

I've heard a lot of chatter about FPS problems with the new clouds. If you are having FPS problems with the clouds, here are a few key things we need for a bug reports:

  • Include a log.txt. We get enough bug reports that if you ignore the "always include a log.txt" note in the bug report, your bug report will go to the back of the line. Everything that we tend to ask users that can be gathered automatically is in the log.txt: your CPU info, GPU info, driver, version of X-Plane, add-ons installed, scenery...there's a reason why we want it!
  • Be sure to note your monitor configuration (how many, how big, what you're doing with them).
  • Include a full res screenshot of the clouds where FPS are bad, preferably with the FPS counter showing. Please don't use JPEG or crop it - I need to see what the rendering engine is doing.
  • Try setting the weather to clear - does that improve FPS?  Try changing your window size. Does that improve FPS?  These differentials are useful.
  • For a given view where FPS are awful, does regenerating the weather (bind a command key) change things or is it always bad?
  • If FPS are always bad, send us a .sit file where, in the cockpit of an aircraft that we ship with the sim, framerate is really bad.  We can load the .sit file, regenerate the weather a few times, and see what's up.
  • Update: one more thing - if you are using a third party aircraft, does changing to one of ours fix the issue? Does changing the shape of the 3-d cockpit (which blocks out clouds) fix the issue?



  • Facebook
  • Reddit
  • Twitter
  • LinkedIn

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.
This entry was posted in Development. Bookmark the permalink.

75 Responses to Beta 15: A Quick Fix

  1. Chuck says:

    Ben you said,

    "Include a full res screenshot of the clouds where FPS are bad, preferably with the FPS counter showing. Please don’t use JPEG or crop it – I need to see what the rendering engine is doing."

    What is your preferred screenshot file type if not JPEG?

  2. John says:

    Hello Ben,

    most simmers get problems with clouds at minium low systems. 2GB V-Ram etc.
    Low GPU.
    1050ti+ users report good results but even low FPS as normal.

    I just tested with GTX660 all settings at minimum. It was worse. Not playable with real weather input. Overcast and broken. 1500ft cloud base.
    Change to clear sky...Fps raise up to playable.

    Another thing:
    With real weather in Cockpit view heading straight not playable.
    Look right or left out of the cockpit FPS raise.Playable.
    Serious.
    I disabled all plugins and data outputs as well GPS .

    Greeting
    John

    • Ben Supnik says:

      Please file a bug - the post includes the information I need.

      • Out of curiosity regarding the situation-files ... if I am on a flight with "real weather" and save .sit, does the weather get saved so that when I reload the situation, it will be the way it was when I *saved* it instead of the way it is "in real world" when I *load* it?

        • Ben Supnik says:

          I don't actually know. I think the weather will get saved but may get overwritten on reload. In theory going back to static weather might load static weather with the current METAR. I haven't explored these edge cases in the v11 carefully.

    • Collin H says:

      I think I can re-produce this on my more powerful machine with a pathological cloud case. I've got 3 overcast cumulus layers stacked on top of eachother. If I look directly into the sun, 19 FPS, if I look about 30 degrees right or left of it, frames jump back up into the 30s.

      Perhaps also of note is the BetterSky Lua script seems to exacerbate this problem, specifically when it sets the dataref "sim/private/controls/clouds/shadow_size" from 1024 to 2048.

      I can file a bug with the rest of the details.

  3. Kirill says:

    It is very unpleasant that there are no stars, the moon and the sun, as well as the sky has no depth, it is very rough, I really want to believe that this can be corrected!

    • Daniela says:

      It has said infinite number of times that those will be corrected at the end of the beta cycle. Please be patient.

  4. GPB says:

    Ben, any word on AI operating at airports with runways having custom pavement textures with transparent pavement type set on?

    Thanks...Greg

    • Ben Supnik says:

      We know about this bug - it's on our todo list, but won't make 11.0.

      We may need to add an extension to the apt.dat for this - we're not sure yet.

  5. XPlanePort says:

    Ben,

    Thank you and all the team members for all your hard work!

    I have question:

    I am building my latest addition to the Airport Scenery Gateway.

    First time attempting to include ATC Flows.

    Everything is working as expected, except:

    I have two AI aircrafts added in X-plane (Cessna Skyhawk (Laminar Research) and King Air C90 (Laminar Research)) to automatically populate one of many ramp starts, so I can test the ATC Flows.

    When starting new flight at the airport, one of the AI starts at a ramp start (as expected) and requests taxi clearance immediately after X-Plane loads. With the winds, I expect the ATC to instruct taxi to (my custom ATC Flow driven) appropriate runway, but the very first plane that contacts ATC, gets instruction to taxi to wrong runway (based on X-Plane default ATC flow).

    I have developer console open and I can see that ATC Flow changes from default to my custom (as it should, based on the current wind), but right after the ATC instructs the first aircraft to taxy to wrong runway (based on the X-Plane default ATC flow).

    I don’t think I am creating the Flow incorrectly, it is being activated, but with slight delay that is just enough to let the very first plane be instructed using the default ATC flow.

    Any comments would be appreciated.

    If it is known please let me know. If you think it is possibly a bug, I will file it properly.

    Thank you!

    • Ben Supnik says:

      Do you cover -every- wind direction with your flows? I wonder if the weather on the first frame is "falling through". Try adding at the end a flow that catches everything (max wind 999, direction 0-360, etc.). Does the bug change to go from catch-all to your flow?

      Either way, please file it - it looks like we have an out of order bug between weather and ATC.

    • Bob Marsh says:

      I have filed a bug on this several times before in XP10 and it must still exist in XP 11. When XPlane first comes up, the state of WX is NOT initialized at the current WX even though current weather is selected in the initial setup. That setup does occur, but only several seconds AFTER the initialization of your aircraft(s). So at that point the operations at your airport are at the DEFAULT weather setting selected by a NON-real WX startup situation. Now with the airport set to that weather and wind direction AI takeoff and landings that are in progress must continue at the runway(s) which conform to the DEFAULT WX until those AI aircraft have completed their assigned Takeoff and Landings. Any other later clearances will be notified of a change in runway assignment pending and asked to hold (assuming that the real weather does change the active RW assignment.) After the wait for the above situation to clear, normal operations will start as expected.
      Obviously, the sampling and setup of initialization for XP should delay the setup of initial AI aircraft until AFTER the METAR data is used to create the WX conditions is established to correct this problem.
      A work around to avoid waiting is to specify your start at some other airport than you want to use at XPlane startup, and after initialization, change your airport location to the real one you want to use. This will make the WX set prior to the dispatch of AI aircraft and use the correct WX data to start with.

  6. XPlanePort says:

    I think I found my answer here: http://developer.x-plane.com/?article=atc-flow-authoring-in-wed

    But in this case, on the initial X-Plane load, should it hold the departures for one minute, in order for the correct (custom) ATC Flow to be picked for the initial weather (wind)? Or, load the correct custom ATC Flow without waiting (right after weather loads), so the first departure is correctly instructed?

    Thank you.

  7. Bill says:

    Ben, nice job to you and the group. I'm getting 30-40 fps with my late 2012 iMac around the Bay area with moderate settings and clouds, default a/c. No issues with the 172.

  8. John says:

    Finally found out that when i disable SASL Plugin Frame rise 10 FPS.
    What did you do ever with the SASL Script ?
    Let it be..it worked all fine until you edit new SASL scripts.

    Not ok.

    Cheers
    John

    • Ben Supnik says:

      I do not know what you're talking about. We do nothing with SASL scripts - they're third party. If turning on SASL loses 10 fps, talk to the aircraft maker.

  9. twgin says:

    Off topic... did you have any luck filling the two positions ?

    Terry

  10. Werner says:

    Hi all,

    I really love X-Plane and enjoyed it since XP10. I am impressed with the capabilities of XP11. However like many other users I really miss a few things in XP11. Having read many blogs and articles, it is my hope that LR will reconsider a few things for XP11:

    Me and many friends in the XP community enjoy flying in photorealistic scenery, like Ortho4xp etc. This worked great in XP10 and was the main reason for me to switch to X-Plane.
    However since XP11 the extended DSF has killed the FPS. Before in XP10 I averaged about 40 FPS while flying in photo scenery. Now in XP11 the system gets overwhelmed by the extended DSF and my frames drop to 1.2 and 4.0 for over a minute when new tiles are being loaded. Yes, you have read correctly. One point two frames per second. This bringes the whole system to a complete standstill for a minute or so. And it looks like that I am not alone. Many of my friends at IVAO have reported similar issues.

    Therefore I would like to plead to LR: PLEASE give us back the possibility to switch extended DSF on and off!

    Another issue is that the amount of autogen trees is fixed. Like my friends I have the problem of too many trees and not enough buildings/houses in XP11; or vice versa.

    Please reconsider the idea of providing an "advanced" function in the graphics section, where users can adjust these parameters like in XP10.

    Really, really appreciate your consideration of the above!

    Another small issue I detected today is that the tug plane cannot be changed in XP11. After changing the tug in AI acft, the system always reverts back to the Stinson L-5.

    Thank you very much for your understanding!

    Best regards,

    Werner

    • Alpenwolf says:

      Same here - having a drop down to 2-3 fps every couple of minutes is pretty annoying and that's since the extended dsf kicked in. Okay, my hickups only last a couple of seconds but it isn't that great anyway.
      No, I will not give away my ortho4xp tiles - no way!
      i7 6700K, GTX 980TI isn't that bad after all - is it? There shouldn't be an almost complete freeze of XP

  11. NickH says:

    Hi Ben,

    Just tried Beta 15 and framerate is better for me on this one.

    I also wondered how high native VR support ranks in the (long) todo list I'm sure you guys have? The flyinside plugin is a pretty sub-optimal experience and I'm really keen to use my Rift properly with X-Plane.

    • Ben Supnik says:

      Native VR is VERY high priority.

    • Kopelent Jürgen says:

      To have advanced settings at XP11 would be nice. I am also not able anymore to use all my Ortho Scenerys with this hardcoded extended DSF, because the gpu load is too much an the reload lags are terrible. VFR pilots don´t need this extended dsf features. To hardcode settings is an big big step backward at XP11. I was loving X-Plane since XP9 about the lot of adjustment possibilities. (one of the best and beautifull things of X-Plane).

  12. Christophe says:

    Hi Ben,

    I filled a bug about the scenery shadow artefacts few weeks ago.

    Just to avoid to send to you some new cases, do you have a roadmap target to fix those artefacts ?

    Thanks.

  13. DreamKeys says:

    Did anyone noticed the stock 747 having problems when controlled by the AI? It does tailstrikes on every takeoff and comes down way below the glideslope on landing, looks like a minor stall. Saw that consistently while working on the YSSY scenery. It ditches always a couple hundred feet in front of 16L in the water and does a Go-Around. I'll file a bug later when I'm at the PC.

  14. About changing window size ... on one forum I was trying to help out a user (this was some pbs back, I think) with their FPS problems. Turned out that they were using a 4K monitor and even though they made the sim windowed and changed it to (by their own words) "to about 1920x1080" they said that there was no noticeable change in FPS. To me this sounded very counterintuitive (because they were suppposedly only pumping one fourth of the pixels to the screen I'd've expected some change). Unfortunately I don't have such a huge resolution available, so I was unable to test it by myself ... and that particular user disappeared and hasn't been seen since (I can only expect they got their problem solved and didn't bother to mention what they'd done).

    • John Rogers says:

      I am running a 4K screen with XP11 with a GTX1080 video card. At that resolution I'm still CPU limited so I don't see much of an improvement if I change the resolution to 1920x1080. That other user may have been in the same place and had their fps limited by their cpu.

  15. Robert says:

    Xp11 Beta 15 : 1)Weather not updating 2)Time not updating. Tried updating program through installer. Xp reports everything up to date . Went to various locations to see if it was working else where. Not working.

  16. John says:

    It's problem about SASL with 3rd party related. Carenado etc.
    I have tested turn SASL plugin off and...frames rise 10fps. No problems with clouds, only with some Add-on planes.
    But even on SASL, it's a hungry FPS beast.
    Maybe further Carenado etc. fix that problem with new scripts to their Add-on planes.

    • Carenado has already stated repeatedly that their aircraft are not updated for X-Plane 11 yet. So when you are using them in XP11, consider yourself in unsupported territory.

  17. Saso Kiselkov says:

    Don't know if it's already been reported, but ground handling is seriously screwed in this PB (might have been in previous ones as well, last one I tested before this is pb11).
    Just tested the FJS 732 in an otherwise vanilla install. At full nosewheel deflection, which should be 60 degrees, and lots of power applied, I'm spinning around one main wheel stopped. No brakes applied, it's like the gear is glued to the asphalt. Also, at anything over 40-45 degrees of nosewheel deflection and it's like the aircraft just hit a wall and grinds to a halt unless excessive amounts of power are applied. This makes precisely pulling into a gate (or anywhere where tight turns are required) excessively fidgety. You're either grinding to a halt, or skidding like crazy because of too much power.
    Whatever you did to the ground roll handling between pb15 and pb11, please undo it for airliners. It makes them virtually uncontrollable on the ground.

  18. Yann says:

    Hi Ben, quick question for you, what is the version of the OpenGL driver that XP11 uses? I read somewhere that is was coded for OpenGL 3.1 but I find it strange as it is quite outdated now... My Nvidia driver support OpenGL 4.4 so wanted to make sure I do not need to override to 3.1 in Nvidia Inspector.

    Thanks in advance!

    • Ben Supnik says:

      We use whatever the latest is the driver offers. Our minimum is either 3.0 with some extensions (Win/Lin) or 2.1 with even more extensions (Mac).

      But...I cannot emphasize strongly enough how totally unimportant this is. OGL isn't like DirectX where there are major incompatible versions and not having the right version is a total show stopper. OpenGL exposes functionality piecemeal via extensions and we can pick that functionality up as we need it. So in v10, we were using instancing while still requiring 2.1 as a minimum. If hardware didn't have instancing, we used another code path.

      So...
      - Don't mess with the nvidia settings - the defaults are fine.
      - Don't worry about this - it's totally unimportant.

      cheers
      Ben

  19. Alex says:

    [Message deleted to due to language.]

  20. Filippo says:

    Hi Ben,

    regarding the enhancements to the engine modeling, especially the turbine engines as Austin wrote about some time ago, are you going to model abnormal turbine starts such as hot starts / hung starts? Just to make a practical example, I often use the default Beech C90 but since I don't have (yet...) a twin throttle panel I start the engines with the condition lever already in low idle position. A real engine would surely make a hot start in such situation, with ITT temperatures going nuts, but the default C90 in X-Plane gracefully starts as if the condition lever was moved at the right time. I also noticed a strange performance drop with altitude (examples: the simulated aircraft struggles to climb above FL200 - real weather enabled - while the real aircraft is credited with a FL290 service ceiling at MTOW; the max power cruise speed of the simulated aircraft at FL160 is 210 KTAS, about 40 knots less than what performance tables say for the real thing).
    I would like to make some extra tests as soon as I have complete performance tables available for the real aircraft. Should it be the case to file a bug if the performance differences are confirmed wrt real performance tables?

  21. John says:

    Next try today with new Nvidia driver and gtx1050ti. Win10.
    The FPS 14 ist not acceptable with stock aircraft. CPU at 60% Ram 70% GPU 50%.
    I dont know what you do in the switch from PB14 to PB15 ???
    There is something totally wrong.
    Greetings

  22. John says:

    Bug report filled.
    But why is such an performance drop from PB 14 to 15 ?

    You get many bug reports about this problem slowly its time for answers ? 😉

  23. AJ says:

    Dear Ben,

    I feel so sorry for you having to repeat "please file a bug" to everyone. How you haven't smashed your PC monitor, or set your office on fire I don't know... unless you have before?

    Anyway, I'm having this bug that I need to tell you about... 😀

    Jokes aside please file a bug everyone!

  24. John says:

    XP11 PB15 VRAM in Game 400MB in use ???
    I have got 4GB VRam.
    GPU: 20 %.

    Seems that PB15 dont't use all resources of the 1050TI.
    Why?

    It's still not flyable on ultra low settings. FPS 10-14.

    • Ben Supnik says:

      This isn't useful data. Here's why: something is wrong in the cloud code. It's killing performance.

      Turning everything ELSE down isn't making the sim faster because the cloud code is going haywire. It _is_ underloading your machine, so that you can have low FPS and underused hardware.

      If you've filed a bug, please wait for us to fix it.

  25. Riccardo says:

    When relased beta 16 this weekend ?

  26. RD says:

    Now ignoring the probable bug in the cloud code, I would like to know: how are they supposed to perform comparing to the old ones (when working as expected)?

    I mean, the new clouds are beautiful, period. In the end, this normally means a step forward on hardware requirement.
    But seeing by the changelog and some of your posts/comments in the blog, you were talking about - performance improvements in clouds -.

    So what is expected? better-looking and faster or better looking but a bit slower?

    Thanks!

    • Ben Supnik says:

      They should be better looking _and_ faster. Two things happened:
      - Max tuned up the art assets.
      - I put in code to (under some circumstances) reduce fill rate costs.
      In one pathological case I've been sent, the clouds are taking longer than before, even when _none_ are on screen. That means something's gone horribly wrong, e.g. it isn't a case where we should be slow.

      • Kris says:

        And this is why I love this dev blog. It's fascinating to see what goes on under the hood with a sim this complex. Keep it up Ben, PB15 looks great, I can't wait to see the final version.

  27. John says:

    Thanks for some answers 🙂 Wish to see my 1050ti hard working in next beta...

    It has something with real flying with the betas...emergency procedures...what happen next. 🙂

    Finger crossed. Many thanks for work so far.

    Greetings

  28. Günther Wittwar says:

    New clouds
    Today I flew at FL200 with a lot of heavy thunderstorms around.
    Below there was a 8/8 cloud layer, but nowhere a Cumulus Nimbus was shown.
    I know, the new clouds are a first (and good looking) step, but there is enough room for development.

  29. John says:

    I must say the cloud integration ist just wow...double check it outside and in the sim. Perfect match. Good work...so far. State of the art...really.

    But i will not reach above 14 FPS. And stuttering.
    Hopefully in Beta16 it will run like a beast at 30FPS.

    System:
    Win10
    Q6600 Intel Quad
    8GB Ram
    Geforce 1050TI VRam 4GB

  30. Albert says:

    Very bad performance with new clouds when loading or refreshing weather!!
    From 50FPS down to 2-3FPS for 10-20sec every 5-10min.
    My system: I7-4790K 4.6Ghz, 16GB RAM 1866Mhz, GTX 1080 OC, 1080p monitor, Samsung SSD 250gb for XP11 only!
    What files do you need for decoding, I want to file a bug report?

  31. Mark says:

    Hi Ben,i still to have a raindrop bug into default airplanes. It will be fixed in the next versions?
    thanks for your work

Comments are closed.