Posts by whitav8

    Same here - corrupted textures.(980X@4.2GHZ, Win10 build 1511, Nvidia 970 with 376.33) Perhaps instead of posting the entire binary aircraft (DR400) with possible licensing/copyright issues, we could simply post the corrected textures? If that is not acceptable to IPACS, please remove the ZIP. The only folks who would need these are those that already have the SDK. Eventually - and perhaps by noticing from this ZIP file which ones are affected (I used filesize only as a determining factor), the image compression tools can be fixed. I assume that IPACS has a full copy of the aircraft with corrupted textures already. These textures are copied into the C:\Users\YOURUSERNAME\Documents\Aerofly FS 2\aircraft\dr400 directory.

    Dave

    @IPACS,
    Thanks very much for the web-mercator equations! That will really help. Could you please say what TM_PI is. EDIT (I guess it is also PI)

    Now, a few questions about scenery development and comparing it to FSX. In FSX/ESP/P3D there are libraries of objects that we can use for a given airport or scenery area in general - all we have to know is a GUID - a unique identifier - and then we can place that object with a given Lat-Lon, angular orientation- primarily a yaw setting, and scale factor. Is there any chance FS2 will sometime support that idea. Then I don't have to have the source for the object - in this case,the max file.
    In FSX, we can generate a simple runway system without having to actually model the runway - just supply the location, orientation, and lighting parameters.For simple runways, that would be a nice feature. There is a freeware tool - ADE(fsdeveloper.com) - that many of us use to make a fairly accurate runway model using an underlay of the aerial image. Also, freeware FsEarthTiles allows us to get an aerial image BMP file with accurate geo corners.
    The above two features would allow users or potentially third party addons like Instant Scenery 3 to be able to help us fill up the world where we really care.

    Thanks for a super product

    Dave

    That should be fine - I think the 480 approximates the Nv970 at least. And it would be good if you have a CPU at 4.0Ghz or better + 16gb RAM and an SSD. Go for it. I run VR with those specs with a custom graphics setting - 1.4 on the Rescale(SuperSampling), high,high, medium, medium,medium. With VR, you want 90fps most all the time (I don't like ASW frame interpolation) excepts for a few high complexity situations and maybe when a new area of scenery pages in.

    @IPACS,
    Thanks much for the SDK release. I've tried everything and thanks for the new airport, new aircraft with great visibility, and the sample data read/write application which compiles and executes with VisStudio2015 just fine.
    I wanted to add a very simple airport near my home in San Diego (Oceanside Municipal) but I really don't get the web-mercator coordinates. I have used websites and actual C++ code to be able to go from Lat,Lon to web-mercator (and back again) but I can't match the numbers that you used for Kingman objects - Example:

    <[tmsimulator_scenery_object][element][0]
    <[string8][type][ground]>
    <[string8][geometry][kigm_kingman_rwy]>
    <[vector3_float64][position][24043.51308231 51802.69537268 0]> <<<<=======

    From editing those web-mercator numbers (by 0.1), and then seeing the scenery shift, I see that the units are 100 meters (I think), but when I enter those coords ( 24043.51308231 51802.69537268) * 100.0 into my converters, I get lat,lons which don't make sense to me (shouldn't they be like the lat,lon for Kingman airport (-113, 35)? Also, what is your longitudinal reference point ? Could you please explain these? There must be some scale factors I don't know about.
    I am using simple equations like the following that I found - they agree with some online converters:
    East = (180.0 + lon) * 20037508.34 / 180.;
    North = log(tan((90.0 + lat) * PI / 360.)) / (PI / 180.);
    North = North * 20037508.34 / 180.;

    lw36,
    Thanks much for sending the ZIP of dr400!! It looks like there are maybe 10 or so ttx (texture) files that are different. You have really helped us debug this - and hopefully IPACS can figure it out. I assume that maybe the nvtt_v64.dll (the Nvidia Texture Tools Library) or cudart64_75.dll (Nvidia CUDA runtime) that is part of the converter in C:\Program Files\Aerofly FS 2 Aircraft Converter may have some issues on some of the PCs - still I have no idea what is the difference that matters between PCs.

    @IPACS,
    I am attaching the ZIP file of my "corrupted" dr400 files - hope this helps figure it out.
    EDIT: I decided to also provide my installed AircraftConverter tools - maybe that is somehow different

    Dave

    lw36 and IPACS,
    Merry Christmas!

    Thanks for the feedback. Something is happening during the conversion on my PC and at least one other user - we will figure it out sometime. I'm guessing that the corrupt generated files are int02_panel_color.ttx (4,370KB) and ext01_fuseandwings_color.ttx (2,958KB) - the originals in the SDK aircraft_workshop\dr400 look fine (49,153KB dates = 9/1/2016 and 12/9/2014). I suppose it could be bump,ambient,etc??

    I have a 980X@4.2Ghz, Windows 10 (Version 1511, Build 10586.713), 12gb RAM, Nvidia driver (tried both 375.95 and 376.33), Nvidia 970

    Not sure if file size will tell you (IPACS) anything but here are snapshots of the generated directory:

    As far as where the actual SDK model results go - it is at C:\Users\YOUR_USERNAME\Documents\Aerofly FS 2\aircraft and C:\Users\YOUR_USERNAME\Documents\Aerofly FS 2\scenery\places.

    I wonder if the aircraft_converter requires a certain CPU i7 instruction set, a GPU 10X0 instruction set, or system DLL. LW36 has an AMD video card I think.
    I've tried hitting the convert button several times, loading the newest video drivers (376.33), re-downloading the SDK and re-installing the aircraft converter - no change - several textures are either corrupted or simply super low res.

    I also am wondering about corruption of the Kingman airport scenery conversion - look at the "paw prints" on the taxiway in this photo:

    BTW, the Garmin moving map can have it's range scale changed by the top button. I am hoping that someday we can supply our own aerial imagery with the SDK tools to improve an airport that doesn't have surrounding imagery provided.

    Overall, the current SDK is a super great start. Hope a less expensive 3D modeler can be supported.

    Here is my attached logfile ( a shortened version due to attachment file size limits) that is in the Windows 10 Documents subdirectory called "Aerofly FS 2 Aircraft Converter" after the corrupt textures are produced. There are many ERROR reports but I don't know if it matters. I also note that the actual converter programs (in "Program Files\Aerofly FS 2 Content Converter" ) use the Nvidia CUDA and Texture Tools - maybe there is an issue with them on my particular video card - Nvidia 970??


    lw36 - what PC system components do you have?

    lw36 and jk1895,
    I wonder what is the difference between our systems?
    I have Windows 10 64-bit Version 1511 Build 10586.713) 980X @4.2Ghz, 12 GB RAM
    Maybe IPACS can figure it out - I am re-downloading the SDK and will try again - but a lot of the aircraft is fine (just the instrument panel and the side of the fuselage are really messed up) and I checked the instrument panel textures inside the SDK folder as well as the side of the fuselage texture and it looks fine. It seems to be an error during conversion (maybe) - but why. As far as video card and driver - I have an Nvidia 970 and running 375.75 drivers.

    In this outside image you can see the speckles on the side of the rear fuselage and aileron but the wings seem clear.

    A couple more videos of shadows of objects on the ground with the most recent release (EA3.66 from the F15). Again, I am trying various Nvidia drivers (currently 375.95) on my Nv970, Windows 10, ULTRA shadows setting. I am still wondering if other users see this or is it my personal video hardware/drivers. I didn't see any sign of it in the several New York demo videos posted yesterday.(BTW, awesome number of buildings!)
    There still seems to be a diagonal geometric plane at approximately center of the screen where toward the eyepoint is clear but on the outer side of that plane (in front of the F15) is oscillating. If this is a low resolution shadow map issue - will it ever be improved - or maybe do I need an Nv1080? It would be great if some others can fly this tight circle around KSFO and report back - thanks

    https://vimeo.com/196762196

    https://vimeo.com/196762020

    To all FS2 Aerobatic flyers,
    While you are waiting for the Reno race course or a Red Bull course in Dubai or something else to challenge your flying, enjoy the two bridges at San Francisco - not under them (too easy) but through them like a very fast car. I was surprised at how accurately the collision detection works and that the Bay Bridge girders aren't just transparent textures but real polygon models (stay low). This looks great in VR!!

    Dave

    https://vimeo.com/194693690

    https://vimeo.com/194693979

    Here is a video from outside the aircraft with full ULTRA enabled. I think the shadows are maybe better but I still see the moving graphics plane (between good and shimmering) that is fixed to the aircraft - perpendicular to the longitudinal axis. From inside the aircraft in the pilot's position, even with ULTRA shadows, it is still not that good.
    I would like to know if you see the same thing, or for some reason, it is just my particular PC (i7 with Nvidia 970)

    https://vimeo.com/194450965

    I'm not using Nvidia Inspector to set anything but it summarizes the Nvidia Control Panel settings:

    I assume this is the test you want me to try:

    "As a test, fly around the same as you did just use an outside view mode, then look at the distant shadows. It should look a little better then."

    Yes, I have tried low, medium, and ULTRA shadow settings - not too much difference. Are there any Nvidia graphics card settings at all that I should try?

    tangojulietpower,
    Glad you joined us - we need a lot of real life pilots to help tune the handling of these aircraft - it sounds like you think the Extra and the Pitts are pretty good. Personally I had to tune the controller gains - especially in pitch down somewhat - it was very twitchy but that is dependent on our own controller (I have a Thrustmaster HOTAS 4) and personal taste. Your comments will help the developers a lot. It would be great to negotiate with them just what ONE aircraft might be the best choice for them to model next.
    If you can - you didn't say - you must try a Oculus Rift or HTC Vive with FS2. Except for the G's, you will feel like you are right there.

    Please let us know about your PC and controller - maybe we can help tune your experience. The greatest part for me of FS2 is the absolute smoothness due to the very high framerate of FS2. Only War Thunder and FS2 can consistently get over 90fps which is required by the VR headset. ATW or ASW frame interpolation schemes just don't do it with high dynamic devices like aerobatic aircraft.

    Welcome

    Dave