1. GPU Fatal Error? Intel HD Graphics 620/630 Crashes? Help us solve them!
    Dismiss Notice

BeamNG: Personal Micro Blog(s) - Feburary 2015

Discussion in 'General Discussion' started by tdev, Feb 8, 2015.

Thread Status:
Not open for further replies.
  1. SoHigh00

    SoHigh00
    Expand Collapse

    Joined:
    Nov 11, 2013
    Messages:
    506
    Hey, I don't know where to put this but sice there are many devs active, Ill put it here. Has there been any more progress on the "race update"? By that I mean of course the actual races themselves...beeing able to create a pathway for the AI to follow, checkpoints, timers etc? Great progress lately by the way ! I also really apreciate the increase in communication :D
     
  2. tdev

    tdev
    Expand Collapse
    Developer
    Staff Member

    Joined:
    Aug 3, 2012
    Messages:
    2,869
    Yes, getting the callstack for the crashdump, generating a HTML overview and then ordering into the crashing filename. Was quite some work to get it working like that, but without this tool it would be even more work to sort the things out by hand.

    Also, currently trying to fix fullscreen mode up in special cases, funky results ensured:
    window_mode.png

    - - - Updated - - -

    added to first post:
    that being said, its WIP
     
  3. Goosah

    Goosah
    Expand Collapse
    Global Moderator
    Staff Member

    Joined:
    Aug 30, 2013
    Messages:
    781
    Last night I worked with Estama, testing the implementation of anisotropic periphery beams in the pressureWheel, with the aim of reducing rolling resistance. My test rig concluded there was no significant benefit :/ Tonight I have something new from him to test!
     
  4. TheAdmiester

    TheAdmiester
    Expand Collapse

    Joined:
    Aug 7, 2012
    Messages:
    527
    This all sounds very oddly satisfying.

    Do you have any screenshots or whatever about the inner workings of this tool or does it involve game stuff that you don't want to disclose?
     
  5. tdev

    tdev
    Expand Collapse
    Developer
    Staff Member

    Joined:
    Aug 3, 2012
    Messages:
    2,869
    Fullscreen fallbacks finally working:

    [xcode=plain]
    1.068737|Attempting to create GFX device 0: NVIDIA GeForce GTX 750 Ti on #1 (D3D9) [with output \\.\DISPLAY1]
    1.069214|shader model: 3.0 , available monitor size: 1
    1.076946|Device created, setting adapter and enumerating modes

    45.937583|AntialiasLevel: 4,BitDepth: 32,FullScreen: 1, RefreshRate: 60, Resolution: 1920, wideScreen: 1080
    45.938148|Win32Window::setVideoMode - invoking curtain
    45.967342| Failed to create hardware fullscreen device, trying windowed mode

    45.968341|AntialiasLevel: 4,BitDepth: 32,FullScreen: 0, RefreshRate: 60, Resolution: 1920, wideScreen: 1080

    45.975015|Win32Window::setVideoMode - invoking curtain
    45.988670| moving window to: 0,0 / 1920,1080
    46.129737| Cur. D3DDevice ref count=1
    46.130502| Pix version detected: 3.000000
    46.132128| Vert version detected: 3.000000
    46.132560| Maximum number of simultaneous samplers: 16
    46.132993| Number of simultaneous render targets: 4
    46.135245| Hardware occlusion query detected: Yes
    46.138107| Using Direct3D9Ex: No
    46.183794|WMIVideoInfo: DxDiag initialized
    47.458568|Initializing GFXCardProfiler (D3D9)
    47.458987| o Chipset : 'GeForce GTX 750 Ti'
    47.459461| o Card : 'RDPUDD Chained DD'
    47.460208| o Version : '9.18.13.4709'
    47.461007| o VRAM : 2048 MB
    47.461351| - Scanning card capabilities...


    ...


    48.404512|--------------
    48.405079|Attempting to set resolution to "1920 1080 false 32 60 4"
    48.408730|Accepted Mode:
    48.408753|--Resolution : 1920 1080
    48.408772|--Full Screen : No
    48.408780|--Bits Per Pixel : 32
    48.408787|--Refresh Rate : 60
    48.408794|--FSAA Level : 4
    48.408801|--------------
    48.409631|AntialiasLevel: 4,BitDepth: 32,FullScreen: 0, RefreshRate: 60, Resolution: 1920, wideScreen: 1080
    48.410180|AntiAliasing has been disabled; it is not compatible with AdvancedLighting.
    48.411102|AntialiasLevel: 0,BitDepth: 32,FullScreen: 0, RefreshRate: 60, Resolution: 1920, wideScreen: 1080
    48.412674| moving window to: 0,0 / 1920,1080
    [/xcode]

    - - - Updated - - -

    Its a quite complex python script up to now. No time to explain it further now, got to fetch some sleep

    - - - Updated - - -

    OK, worked all night through: refactored logging: new debug levels (debug + info), improved file logging (and log rotation) and improved console logging.

    torque3d.log now with start time and log level (D = Debug, I = Info, W = Warning, E = Error)
    [xcode=plain]
    //----------------------- 2/9/2015 -- 09:03:24 -----
    0.850217|D|--------- Loading DIRS ---------
    0.852234|D|Executing art/main.cs.
    0.853372|D|Executing scripts/main.cs.
    0.856132|D|Executing core/main.cs.
    0.859464|D|Executing tools/main.cs.
    0.862955|D|
    0.863151|D|--------- Parsing Arguments ---------
    0.865965|D|Executing scripts/client/defaults.cs.
    0.866994|D|Executing core/scripts/client/defaults.cs.
    0.875414|D|Executing settings/game-settings.cs.
    0.876971|D|Executing scripts/server/defaults.cs.
    0.878220|D|Executing core/scripts/server/defaults.cs.
    0.879561|D|Executing scripts/server/prefs.cs.
    0.881546|D|Executing core/scripts/client/core.cs.
    0.885219|D|Executing core/scripts/client/audio.cs.
    0.890714|D|Executing core/scripts/client/canvas.cs.
    0.893439|D|Executing core/scripts/client/cursor.cs.
    0.896252|D|Executing core/scripts/client/persistenceManagerTest.cs.
    0.902928|D|Executing core/art/gui/profiles.cs.
    0.940281|D|Executing core/scripts/gui/cursors.cs.
    0.942858|D|Executing core/scripts/client/audioEnvironments.cs.
    0.958400|D|Executing core/scripts/client/audioDescriptions.cs.
    0.963432|D|Executing core/scripts/client/audioStates.cs.
    0.967812|D|Executing core/scripts/client/audioAmbiences.cs.
    0.970106|D|Executing core/scripts/client/oculusVR.cs.
    0.973382|D|Binding server port to default IP
    0.973812|D|UDP initialized on port 0
    0.974046|D|no Oculus Rift support :(
    0.974459|D|Attempting to create GFX device 0: NVIDIA GeForce GTX 750 Ti on #1 (D3D9) [with output \\.\DISPLAY1]
    0.974746|D|shader model: 3.0 , available monitor size: 1
    0.982777|D|Device created, setting adapter and enumerating modes
    0.984026|D|AntialiasLevel: 4,BitDepth: 32,FullScreen: 0, RefreshRate: 60, Resolution: 1024, wideScreen: 768
    0.988488|D| moving (centering) window to: 448,141 / 1024,768
    1.101577|D| Cur. D3DDevice ref count=1
    1.103517|D| Pix version detected: 3.000000
    1.104652|D| Vert version detected: 3.000000
    1.105823|D| Maximum number of simultaneous samplers: 16
    1.108223|D| Number of simultaneous render targets: 4
    1.109518|D| Hardware occlusion query detected: Yes
    1.110163|D| Using Direct3D9Ex: No
    [/xcode]

    This is how the windows console looks like (colors, error counters, no debug messages)
    logging.png

    - - - Updated - - -

    OK, refactored complete logging subsystem in T3D and libbeamng. Should provide good debug and proper user feedback now :)
     
  6. B25Mitch

    B25Mitch
    Expand Collapse
    Level Designer
    Staff Member

    Joined:
    Aug 27, 2013
    Messages:
    268
    Just finished UV unwrapping the Ibishu 200bx. This is a very tedious task in which I have to map every face on the model to a space on a texture map, 4096x2048 pixels in this case (textures are created in double resolution and then scaled down).


    (imported from here)

    Each UV island requires a small margin around it to avoid conflicting with the adjacent UV island, and space must be conserved as carefully as possible to get the best quality out the texture resolution.

    Fun fact: U and V are special coordinates describing texture space, kind of like X and Y :)

    For those of you who are wondering, the 200bx that was leaked last week isn't at all representative of the quality we're currently producing - that was an outdated mesh from last year, used to aid jbeam creation. Rest assured the model quality will match or exceed that of the Sunburst.
     
  7. IBsenoj

    IBsenoj
    Expand Collapse

    Joined:
    Sep 28, 2013
    Messages:
    833
    Wow. That is amazing, I had the most epic music playing right when that picture loaded lol. Anyways, Thanks for the update! Amazing work once again, you guys never let us down. :)
     
  8. monte379

    monte379
    Expand Collapse

    Joined:
    Dec 24, 2013
    Messages:
    792
    Wow might exceed the Sunburst model you Devs are the greatest!!:)
    Nice work Mitch!!
     
  9. tdev

    tdev
    Expand Collapse
    Developer
    Staff Member

    Joined:
    Aug 3, 2012
    Messages:
    2,869
    still working on this massive change ...
    log.png
     
  10. Hati

    Hati
    Expand Collapse

    Joined:
    Aug 5, 2012
    Messages:
    1,366
    that -- looks nightmarish
     
  11. TYLERTJ

    TYLERTJ
    Expand Collapse

    Joined:
    Aug 18, 2013
    Messages:
    677
    Really looking forward to the 200BX, but I think the T series should be a priority. AFAIK, it's complete other than some textures and interior, no? Anyway, progress has been REALLY great lately:)
     
  12. monte379

    monte379
    Expand Collapse

    Joined:
    Dec 24, 2013
    Messages:
    792
    It looks like the 200BX is going to deform nicely:cool: And I'm loving the progress lately:)
     
  13. SoHigh00

    SoHigh00
    Expand Collapse

    Joined:
    Nov 11, 2013
    Messages:
    506
    Lucky Gabester escaped this time :p Now seriously... Really awesome work Mitch, as it looks really nice and clean :) UV mapping really takes patience but it' so rewarding to have a near perfect map for texturing. I can't wait to see what you do with this one!
     
  14. Dummiesman

    Dummiesman
    Expand Collapse

    Joined:
    Sep 17, 2013
    Messages:
    4,271
    Hm, what kind of massive change?

    (hopefully not too far off topic)

    To expand on that, U an d V are typically points between 0 and 1. 0 being the left (in U coordinate) or top (in V coordinate) of a texture, and 1 being the right and bottom respectively. For example, here is a part of the texture with the UV coordinate listed


    (imported from here)
     
  15. TheAdmiester

    TheAdmiester
    Expand Collapse

    Joined:
    Aug 7, 2012
    Messages:
    527
    So UV coordinates are essentially the percentage covered of a square up to the designated point?
     
  16. Aboroath

    Aboroath
    Expand Collapse

    Joined:
    Aug 25, 2013
    Messages:
    3,789
    This is sheer awesomness. Always wondered what UV meant in relation to texture mapping. I understand a bit more now when someone says this
    process is a headache. I can assume by the sheer detail in the 200BX texture "sheet"?....this thing is going to look stunning.

    Question: When scaling down the size does the resolution go with it? Great posts guys.
     
  17. Dummiesman

    Dummiesman
    Expand Collapse

    Joined:
    Sep 17, 2013
    Messages:
    4,271
    Basically yes, can be a rectangle if the texture size ratio is 2:1.
     
  18. AdR

    AdR
    Expand Collapse

    Joined:
    Feb 6, 2014
    Messages:
    306
    So you'll end up with a 2048x1024 texture for all the metal parts, right? that's too small for such amount of small parts in there. Some will just lose their detail on the resolution change.

    Why not using one 2048 map for the body (all exterior parts, plus makes a good template for future skins) and another 2048 for the inner metal parts, to keep the detail?. The normal map will look a lot better like that, specially on the seams and small details like emblems
     
  19. gabester

    gabester
    Expand Collapse
    Vehicle Artist
    Staff Member

    Joined:
    Jun 6, 2012
    Messages:
    2,675
    2048x1024 with the UVs he's shown above will be plenty. The existing cars like the Bolide, Grand Marshal, and Moonhawk use that size texture for their body/underbody/frame/suspension and look fine.
     
  20. crimsonskull

    crimsonskull
    Expand Collapse

    Joined:
    Jan 1, 2013
    Messages:
    246
    umm, where is the other side of the car?
     
Thread Status:
Not open for further replies.
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice