Graphics improvements suggestions

Discussion in 'General Discussion' started by hamiltonfan2205, Nov 16, 2014.

  1. Tosch

    Tosch Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,872
    Likes Received:
    51
    If you look at the grandstand alone you see some of the most obvious issues with the rF2 graphics.
    -The roof is way to bright because there is no ambient occlusion. It's not a problem of the engine (ao is a performance killer), the modder should simply add an ao map.
    -The crowd billboards are way to bright. The sun comes from the right side and most of the people on the grandstand should be in shadow, but billboards (tree billboards too) don't receive shadows.
    -The saturation of the textures is all over the place. This may look ok in rF1 but hdr multiplies these differences and it looks like a cartoon.

    Here is a photoshoped version of the same picture. Added AO to the roof, darkened the crowd, reduced the saturation.

    GRAB_015sat.jpg
     
  2. Gijs van Elderen

    Gijs van Elderen Member

    Joined:
    May 1, 2014
    Messages:
    1,396
    Likes Received:
    466
     
  3. Gijs van Elderen

    Gijs van Elderen Member

    Joined:
    May 1, 2014
    Messages:
    1,396
    Likes Received:
    466
  4. tjc

    tjc Member

    Joined:
    Jun 4, 2012
    Messages:
    5,842
    Likes Received:
    327
    Ok, I`ve said now at length (to some peoples, especially the ISI devs) full on boredom that the HDR in rf2 needs a serious overhaul. It causes some quite serious issues, not least the near black pixels bug(s) causing the purple/green "oil" patches as well as some other things mentioned here and in other threads but I don`t get the "blocky" shadows thing at all...

    As Tosch said above, I think there are issues with the shadows as he stated, some objects/people that should be in shadow aren`t etc but I don`t see the blocky issue some have mentioned.

    I`m not at my sim rig atm and won`t be for some time yet but I use maxed settings for everything and iirc shadows at max and optimal and have never seen "blocky" shadows (apart from certain times when they may have been "broken")

    I don`t see any blocky shadows in the screens below and they range from very early builds to much more recent ones...

    [​IMG]rFactor2 b85 by tjc - Sim Screens, on Flickr

    [​IMG]rFactor2 by tjc - Sim Screens, on Flickr

    [​IMG]rFactor2 by tjc - Sim Screens, on Flickr

    As I say, there have been times when certain builds have had shadow issues but in general I don`t see this issue.

    All the screens above are unedited...
     
  5. Spinelli

    Spinelli Banned

    Joined:
    Jan 28, 2012
    Messages:
    5,290
    Likes Received:
    30
    Visor raindrops aside, is it just me or does the track surface in that video look way more complex then what we have now. It looks like there are different levels of water on all sorts of different areas of the track - as if the water levels are actually dynamically "connected" to the physics/physical properties of the track surface in a much more complex way than we have at the moment??...
     
  6. Guineapiggy

    Guineapiggy Registered

    Joined:
    Apr 22, 2011
    Messages:
    535
    Likes Received:
    0
    Really? I can see the blockiness even through the heavy blurring. It's not pixels as big as cats stuff but it's plain to my eyes when you look at those in full 1920 x 1080. Besides, the sample is disingenuous - it's most noticable when thin or complex objects cast shadows, not large smooth ones. Those shadows really don't look much better than RF1 shadows. That aside they're way too dark and flat - there's *no* suggestion of ambient light interacting with them. The further away the shadow is from casters in such situations the lighter it should be. A track covered in shadows that are all the exact same (seemingly excessive) darkness looks kinda crappy.

    [​IMG]

    The contrast with the screenshots and this photo also demonstrates how woeful the HDR and reflections/materials are in rF2.
     
    Last edited by a moderator: Nov 23, 2014
  7. Prodigy

    Prodigy Registered

    Joined:
    Nov 2, 2012
    Messages:
    946
    Likes Received:
    59
    nVidia has Percentage Closer Soft Shadows (PCSS) system

     
  8. hammer666

    hammer666 Registered

    Joined:
    Oct 29, 2014
    Messages:
    96
    Likes Received:
    0

    That video makes me cry a little inside. :(
     
  9. stonec

    stonec Member

    Joined:
    Jun 19, 2012
    Messages:
    3,224
    Likes Received:
    1,340
    From latest build notes: Lightened up shadows by artificially boosting ambient light level
     
  10. tjc

    tjc Member

    Joined:
    Jun 4, 2012
    Messages:
    5,842
    Likes Received:
    327
    Erm, yeh... really. :)

    You make a fair point about the thin/complex shadows though, they definitely are quite blocky/jaggy but in general, for the shadows shown in my screens I think the shadows look ok...

    They are too dark and they do ( as do a lot of things in rf2 at this time) still need work of course but at some angles, purely as far as the blocky thing goes, I think they look "ok"

    :)
     
  11. Guineapiggy

    Guineapiggy Registered

    Joined:
    Apr 22, 2011
    Messages:
    535
    Likes Received:
    0
    Still don't seem right honestly.
     
  12. blakboks

    blakboks Registered

    Joined:
    Oct 20, 2010
    Messages:
    843
    Likes Received:
    30
    View attachment 14988

    I'm working on some shaders 'for fun'. I'm sure they're horribly inefficient at the moment, and they're not quite 'done' yet. If ISI "flicks a switch" (which is a huge oversimplification) and gives us reflection maps that are "blurred" (ideally by the GGX lighting distribution function), then we'll see ambient lighting get better by even more than you see here. Right now, what I have is too specular (too shiny), but you can see some things that are being talked about here--most notably, the way the shadows aren't just a solid color, and actually have a gradient to them. Anyway, just thought I'd share to show that some of these graphics issues are relatively simple to fix (if even I can do some)...however, there are some much larger issues that would take some seriously heavy re-working of the graphics/lighting/rendering engine.
     
  13. matf1

    matf1 Registered

    Joined:
    Dec 27, 2011
    Messages:
    671
    Likes Received:
    2
    Oh my! That does look sensational.

    Sent from my Nexus 4 using Tapatalk
     
  14. DurgeDriven

    DurgeDriven Banned

    Joined:
    Mar 20, 2012
    Messages:
    6,320
    Likes Received:
    42
    Looks great
     
  15. Saabjock

    Saabjock Registered

    Joined:
    Nov 13, 2011
    Messages:
    898
    Likes Received:
    14
    Which build is that vintage screenshot from?
    I'm looking at the hay bales..they don't appear to be sitting on the ground (the bottom edges are perfectly straight/flat).
     
  16. tjc

    tjc Member

    Joined:
    Jun 4, 2012
    Messages:
    5,842
    Likes Received:
    327
    Build 85 Saabjock, it actually says that under the screenie... :p (well it says b85...)

    That`s actually the build I joined rf2 at...
     
  17. K Szczech

    K Szczech Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,720
    Likes Received:
    45
    For different reflections on materials blurring would help and I would welcome such a feature. Using cubemaps for ambient lighting is a different story, though. They would need correct level of light in them, which at this point is not there. When rendering reflections rF2 uses simplified shaders, which change appearance of objects and do not support shadows.

    Using this kind of cubemap as input for any lighting algorithm may introduce more error than gain.

    I think what we see here is actually a reflection algorithm, not ambient lighting algorithm.
    Note that ambient lighting algorithm, like diffuse lighting algorithm, should not be related to eye vector, which seems to be the case in your screenshot.

    Ambient lighting based on cubemap will not produce gradient on flat surface. It will create diversity of ambient light intensity for surfaces facing different directions, but not for surface facing one way (like road). This is why these kind of approaches are called directional ambient.
     
  18. Marc Collins

    Marc Collins Registered

    Joined:
    Jan 11, 2012
    Messages:
    3,168
    Likes Received:
    162
    I have asked this before, but what would be required for rF2 to natively display the clearly superior image you touched-up with PS? Is it tweaks to HDR and shadows, etc., or a total overhaul?
     
  19. Rony1984

    Rony1984 Registered

    Joined:
    May 10, 2012
    Messages:
    488
    Likes Received:
    23
    good question. I would like to know the answer as well. Thanks to the experts here for there input.
     
  20. stonec

    stonec Member

    Joined:
    Jun 19, 2012
    Messages:
    3,224
    Likes Received:
    1,340
    I think first difference which everyone immediately sees in Tosch's processed image is the lesser saturation. ISI has stated that saturation of their HDR is as it should be, so I don't see that getting changed. Most people seem to prefer a more monochrome image (even on the ISI tracks, which have adjusted all textures for HDR). iRacing for example has an unrealistically de-saturated look, but not many complain about that. It seems it would be the safer way to go to keep people happy.
     

Share This Page