Thinking about improving FPS, a path to optimizing?

Discussion in 'General Discussion' started by Emery, Oct 19, 2017.

  1. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    Hi.

    For a 132cm screen, using 4K and max NVIDIA settings improve in a gigantic way overall graphics with few FPS lost : its very surprising. I have lost ~50 hours of my life to make hundreds and hundreds tests and the best result is 4K DSR + max nvidia settings + PP off + shadows off + reflections off + low AA (with 4K level aliasing is almost perfect, far better than 1K + level 8) + real mirrors off (I dont need and thats simple : at some moment real mirrors = 15 FPS lost and I make this test a lot IN REAL TIME).

    Very very strangely, the overall quality VS the overall performance (stable performance) is very very much better than all settings max + full hd + max AA : so yes its a very original approach....

    For your information increase the resolution is very very much better than full hd + strong AA : not really need better GPU but overall quality is incredibely much better
    Thats a fact but you have to test this with a lot of precision to obtain this result.

    Concerning your, effectively thats a small screen so it is a little less usefull for you perhaps.

    I am not sure you understand that with full hd + level 8 AA aliasing is not very good VS my approach there is only AA off or level 1 or level 2 so THERE IS A REAL PERFORMANCE COMPENSATION between these two strategy, and from my approach I have a strong improvement in all overall quality. Spend a lot of performance to improve only aliasing, not in all the scene, with a picture very less "beautifull" and poor overall quality for me (thanks to my hundred tests) it is very more profitable to use my approach.

    BUT I DONT CARE ABOUT MY APPROACH. We dont care about dadaboomda ego.
    I just want to help people. For me speaking english is not a pleasure, I only want to help people.

    Hope this clarification is more comprehensible ;).
     
    Last edited: Oct 31, 2017
  2. Davvid

    Davvid Registered

    Joined:
    Feb 11, 2012
    Messages:
    163
    Likes Received:
    29
    How do you turn off real mirrors without disabling virtual ones too?
     
    Filip likes this.
  3. WiZPER

    WiZPER Member

    Joined:
    Oct 5, 2010
    Messages:
    1,521
    Likes Received:
    186
    Are you skizo, Smeagol ?
     
  4. Mulero

    Mulero Registered

    Joined:
    Jan 16, 2012
    Messages:
    351
    Likes Received:
    153
    It is not possible although it is indifferent. It does not matter if you have a single real or virtual mirror or hundreds of them that the loss of performance will be the same.
     
    SPASKIS, Davvid and Filip like this.
  5. Filip

    Filip Registered

    Joined:
    Feb 9, 2017
    Messages:
    1,006
    Likes Received:
    774
    Yesterday I tried cca 15 different official and unofficial car mods on Sao Paulo, same session settings.
    I was having between 40 and 90 FPS on race start.
    Better looking mods all had lower FPS as expected with Radical GT3 having lowest FPS.
    But there was one big exception.
    With FR 3.5 I had highest FPS although it's one of the best looking cars imo.
    Is it super optimized or good fps comes from lower poly numbers?

    Btw FOV has considerable influence on FPS too.
     
  6. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    There is a command for real mirrors and fake mirrors. I had compared all mirrors off VS real mirrors in real time : huge lost of FPS between ~5 and ~15 FPS with overclocked +23% gtx 780

    I hope thats more comprehensible.
     
  7. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    Do you insult me or saying that I have a psychiatric issue in reading 2 sentences ?!

    Not cool man. And I dont understand "smeagol" ?

    Yes you upset me so if that was not the goal, clarify your statement : I just remind you that we are not on this forum for this and that life is great enough not to add.
     
  8. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    I dont use fake mirrors but I make these tests in real time I remind that :
    - real mirrors kill FPS
    - fake mirrors seemed to have very few.impact on FPS and perhaps no impact (I dont see any differences)

    But like I said, I use real mirrors off or on but no fake mirrors which I found very irrealistic (like hud, lcd, track map etc.)
     
  9. Mulero

    Mulero Registered

    Joined:
    Jan 16, 2012
    Messages:
    351
    Likes Received:
    153
    Tested right now to make sure and there is no difference between real or virtual mirrors. Only difference is between mirrors on / off.

    Drop of performance has been quite small. From 110FPS to 105FPS in Silverstone with the USF2000. I think that the loss of performance is much greater the smaller the adjustments are.
     
  10. Davvid

    Davvid Registered

    Joined:
    Feb 11, 2012
    Messages:
    163
    Likes Received:
    29
    Thanks Mulero, that's what I always thought about mirrors. And I can't imagine how one could drive without mirrors at all. It would be totally irresponsible.
     
    Filip likes this.
  11. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    Ok, what GPU have you ? Concerning fake mirrors I could agree no problem but with gtx 780 oc +23% I am 100% sure that REAL mirrors is one of the worst thing for performance for my GPU ("medium end computer")

    Between ~5fps lost to ~15fps. Most important problem : this lost of FPS is very unstable : exactly like shadows and reflections which have huge and very unstable impact in FPS (according to number of trees/objects : lot or few shadows and reflections)
     
  12. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    Totally agree for online users and users which play with AI.

    But there are other user like me which train offline without AI. Lose between 5 and 15 fps very randomly just to have 0.1 % of better immersion, while the only goal is to train....

    Just take this into account.
    All rf2 users are different and use differently rf2.
     
    Davvid likes this.
  13. Mulero

    Mulero Registered

    Joined:
    Jan 16, 2012
    Messages:
    351
    Likes Received:
    153
    @dadaboomda I have an RX570. Mirrors were also terrible for me before. I had a 5850 and used a low graphics setting to get 80FPS at 720p. Use of mirrors meant a loss of more than 30FPS. Now with a high configuration only 5FPS.
     
  14. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    If some guys want to use real mirrors but have low/medium GPU, I strongly recommand to test (I dont make the test) this : in player.json you can manage the distance of what you can see in mirrors.

    Thats probably a good compromise.

    Hope that can help ;)
     
  15. Davvid

    Davvid Registered

    Joined:
    Feb 11, 2012
    Messages:
    163
    Likes Received:
    29
    Yes, of course. When you're alone on the track that's no problem. Mirrors could be very fps hungry. I have tracks that lose 80-100fps (e.g. from 250 to 150 fps)
     
    dadaboomda likes this.
  16. Travis

    Travis Registered

    Joined:
    Dec 7, 2012
    Messages:
    187
    Likes Received:
    204
    Latest GT3 update benchmarks starting from 20th position, 19 AI cars ahead of my car. Default weather and TOD for Matsusaka.

    • Matsusaka 20 McLaren GT3 cars, start grid, fresh profile graphics settings, 1080p, AA level 2, post FX medium: 109 FPS
    • Matsusaka 20 Random URD EGT cars with custom liveries, start grid, fresh profile graphics settings, 1080p, AA level 2, post FX medium: 144 FPS
     
  17. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    Dif
    Difference is huge for just different cars mods ! That is weird.

    A thing is strongly important in a rf2 test, many people forgot :
    - not only the same weather but the most important is THE DAY TIME : have you tested with exactly the same time ? Because you can have 100 FPS at 12:00 (less shadows) and have 50 FPS at 08:30 or at 18:00.........

    To remind : shadows is the worst for performance (worst than PP, AA environment redlzction --->>> the other bad settings for performance)
     
  18. Filip

    Filip Registered

    Joined:
    Feb 9, 2017
    Messages:
    1,006
    Likes Received:
    774
    I'm getting 40 - 90 FPS depending on car mod in absolutely same conditions!
    But it has always been like that.

    And it's not only on grid. When leaving pits with nobody in sight there are big FPS differences between different vehicles.
     
    LokiD and dadaboomda like this.
  19. LokiD

    LokiD Registered

    Joined:
    May 3, 2017
    Messages:
    1,282
    Likes Received:
    1,208
    somehow I've always found mods, less fps hungry!?

    but the same is for the radical? That's not being looked at? Also the led lights not shining correctly at night.
     
  20. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    709
    Likes Received:
    200
    There is a very very very strong lose of FPS when we use ac cobra compared to other ISI cars. Especially if environment reflections are turn ON.

    So you're right !

    Sorry I dont take the time to answer to one of your post concerning real driving VS rf2. I will try to do this when I could. But I don't think.my view is very pertinent because I have for my age a "strong" experience but I made this experience with a low cost car (low powerfull but exeptional engine fortunately, very very bad chassis + suspensions + frequently I bought used tire with the worst grip of the world etc... I deactivated ABS to train.)
    But the thing to train for me like I did was a exceptional training / experience (0 grip, good acceleration, very bad brakes etc.)
     
    Filip likes this.

Share This Page