Next Gen Simulation

Discussion in 'General Discussion' started by vegaguy5555, Feb 3, 2018.

  1. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    712
    Likes Received:
    202
    You think totally, but then totally different from me on this point.
    But I fully respect your point of view, which is not mine.

    Each user is different, and these needs are different.

    I'm really focused on realism :

    DX9 graphics totally in ultra with AA at most suited me perfectly (really).

    Suddenly, the improvement of realism (not only physics, realroad, laser, wet road and aquaplanning, etc etc etc) is for me, simply the reason for being RF2 (for me).

    On the other hand, the GPU + CPU performance is an extremely high priority too :

    1. S397 must not leave aside some of these users. It would be really not normal and not correct.
    2. Then everyone will understand that the very low latency + the perfect fluidity + the absence of stuttering + the absence of lags ................................... is just 1 billion times more important that the beautiful graphics, because we are not in a video game itself, but a real hardcore simulation, which I think targets the driving enthusiasts (who privilege realism) and not the players of video games (which privilege immersion, ergonomics, visual etc. and generally choose PC2 / AC / Raceroom).

    I let you think about what developments are very widely favored since the arrival of S397 :

    For these two reasons that I prefer (realism + performance), I am extremely frustrated.

    We are all different / we all have different expectations :

    That's why some of the users of RF2 are delighted with the developments made or planned by S397.....................for another part (less numerous apparently), there are a lot of frustrations and deceptions.
     
  2. patchedupdemon

    patchedupdemon Registered

    Joined:
    Jan 3, 2017
    Messages:
    1,602
    Likes Received:
    1,151
  3. burgesjl

    burgesjl Registered

    Joined:
    Feb 2, 2013
    Messages:
    344
    Likes Received:
    540
    Well, let's look at four areas for potential improvements; physics, graphics, game environment, online.

    For physics, the basic models haven't changed in about 5 years. The tire models evolved (brush models plus iRacing fundamentals approach) but they seem to have reached a plateau, and still they do not respond properly at or over the 'grip threshold'. Track surface (heat, rubbering, rain/puddles) continue to evolve. But there's still something fundamentally wrong with some of the physics; cars behave very poorly in some situations that is not life-like and are still quite simplified, and especially for collisions. Given limitations in CPU speed increases, I don't see how this can fundamentally change (unless someone develops a parallel physics engine) so we are on a path to continual refinement in all these areas. Frankly, this is the most disappointing area for 'next gen'.

    Graphics have taken leaps-and-bounds towards being more photorealistic, and when we all have 4K screens with HDR or VR headsets with 4K-per-eye then we'll know we have made it. There's certainly some improvements like 3D trees that would be nice, but we've got track surfaces with high enough polygon counts and all the surrounding buildings etc. to be pretty much spot on here. Lighting and HDR will continue to improve. So again, I think this is basically continued evolution for 'next gen'.

    By game environment, I mean giving offline racers the game modes (career modes, etc.) for immersion this way, and AI to drive against. I actually think AI is the area most ripe for developments. We know the AI 'intelligence' in racing (training to race like a human would) is still not there, they also don't make the same mistakes we do. Also, they have to use dumbed-down physics. Game modes are usually easy to program and just time consuming. Big area for 'next gen'.

    Finally, we have online. This is both the racing itself, and the competition infrastructure. For racing, there is only so much you can do to the latency of the internet; our pipes get bigger but latency isn't improving much, and we still get losses and out-of-order packets and such due to the fundamental design of the protocols (which guarantee delivery, not time). See Q1 on collision detection and Q3 on AI for how that aspect of things will get improved. So the area for development is competition infrastructure, and its still amazing to me that after 9 years, iRacing is the only company to have truly developed this. For everyone else, its embryonic or 'DIY' from some very basic dedicated server capabilities. It's use of the Cloud. Here is 'next gen'. iRacing has lots of faults and they are very lazy in thinking of how to organize competition; its ripe for innovation. Thousands and possibly millions of sim racers around the globe, but we can't figure out how to organize things to get them all together to compete against one another. Leagues with 10 regular racers; servers empty.
     
    vegaguy5555 likes this.
  4. vegaguy5555

    vegaguy5555 Registered

    Joined:
    Jan 28, 2017
    Messages:
    960
    Likes Received:
    505
    The 917k is one of the cars that feel good both here and AC but are yet quit different over all in either sim. I would think if two separate sims are 100% real would feel the same in the same car?
    But when I sit back and think of all the detail that is missing from this old square, bias ply tire rough machine its hard to be sold on the effect. I know the programmers can only program so much power through any wheel but the reaction of all this missing detail would still have an effect.
     
  5. ECAR_Tracks

    ECAR_Tracks Registered

    Joined:
    May 1, 2016
    Messages:
    456
    Likes Received:
    465
    Funny fact is Microprose GP1 had those in 1990@16 bits
     
    Louis likes this.
  6. Nielsen

    Nielsen Registered

    Joined:
    Feb 29, 2016
    Messages:
    181
    Likes Received:
    60
    My focus and criticism of the current racing sims is aimed solely on their tire models and more specifically on the way different sims loose grip and (sometimes) regain it.
    In RL racing this is the absolutely crucial part. And the reason a majority of track day entusiasts get their kick.;)
    Thats also the reason I after 5+ years just before christmas ended my iceRacing membership.
    The iceRacing TM cannot cope with driving on the edge - because the regaining part is completely flawed(IMO).:rolleyes:

    At the moment Im loving Racerooms TM because this swedish team has more or less nailed this part(again IMO).
    In my book the cars in rF2 and pCars2 behave different but both (somewhat) believeable when the cars loose grip and (often:) regain it.
    Conserning this "postulate" that there are "possibly millions of sim racers around the globe" then I doubth it.
    If we count all the ACTIVE users of the different socalled racing sims together then we get a much lower total number.
    And worse they are destibuted/deluted over so many "sims" that each sim can only hope to be profitable niche.
    Hehe I dont know if it would be possible but if people with all different HW and sims(!) could compete online in the same races - on equal conditions(!) - then we could see a real racing sim ESport emerge. ;)
     
    patchedupdemon likes this.
  7. Louis

    Louis Registered

    Joined:
    Dec 27, 2016
    Messages:
    1,379
    Likes Received:
    840
    Yeah, but they didnt had working wipers... :D
    ...and realroad, third party mods...
     
    ECAR_Tracks likes this.
  8. PLAYLIFE

    PLAYLIFE Registered

    Joined:
    Dec 26, 2010
    Messages:
    743
    Likes Received:
    126
    The limitation to gaming is the talent of the people who are building physics models. I doubt the most talented people in this field are working on public software.
     
  9. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    712
    Likes Received:
    202
    Of course.
    And the Michael Borda team is incredibly talented.

    But also especially the time spent on physics (and all other elements of realism).

    You should not waste your talent by focusing on elements that are not directly related to the simulation aspect (do not leave them totally out of course).

    There are video games for this like the pseudo competitor Project Cars 2.

    Unfortunately, each editor must choose these priorities not in what is essential (simulations = priorities should systematically be the elements related to realism), but according to what these users want.

    There is nothing more than to hope to be part of this majority, otherwise you are frustrated, and in the case of RF2, you have no other alternative choices (hoping that the future competitors of S397 will huge jump forward on the simulation aspect.).
     
  10. Filip

    Filip Registered

    Joined:
    Feb 9, 2017
    Messages:
    1,208
    Likes Received:
    937
    The limitation to gaming is also average user hardware.
    I wouldn't be surprised if public software is more optimized than private specific one. Public one has to work without bugs everywhere, I presume it makes quite an impact on development.
    And then there is money involved.
    For multi-million professional simulators they can probably afford all kinds of tests and measurements to collect data but game devs have to compensate with their creativity.
    So I don't know if that talent claim holds.
    I could be wrong though.
     
  11. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    712
    Likes Received:
    202
    I think that by "talent" he meant:

    Knowledge / skills, associated with a certain passion.
    I understood it like that.

    And I think ISI / S397 has that strength.
    And most importantly, I do not think the other publishers have it, at least not so much.

    Conversely, other publishers have forces that S397 does not have. An example among others:

    Capabilities to make a software / graphics engine that allows ~95% of their users to have a performance/quality ratio really acceptable in all circumstances of play (45 AI + rain + wet road and projections + hour of the day not favorable to the best performances etc etc).

    As a reminder, performance is the only thing from the point of view interaction simulation/user hardware that conditions the fact that the player will drive in acceptable conditions or not.

    Some publishers have perfectly understood (eg Assetto Corsa).
     
    Last edited: Feb 9, 2018
  12. AMillward

    AMillward Registered

    Joined:
    Apr 4, 2017
    Messages:
    1,879
    Likes Received:
    1,837
    AC doesn't require the CPU that rf2 does though becuase it's a simplified physics engine. This is why the 'my rig will do XYZ but won't run RF2 properly' argument is ridiculous.
     
  13. avenger82

    avenger82 Registered

    Joined:
    Dec 29, 2016
    Messages:
    811
    Likes Received:
    342
    Can you give some examples?
     
  14. AMillward

    AMillward Registered

    Joined:
    Apr 4, 2017
    Messages:
    1,879
    Likes Received:
    1,837
    Only car I can ever recall anyone having issues with was the 2010 FR 3.5
     
  15. avenger82

    avenger82 Registered

    Joined:
    Dec 29, 2016
    Messages:
    811
    Likes Received:
    342
    I was expecting some third party mods for sure are not quite realistic.
     
    Last edited: Feb 9, 2018
  16. SPASKIS

    SPASKIS Registered

    Joined:
    Sep 7, 2011
    Messages:
    3,155
    Likes Received:
    1,426
    I wonder why the comparisons between sims and real life are always done based on sensations and not on data... Doesn't AC have a telemetry plugin?

    It should be quite easy to determine which sim does best, both overall and in a specific area. For sure you need reliable data to compare with.
     
  17. dadaboomda

    dadaboomda Registered

    Joined:
    May 15, 2016
    Messages:
    712
    Likes Received:
    202
    Everyone here knows CPU performance is bad and PARTIALLY justified.

    All that concerns the FPS and not the lags, ie PCIe + GPU is also very poorly optimized.

    This does not justify in any way that RF2 is the simulation/ car video game has more poorly optimized.

    The performance gap with these many games/simulation is much too high, for a lower graphical result.

    My remark was not stupid because I took into account the specificity of RF2 regarding CPU usage.

    By cons what is sure is that you lack critical thinking, you say AMEN has everything that S397 does or predict, and I have clearly noticed for many weeks that you make a fixation on me. coooooool
     
  18. Filip

    Filip Registered

    Joined:
    Feb 9, 2017
    Messages:
    1,208
    Likes Received:
    937
    Will telemetry data help to determine which sim losses/regains grip realistically for lets say catching oversteer at specific slip angle when overdriving a car?
    It's an honest question, I don't know much about that stuff.
     
  19. Ari Antero

    Ari Antero Registered

    Joined:
    Jul 27, 2012
    Messages:
    1,882
    Likes Received:
    829
    I wouldn't claim that CPU performance is bad in rF2. I7 3770k was released in 2012 and it is still descent CPU to gaming and in my mind it is mid range CPU today. If you look pic below you can see that CPU load is 21% max CPU load is about -+45% with 76 AI`s. If you look other loads in pic below you might find out what is reason to why some is struggling with performance in rF2 ;)

    SLI at Nola 2.jpg

    Gpu loads I7 3770k.jpg
     
  20. AMillward

    AMillward Registered

    Joined:
    Apr 4, 2017
    Messages:
    1,879
    Likes Received:
    1,837
    I don't have a fixation on you. You just post the same thing time and time again. I've got an FX6300 and it's not a space shuttle by any means and runs fine, as does Ari's who also doesn't have the most advanced processor in the world
     

Share This Page