Hello. I am changing my PC. Everything changes and everything will be very very powerful (i7 9700K, Asus maximus hero XI, big water cooling [Corsair h150i], 16 gb 4000 Mhz CL18 ......). Unfortunately, I could not change my GPU : GTX 1080 8gb VRAM. Except I think that in a number of circumstances, my FPS dropped drastically, partly I think because of the lack of VRAM of my GPU (8 gb). What I like is play at any time of day and night (so sunrise/sunset/low sun), with a timescale of at least X5, rain 50% of the time, with AI from various cars or classes (~25),...... and obviously nice graphics. I play exclusively in VR. My question is : What are the main (say 5) ingame settings + resolution/PP/AA + race settings/game settings that are the most greedy in VRAM ? So that I lower them first. Thanks for reading me.
Opponent detail above high is unnecessary, same as aa level above 3, then track detail ultra to high, number of very high opponent count ( visible or not doesnot matter at all ), also i read somewhere there is a trick with player.json that set opponent detail to low in game setting and then opponent bla bla ( cannot remember ) in json to 3 is save vram
Textures (track and opponent detail), screen resolution and perhaps AA to some degree. What you should do is monitor VRAM usage with a software like Afterburner (I think Windows 10 task manager will now include it as well) and see if you get close to 8 GB, if not, then no problem. Sunrise/sunset should have no effect on VRAM, as it's using the exact same textures, just more computation required with shadows etc.
I have never played with VR. But I think the biggest performance eaters are, like without VR, shadows and rain. I would adjust their effect settings to a low level.
I did some tests and found out AA level (Level 3 and above) and SteamVR supersampling effects vRam usage dramaticly. Texture detail and car count also have some impact but not as much as AA and SS levels. Worse thing is, no matter what i do, i couldn`t go lover than 7.0 GB usage. It is clear there is a serious optimization issue. Low vRam causes CPU times go up so, if you have performance problems and high CPU times, check your vRam usage. i7 9700K@4.9 Ghz 2080 RTX 8 GB vRam 16 GB System Ram Valve Index
Is there anybody know what the aa levels refers? I think nobody know what the level refers, even marcel donot know, it is asked before a lot but we couldnot get a proper answer yet
He means what AA ingame refer, for example : Level 1 = X2 MSAA Level 2 : X4 MSAA Level 3 : X8 MSAA Level 4 and 5 : ??????
https://www.tested.com/tech/pcs/1194-how-to-choose-the-right-anti-aliasing-mode-for-your-gpu/ There, the difference is explained quite well. But I think that depends on the monitor and graphics card, if you can even see big differences. For me, the activation of FXAA in the graphics options has been to make the image smoother and stabilize the FPS, along with the VSync setting on monitor.
We all know what is anti aliasing. But we donot know what the aa levels in rf2 refers to ( 2x msaa, 4xmsaa etc) as drivingfast shown as an example. This is asked many times and also there is a topic for it in here, in radepartment forum, it is asked in discord too, but none of the studio397 guys answer it, and i guess that they do not know what is it exactly.
Like I already wrote. It will depend on the graphics card and the monitor if you notice changes in the anti-aliasing settings. I notice that the picture is smoother if I set the shot higher. Between level 1 and 4, I notice differences. At level 5, however, I can not tell the difference to level 4.
Hello @Christopher Elliott . I think my use of rF2 poses problems of lack of VRAM : Offline, 61 AI, Nurburgring combined, day and night, rain, multiclass,....... I know you're very busy. Nevertheless, could you take 5 minutes to tell me what graphical parameters influence the most the VRAM ? I have no technical knowledge, but I have the impression that it is : Texture details, track details, cockpit details, number of AI and AI visible and can be the shadows ? Thank you.
If you monitored vram usage via afterburner like stonec suggested 3+ weeks ago, you could have tested 1 setting per day and had answers for yourself (and everyone else) a week ago. I'm not sure why you think Christopher has such a list at hand.