Maybe they'd just think "Holy Cow - these guys really take sim racing seriously" if they're putting this much effort into it. Or maybe they'll think that "people are really helpful here - seems like a great community". It would be different if people were saying "rF2 graphics suck". I don't see that anywhere. All I see is people trying to help one another and sharing their findings which benefits everyone.
don't be silly, i don't care a sticky post. the only thing than i see is a race behind a expensive hardware because nvidia don't solve the base problem. Actually i've got a gtx 770 and with aa x8 + sgss in x8 goes frame by frame. i use this files, looks pretty well and i'm happy with more than 80fps in a race start. if you don't care, leave some users can probe it.
Durge some times you go over the top we all do it, but you should complain to ISI about all the different threads if you want less or it cleaned up ask ISI to clean it up but don't have ago at ppl for doing the right thing. Unless Tim is in here to do it you know very well it wont get done as they don't have enough moderators. As long as it is something new on the subject it needs its own thread but yes i do agree some could be merged or posted else where. this 1 is fine.
Hi. I have tested all of the methods from other topics about nvidia aa quality and till today I've been using 2x multisampling with 2x sparsegrid super sampling becasue that fits my system the best. My card is gtx760 OC and cpu phenom 965 OC'd 4x4GHZ. So tonight I did some test about fps and different AA types. I run sim maxed out on single 1080p TV. My goal is to be over 60fps in all of the situations because I race online sometimes with 20+ people. Only thing I did not have maxed out was AA. Last few races I ran 2x2 AA type and it was fine except LeMans track which I had to lower some details in order to mantain good fps. Tomorow is new race for me on Portugal track with URD T5 mod so I tested out fps only in garage with 2 different cameras. I only use cockpit one but I wanted 2 results. Portugal track : 2x2 method - Cockpit = 69 Bump Cam = 78 AA LVL 3 only in game config - Cockpit = 84 Bump Cam = 95 AA LVL 3 + your file - Cockpit = 85 Bump Cam = 96 Funny thing is that I gained 1 frame from LVL 3 with that file But graphics quality with your file is so much better. I didnt tested 4x4 or 8x8 because my configuration is not powerful enough. Later I tested graphics on sebring because of those wires in background and here are the differences. Sorry because I cut original files. I run so many widgets from RfDynHud so i wanted to move them out and point just on that scene. View attachment 14886 View attachment 14887 View attachment 14888 Now question: Is there anything else what i can change and benefit from ? I've been reading about changing texture sharpening in PLR file? Does that works with your method and which one would be ok to run. I saw somewhere that "2" is good all around. All of those test are done on x32 version of game because RfDynHud. And last one sorry on my really bad English but I did my best
Can people tell me how this is different from using the multiple "texture sharpening" settings in the player file? Does this LOD bias file have the clean, anti-aliasing picture of "texture sharpening" @ 0 (off) but without destroying all the further-down-the-road texture details that TS @ 0 kills??... What's the difference? Thanks guys, I'd love to know.
Little update on my previous post. Ive tested texture sharpening 2 instead of 5. Everything was ok and kinda look better @ Portugal track but later I changed on Sao Paulo and it was dissaster. Track and all around was so blurry and bad looking. When i changed back to 5 everyhing was better and i could see sharp textures again. So fixed value is probably not good overall option ?
Let's go from the beginning. In rF1 there is no TS (Texture Sharpening) parameters. In my league we still using those engine based sims like rF1, GTC, ACTC. With nv3dum.dll in root folder you can clamp negative LOD, so the image got more quality. when i came to rF2 i saw the same cutting on tension wires. So i try with the same nv3dum.dll (nv3dumx.dll 64bits) and again the image got quality. Personally compared levels of TS it's close to 0=off. maybe a little more blurred but no as TS in +1.0. Personally i use this files with TS in 4, -2.0. Thats all. i'm not an evangelizer. You wanna try? ok. you don't want? thats ok too.
Ok so this file does the same thing as TS in the rF2 player file? So it was needed for rF1 but not needed in rF2 since we have TS in the controller file? Sorry, I'm still not getting it....
Clamp only impacts those parts of the image that have a negative LOD bias. It does not affect the entire image. Anything that has LOD bias greater than zero is not touched. It should work well with anisotropic filtering. I don't know how TS works in rF2. If anyone has any knowledge/information concerning TS, please chime in here. This site may be of interest to some: http://wiki.step-project.com/Guide:NVIDIA_Inspector
I'm sorry but I gave up on excusing him long time ago, he's turned off for me. What the heck, ya gotta sign up with forum police to legally start a thread, crap, just quit reading all of them Durge!!! Good grief. I know another person in flight forums, known as Snave........ he's got a problem too.
Hey man, thanks you share these files with us!! is there a way to update the patch by myself every time a new driver was installed?
You seem to miss the point as always. If there was a nvidia and amd sticky thread all this stuff would get pinned to the first post by the author and PEOPLE WOULD BLOODY FIND EVERYTHING A LOT EASIER !@! rather then 10 graphics threads a page drop to page 2 Personally after 3 years I rather the main forum be about other stuff. Just want all people to find all relevant stuff fast and easy. Is that bad.
"A kind word turneth away wrath" I long for the day when EVERY f**king thread here doesn't erupt into some form of argument... before we get to the 'meat and potatoes' of what it's about. It's getting very old...very fast. I want everybody to STOP, take a breath and look at every response..then tell me if this was worth a fight.
+1 Stick to the topic. Save other comments for somewhere other than what is supposed to be the content-related thread.
http://us.download.nvidia.com/Windows/344.75/344.75-win8-win7-winvista-desktop-release-notes.pdf Confirms Clamp setting still does not work. I presume this is permanent. For those who know more than me, why is this 344.75 release dated September, but it's not even out yet?
Instructions might help And these files will get picked-up by a lot of anti-virus scanners. Be prepared for that.
