In my opinion the track is perfectly optimized. Like all VLM tracks there are a lot more, but much smaller, gmt files. I think this is one of the reasons why VLM tracks have a better performance than most of the other tracks. A thing that is worth to look at is the amount of objects assigned to the environment reflection mapper (road reflections). Sebring has a lot of them. Have you tried to disable road reflections?
I don`t have any problems at the new Sebring (or did I with the older version) I have solid, steady fps only dropping a couple of frames with 16 cars on the pit straight. All settings maxed... level 6 AA. I haven`t got any problems running the NSX with b946 either... Not meant in a bragging way, simply saying I`ve not encountered these problems others seem to have.
I've got an older PC, but I get good performance with this track. 44 max FPS, 41 avg FPS, 37 lowest FPS around this track with 15 of either the Clio or Honda (ISI cars). Looks nicer than the older version and runs better for me too. Thanks for this update!
Excellent update, the only minor eye sore I have noticed is in the pits some tyres on stands are very blocky, they look octagonal. Apart from that I cannot fault it.
TJC missing some detail e.g. vsync or max fps or are you running full out? e.g. I could lower my max FPS to 80 and I'll never see a frame drop on Sebring. My specs are slightly better than yours. However GPU usage will still show higher vs say Sepang...
So now I have done it I can tell you Tosch that it made little to no difference for me. Went from high on both road and environment reflections to off and was 30-40 at pit straight to 40-60 jumping all around the circuit. Tried lowering circuit detail with texture detail to medium from full. Same, almost no difference. No other track that I know of or remember from the top of my head has this kind of performance. Every other track is just sky high with fps values with max settings, no dips whatsoever. With Sebring, well it's beyond anything else. Noooooooot funny at all since it's one of the most enjoyable tracks out there. I'm gonna leave it there I guess for the time being until I figure out sth else. I might try reinstalling rF2 with all the tracks etc. However that might be probable when next build comes out. Usually do that with every odd build release. Did it with 930, autoupdated to 946 so next one is going to be fresh install again.
Hey Panigale... No v-sync - no capped fps. I`m running full on/flat out... Will check fps etc at actual race time (race start with all cars on grid) again but from memory, fps was still good/high and I don`t get any stutter of any kind at all.
I tested each of the ISI tracks plus Longford & Sebring for framerates with the Formula Renault to see how Sebring compares. Everything maxed out, HDR & FXAA, Level 8 AA, 1920x1080 resolution, no vsync, only the single car, practice session, one lap AI driving. It's not the most exhaustive test because I used only the ctrl-F counter. Monte Carlo, Brianza, Indianapolis, and Lime Rock all had low moments in the 63-66 fps range. Sebring did the best with lowest = 107 fps which occurs while you're on the racing line where pit road begins. So from that perspective, Sebring is very highly optimized. Switching cars to the NSX-R incurs up to a 20% decrease in framerates compared to the FR 3.5. BTCC Honda is about a 10% decrease. Poking around Sebring some more, the start of the race (with one car) has a 7% decrease in framerates just sitting on the grid waiting for the lights, but quickly goes back to normal at about the first turn. The low framerates at the beginning of the pit road improves after the first lap is completed, which leads me to believe it is due to memory caching. The affected area is very small, only a couple car lengths long.
And just to see if framerates scale according to number of cars, I added 9 AI. Indianapolis worst was 41 fps for NSX, 54 fps for FR 3.5. Sebring worst was 57 fps for NSX, 71 fps for FR3.5. Worst case was always while sitting at the back of the grid waiting for the start lights. Conclusion still suggests that Sebring is pretty dang solid with lowest framerates 30% better than Indianapolis.
Emery, Race is the killer on Sebring, I can get on with practice, (sans the 200 RVs) pretty well with less AI. Set to race, with some clouds, movies on, sun occlusion, crowd and env movement on, shadows maxed, etc, at 6pm. It is full on beautiful Sebring, flags waving in the air on the RVs, but this is were it shows it's appetite for GPU use. Same benchmark I use at other ISI tracks. Tosch may be on to something because I see road reflections on dry pavement to the point of showing the sign spanning above it. (Tuttle providing some input to the team would be great.) It probably all adds up since there is more detail and action taking place, camera's flashing, fireworks, etc. than ISI tracks. Again to the Virtua LM guys, huge thanks for this effort, I am very grateful, want to make sure that comes out since it sounds like a lot of complaining. It is probably going to make me get a 980 but it will be worth it!
This is my experience, too, but the reflections may be the key for less powerful video cards. Panigale, you could experiment with the difference in FPS with road reflections off, low and high at these tracks. On my high-end card, there is no FPS hit and sometimes even an FPS gain when using reflections at high!!!!! They are not properly optimized yet in rF2 and Sebring seemingly has some "extra" refectivity that may not be realistic. Combine these together if your card normally has performance trouble with reflections and there's your answer. Or maybe it's env reflections, too?
@Marc I'm using an ASUS 970 OC, what high end card are you using? Not much better than that avail save for a 980 and a Titan...
OK, just grasping for straws here. I know how frustrating it is when something doesn't work on your system that works fine on others. If reflections usually cause little to no FPS hit for you, then that's not the explanation. Not that it matters, but I am using a R9 290X, but AMD cards definitely do not perform as well in rF2 as NVIDIA.
Sebring graphics bug? Is anyone else experiencing odd issues with alpha/transparency as in the pics below? View attachment 16338 View attachment 16339
@Paningale - I understand, just confused as to why it's not working for you and it does for me, especially if I've got a 770 and you've got a 970. I'm moving to a 970 (should be here today), so I'm concerned! Here's a screenshot with 30 cars in the conditions you describe (clouds ahead have moved nearly out of screenshot). I started that session with 45 fps on the standing start, but otherwise never saw it dip below 60 fps. View attachment 16340 I let AA be handled in-game (Level 6), both env & road reflections are Low, and FXAA is off. Those are my compromises between image quality and performance. Everything else is maxed out. Normally I'm capping at 100 fps in the player.json, but for these tests I wanted to see what the range was and disabled the capping.
Hey Emery, not sure I would say it isn't working per say, I'm getting decent performance maxed race day mid to low 80s with 16 AI isn't bad at the pit. It is just that I never dip below my 105fps setting in ISI tracks. Looks like this is only the case for a couple other users. Although who knows, I have heard people say that 780s are better for some games than the new 980s. Will be very interesting to hear what you see with your 970. The MSI card would have given me a few more FPS but I liked the backplate on the Strix. LOL. I have AA at level 4, mine only shows up to level 4 in the video settings. It used to show up to level 6 or 8 but haven't seen that in over a year I think? In the options I have texture mapping at x16 anisotropic. I do have FXAA on as well. I also have sun occlusion enabled in the player json so maybe that is part of it. I double checked my mobo to make sure the PCI slot was Gen3 (my board lets you set it to gen1-3 or auto). Last idea is to dig around in the Nvidia profile and see if something is going on there by default (in the 9 series cards) that wasn't in my old card. Would feel like a total noob if that is the case. I'll also try setting sun occlusion to off. Please let us know your results when you get your 970. Cheers!
