NVIDIA (and possibly AMD soon) PURPOSELY ADDING INPUT LAG!!!!!!!

We should all be on 2ms gtg and 120hz monitors these days anyway (and yes using vsync) - that is where the tech is at. At no point did I argue that lag makes no difference - like it seems so many tried to put on me. All I said is that the 1fps functional solution - as implemented in the drivers, by the manufacturers, to keep the video presentation coherent - will not be something that can be noticed. And since we will all be on the same level playing field - it makes even less of a difference. And therefore pants CAN stay on.

I repeat - pants CAN stay on. And I stand by that and all other statements.

 

To me additional frame (16,7 ms) from drivers would prevent from getting/using that kind of system. 16,7 ms additional lag can be seen visually, 8 ms is starting to be invisible but I think even 4 ms can be felt from difference in driving experience (increased confidence and ability to be on limit/catch slides better).

 

If using VSync, yes.

 

In terms of monitor induced input lag, a CRT is 0.001 ms of input lag pretty much. However, even with VSyn disabled, there is still lag from hardware communication, ffb motors, etc etc, but yes, in terms of monitor, CRT is still the only way to be at, or just about, 0 ms. Oh by the way, from what I read so far, OLED will NOT be an improvement in input lag relative to current day monitors.

- Do your research on monitor/TV input lag tests (a few different methods to do these tests, some arent as reliable as others), so just research on a low input lag one, and get that one (my current Samsung PX2370 averaged 3.3 ms of input lag according to a very reputable website's tests)
- Play at 120 fps (ideally using 120hz as well, but even if only using a 60hz monitor), or minimum 90.
- Disable Vsync
- Then comes the other things we mentioned (pre-rendered frames, sli frame latency thingy, etc), those might be game/hardware dependant though, but the first 3 things listed can be applied for everyone.

There is also this Windows Timer Fixer thingy that made a difference for me, supposedly Windows 7 has some kind of timing thing messed up compared to Win XP (fix not needed for Win XP), I will start a thread on it later.

 

Yes please! Every lag improvement helps (especially on my higher lag system).

 

seems ?



Show me the money.......


People say it is better smoother more responsive I do not doubt.

Shows me a personal long standing laptime improvement as evidence.

Improved responsiveness and steering then you should be more consistent and eventually faster.

 

That's exactly what i was, more consistent/better at controlling and holding the car when at/near the limit, especially when very quick controller inputs were required.

When I went from my CRT to a 20ish ms input lag IPS monitor, I all of a sudden wasn't correcting the rear on time, was tending to over correct and snap the other way, just felt like my steering was always just a tiny, tiny bit behind or something. It was driving me INSANE. I spent probably 2 weeks checking settings, drivers, re-installs, EVERYTHING. IT EVEN MADE ME GO OUT AND BUY A NEW WHEEL. Getting the G25 definitely seemed to help A LITTLE BIT, which leads me to believe that my previous wheel (Black Momo Racing) had a bit more lag/latency than the G25, HOWEVER it still wasn't as good as before. This is what led me to studying input lag, I finally decided to get a new super low input lag tested monitor (Samsung PX2370 @ 3.3 ms) and VOILA, huge improvement. I then installed Momo Racing instead of G25, again, huge improvement over previous monitor.

I thought I was going crazy, or that there was something severely wrong with my equipment. I was so convinced because of being able feel the lag, that I even ended up buying an expensive G25 (close to $400 back then). This was before I barely even knew about input lag, so it was 100000% not just in my head. I would have just saved my money and not bought a G25 if I knew it was a monitor issue.

 

I did lots of back to back testing. There is absolutely no question it improves consistency. Not necessarily raw laptimes but concistency. If you are an offline/hotlaper type you may not care. But if you are doing long races with humans that do unpredictable stuff that you need to respond to, that I where it really helps. It is odd we are arguing this for so long. In every league I have raced with when the topic comes up on the league forum everyone is like, "oh yeah, you didn't know about that? Duh!". It is common knowledge there.

I actually never used VSync for the longest time, not because I knew it caused lag, but because I was one of those guys who wanted to see the Framerate counter really high. Then it went through a phase where my machine was crashing a lot so I decided to turn on VSync to restrict the Framerate and possibly reduce heat. I went from finishing in the top 3 all the time in my league to finishing 5 if I even finished at all. This went on for a few months before I found a thread on RSC (the old RSC) about VSync and Niels Heusinkveld pointed out some scientific testing of his own that measured the lag using a webcam and a game pad with a rubberband. He would then inspect the video to see how many frames of lag there were. I always ran with the animated wheel off but in his test video he had it on as a way to prove the lag. So I turned my animated wheel on and sure enough mine was lagged. So I turned off VSync and the lag was gone. Animated wheel was in sync with my real wheel. So then I started using "Max Framerate" to cap it to 60 which also combats the tearing. After that I went back to the front of the grid.

VSync was something I discovered back in 2008. I have been trying to warn people every chance I get. I normally just get the middle finger.

Again, if you are a hotlapper or the type that only races the predictable AI then who cares. If you do real competitive racing with humans then you are going to seriously benefit. I do 12 hour endurance races, so consistency is more important than anything else.

I could probably even dig up the results from the few months I was running VSync so you can see the contrast in my performance. I will do some digging.

Edit: okay it was 2009 that I discovered input lag, not 2009 and it was NoGrip where I discovered it, not the old RSC. Here is the thread:
http://www.nogripracing.com/forum/archive/index.php/t-162910.html

Now I have an exact timeframe to go back into my results to find the improvement in my performance.

 

i noticed micro-stuttering in rf1 and followed noel's advice of maxing my framerate to 60. now a get a lot more stuttering, like you see the cars almost constantly stutter!

tried disabling SLI, same result, so max_fps 60 = massive stuttering!

this all was tested in tv-cam.

 

ah, my bad, it had to be a negative value for rfactor 1 , everything is smooth now...