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

I guess I should read all of the thread before replaying. You guys already disputed the same stuff. Hahaha

Anywas, I see the point he is trying to make. He thinks that 1ms isn't going to matter because we can't tell the difference anyways. The thing he fails to understand is that although we may not feel the difference, the physics can. Say the car starts to slide and we have 20ms to make a correction before it is too late. Say our brain takes 19ms to respond and you have 2ms of lag. That 2ms puts you outside that window. It is now too late. The fact is our slow brains can't start working until the image is on the screen but all the while the physics keep running in real time.

 

19 ms thats impressive, more like 200 ms

 

NV and AMD are not adding input latency- they are fixing the problems with stuttering and frame dropping. PCPer has an article that explains the problem of micro-stutter. So Chicken Littles of ISI, take some advice from Douglas Adams- DON'T PANIC

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-New-Graphics-Performance-Metric

 

Coincidentally I just watched this video. Walter Röhrl on the Nordschleife with his old Porsche 911 S:

At the end at 3:20 he says, in german though, I try to translate:
"Here the experience (him) being a rally driver comes to play, who is always cautious to immediately detect and immediately react. If your only a bit to slow you have no chance to catch the car."

Now there is the question what is a significant lag. A friend of mine doing races in real life too and often drives on my RF2 system (with +25ms more lag than my very low lag monitor). When trying for the first time he said that there is something wrong, he is always to late and can't catch the car. "It seems there is a big lag?". He doesn't know anything about computers, racing sims and input lag!

My guess: Every lag is significant.

EDIT: LOL, just noticed the video does have subtitles.

 

Very interesting.

 

The lower the better - it's mathematical fact! Just like one elderly man commented my mosquito trap. He said some people are saying those are not working but in his opinion it's mathematical fact that when you kill one mosquito then there is one less!

 

Exactly exactly exactly, there is no argument here. Math proves it, peoples experience prove it, the physics of how the car seems to behave relative to your inputs proves it, it's just proven lol.

 

What can we do, besides having a faster card and thus higher FPS, to decrease input latency?

What settings, nvidia settings for example, do people find optimal and why (SLI here)

 

The most important things to reduce the input lag are

1. The model of monitor/tv itself (do your research!!!, and remember that the much often advertised response time numbers have nothing to do with input lag)

2. Making sure vsync is disabled

3. Playing at as high fps as you can. I always shoot for locked/limited to 120 fps, or 90 fps if I just don't got the power for 120. I have s 60hz monitor only, but this still applies.

4. Haven't done this with RFactor, but on some games (like Battlefield Bad Company 2), it has a pre-rendered frames setting. It was stock on 2 or 3 (I think 2), setting this to 1 dramatically reduced a lot of input lag, however there was so much input lag to start with before I changed this setting that i think this was more a game specific thing. However keep "pre-rendered frames" in mind though if you see it in other games.

5. Not using Nvidia SLI (ever since I heard about this 1 full frame delay that NVidia uses in SLI situations to avoid stutter). I am a big multi-GPU fan so that's why I'm so disappointed in this.


3 extra points:

- I hope NVidia give us the option to disable this "1 full frame delay in SLI situations to help reduce stutter" option. I would then leave it enabled for most times, for most games, and just disable it when racing.

- Again, if AMD also uses this tactic in their new drivers, I hope they give the option to enable/disable it.

- I heard that this implemented NVidia multi-GPU frame delay thing goes up to 2 full frames when using 3 cards!!! And to 3 full frames when using 4 cards!!! WTF?!! I hope this isn't true, if it is then please give us the option to enable/disable.

 

6. Don't use "windowed mode" in rfactor2 video setup

 

My numbers were simply for discussion. I forgot to put that disclaimer in there for those that fail to follow the logic.

 

Actually their is no point following logic based on incorrect data. Using 19ms when the average human reaction time is close to 200 skews the data and since this incredibly bad 16.7 ms lag time is so crucial you would think the accuracy of the data being collected would be important so we get an accurate picture of the severity of the problem

 

I prefer smooth over tearing and stuttery video any day, also your screen wont render any faster then your refresh rate so any fps above that causes tearing as the frames arent synced. But 120fps on a 60hz screen still only draws 60fps so your not going to see any more frames no matter what you think. This means there is a lag between frames of 1/60th of a second so anything that happens between each frame being displayed on your screen wont be shown until the next one is drawn. Sure there may be some slight delay using vsync but there are also more delays in your own reaction time and then also there is lag from control inputs.

I drive sims for a bit of fun and if it makes me a 10th slower but have a great smooth and detailed image id prefer that, over turning everything down and getting torn up ugly graphics at 1000fps on my 60 hz screen rendering so many images my screen cant even display.

 

Also proves my many years standpoint on automatic upgrading to newest drivers - don't do that! If you don't experience major issues with your current graphics drivers, just keep them. Newer drivers may solve some older problems but almost always introduce new too!

 

Please guys. Stop putting human reaction time into this discussion. You're really just embarrassing yourself.

Human reaction time comes after the image is on the screen. If the physics engine is running ahead of the image then you're already at a disadvantage before your slow reaction even enters the equation.

This topic comes up over and over and over. In all forms of games and it always ends the same way. People can't wrap their heads around the concept and miss out on the advantages of turning off VSync. Then you have those open minded people that spend the time to minimize lag and come back saying how much it improves consistency.

Now, let point you to a video, you tell me if lag is undetectable.

http://youtu.be/WjHqXKqeDgE

Another thing. The benefits of 120fps on a 60hz monitor aren't really so much about lag. The FFB in rFactor is tied to the frame rate. So there is that advantage. The other advantage is you have much less tearing at 120 then you do at 60.

 

Also, lag doesn't necessarily make you slower, it makes you less consistent. I have accidentally turned on VSync before and thought I must have woken up that day with a brain tumor. You have to be as close to perfect sync as possible if you want to consistently save the car when pushed too far.

 

Realistic wheel lag for skilled drivers hey ..............then freely use virtual mirrors, fictional readouts, headphones, tmpspeak and 4G.





Oh dear very realistic

 

120fps makes a difference even on a 60hz monitor, you can't see the difference, but I am not talking visual, I am talking about controller and pc/game communication

 

Spinelli

Yes your right, 60hz or 120hz, shoot for a fps (80% of the time ) of 90+ (+100 preferable ) & your sim feels solid & connected

Turning down graphical detail/settings is a SMALL price to pay for the upside here how ever much you like eye candy

Unobvious bottle necks in pc hardware (how ever high spec & over clocked our kit is ) can be the weak point in our sim racing experience

.....another has finally seen the light (-;

 

So many experts proclaiming they can tell the difference between ms.......in fact almost every person here claims they can always tell the difference. Strange. I would love to see the empirical results. Just like wine tasters I expect - FOS in most cases. I wonder why there is no linear connection between who has what hardware (or FPS), and their position in a race. Why there IS such a thing as a great gamer (who can beat you on "slower" hardware). There is probably not a person here who can click a mouse within 150ms of a cue.

Learning how to play a game, is so much more important than gaining 1fps. But, argue away.

I recommend working on your own reaction times, and how far ahead you are driving, rather than screaming in capitals about driver manufacturers.