DrivingFast
Registered
Step 1 :
Find out your real refresh rate (important step).
No, your actual refresh rate is, in the majority of cases, not 60 or 59: it's usually fractional.
Most monitors use 59.94 Hz for "60 Hz" and 119.982 Hz for "120 Hz" for example.
You can detect your actual refresh rate on these sites:
http://www.testufo.com/#test=refreshrate
https://www.vsynctester.com
NOTE: this step is essential and must be very precise and done with great seriousness.
- Let the test run for a while (normally 5 minutes is enough but I suggest 15 minutes).
- Do not run both tests at the same time.
- No background tasks should run.
- Above all, do not forget to turn off the "put on stand by the screen", or simply turn off the "paused screen" (sorry for the bad translation).
- It is important to use a browser that works and update (Chrome and Firefox should be correct).
- If you use Windows 7, Aero must be activated (the test requires the Aero vsync to detect your refresh rate). I believe this is the same for Windows 10.
- Make sure GPU acceleration is enabled in your browser, otherwise you will get wrong results.
- On testufo I think it is necessary to activate the full screen mode (there is an icon).
- I suggest not touching anything for example with the mouse.
If both sites give you very close results, you can be sure that the number you got is very accurate and very close to your actual refresh rate.
I suggest, but it's me, to do the test at least 3 times 15 min, warm TV and healthy windows/computer.
You need only the first three decimals of the number detected :
If the result is something like "59.940875" for example, you can consider "59.940" as the number to remember.
Write down this number and keep it for step 2.
2nd step :
Use the new RTSS riva tuner fractional frame limit feature.
Download rivatuner latest version.
Example:
For a 59.940Hz monitor, you ceiling at 59.930FPS (59.940 - 0.01 = 59.930). For a 119.982 Hz monitor, you ceiling at 119.972FPS (119.982 FPS - 0.01 FPS= 119.972 FPS).
How to do ??
1 / Start RIVA TUNER and create a profile for RF2 or any simu, then close RIVA TUNER
2 / Go here: C:\Program Files (x86)\RivaTuner Statistics Server\Profiles
3 / Right click on the profile, security, select your user name, modify permissions, check total control in order to modify the profile file
4 / Open the file, where there is "[Framerate]", copy the following lines by changing the first value by :
if your monitor refresh rate is for example 59.999 (figure you found in step 1) like me, make 59.999 - 0.01 = 59.989 and enter 59.989.
In the game profile file:
[Framerate]
Limit = 59989
LimitDenominator = 1000
Step 3:
Disable vsync by the game, AND ACTIVATE BY NVIDIA CONTROL PANEL (standard vsync)
Result:
Above all, much lower entry delay (latency), associated with perfect fluidity from the moment that you have at least the same number of FPS as your framerate (fluidity identical to the vsync without the big delay (latency).
You will press on a device, and with a very short delay you will have the answer by the game (then the answer to the following screen which depends on the entry delay of your monitor).
To be very clear dear RF2 lovers, you will be very slightly better in virtual driving while maintaining a fluidity, guarantor of lots of good things.
Works especially if the hardware/graphics/game combination allows to obtain the majority of the time the number of minimum FPS corresponding to the framerate of your monitor.
This will keep pre-rendering and other blank image buffers by preventing the game from rendering images faster than your refresh rate.
And blank frame buffers mean NO LATENCY added due to buffering.
If you do not do the above, setting pre-rendered images larger than 1 will cause more input delay, and may cause poor image stimulation in some games. So put "1" at "maximum pre rendered frames", that's important.
And even if you do the above, setting this value to 1 always helps to guard against temporary fluctuations in the frame rate (no frame capper is perfect), which gives an offset value of more consistent entry.
Keep in mind that all this requires a PC capable of rendering images fast enough.
As the above keeps blank rendering/image buffers, this means that there is no protection against peaks in rendering time of images.
If your PC can not maintain a frame rate of 60 fps most of the time (or other monitor refresh rate), it may not work well (to be tested).
If your PC is fast enough, then this is a great way to greatly minimize the vsync input delay and optimize fluidity by having image times really matching your refresh rate.
EDIT:
Almost the same latency than vsync off and gsync, for example :
- vsync ON uncapped: 98ms
* VS *
- vsync ON capped: 48ms (50ms reduction, which is 3 frames less latency; 16.7 * 3)
- vsync OFF capped: 34ms
- gsync capped: 39ms
It works fine for me.
This information comes from the experienced user "RealNC" on the GURU forum.
I strongly modified his post to make it easier to exploit.
Thanks to him.
Find out your real refresh rate (important step).
No, your actual refresh rate is, in the majority of cases, not 60 or 59: it's usually fractional.
Most monitors use 59.94 Hz for "60 Hz" and 119.982 Hz for "120 Hz" for example.
You can detect your actual refresh rate on these sites:
http://www.testufo.com/#test=refreshrate
https://www.vsynctester.com
NOTE: this step is essential and must be very precise and done with great seriousness.
- Let the test run for a while (normally 5 minutes is enough but I suggest 15 minutes).
- Do not run both tests at the same time.
- No background tasks should run.
- Above all, do not forget to turn off the "put on stand by the screen", or simply turn off the "paused screen" (sorry for the bad translation).
- It is important to use a browser that works and update (Chrome and Firefox should be correct).
- If you use Windows 7, Aero must be activated (the test requires the Aero vsync to detect your refresh rate). I believe this is the same for Windows 10.
- Make sure GPU acceleration is enabled in your browser, otherwise you will get wrong results.
- On testufo I think it is necessary to activate the full screen mode (there is an icon).
- I suggest not touching anything for example with the mouse.
If both sites give you very close results, you can be sure that the number you got is very accurate and very close to your actual refresh rate.
I suggest, but it's me, to do the test at least 3 times 15 min, warm TV and healthy windows/computer.
You need only the first three decimals of the number detected :
If the result is something like "59.940875" for example, you can consider "59.940" as the number to remember.
Write down this number and keep it for step 2.
2nd step :
Use the new RTSS riva tuner fractional frame limit feature.
Download rivatuner latest version.
Example:
For a 59.940Hz monitor, you ceiling at 59.930FPS (59.940 - 0.01 = 59.930). For a 119.982 Hz monitor, you ceiling at 119.972FPS (119.982 FPS - 0.01 FPS= 119.972 FPS).
How to do ??
1 / Start RIVA TUNER and create a profile for RF2 or any simu, then close RIVA TUNER
2 / Go here: C:\Program Files (x86)\RivaTuner Statistics Server\Profiles
3 / Right click on the profile, security, select your user name, modify permissions, check total control in order to modify the profile file
4 / Open the file, where there is "[Framerate]", copy the following lines by changing the first value by :
if your monitor refresh rate is for example 59.999 (figure you found in step 1) like me, make 59.999 - 0.01 = 59.989 and enter 59.989.
In the game profile file:
[Framerate]
Limit = 59989
LimitDenominator = 1000
Step 3:
Disable vsync by the game, AND ACTIVATE BY NVIDIA CONTROL PANEL (standard vsync)
Result:
Above all, much lower entry delay (latency), associated with perfect fluidity from the moment that you have at least the same number of FPS as your framerate (fluidity identical to the vsync without the big delay (latency).
You will press on a device, and with a very short delay you will have the answer by the game (then the answer to the following screen which depends on the entry delay of your monitor).
To be very clear dear RF2 lovers, you will be very slightly better in virtual driving while maintaining a fluidity, guarantor of lots of good things.
Works especially if the hardware/graphics/game combination allows to obtain the majority of the time the number of minimum FPS corresponding to the framerate of your monitor.
This will keep pre-rendering and other blank image buffers by preventing the game from rendering images faster than your refresh rate.
And blank frame buffers mean NO LATENCY added due to buffering.
If you do not do the above, setting pre-rendered images larger than 1 will cause more input delay, and may cause poor image stimulation in some games. So put "1" at "maximum pre rendered frames", that's important.
And even if you do the above, setting this value to 1 always helps to guard against temporary fluctuations in the frame rate (no frame capper is perfect), which gives an offset value of more consistent entry.
Keep in mind that all this requires a PC capable of rendering images fast enough.
As the above keeps blank rendering/image buffers, this means that there is no protection against peaks in rendering time of images.
If your PC can not maintain a frame rate of 60 fps most of the time (or other monitor refresh rate), it may not work well (to be tested).
If your PC is fast enough, then this is a great way to greatly minimize the vsync input delay and optimize fluidity by having image times really matching your refresh rate.
EDIT:
Almost the same latency than vsync off and gsync, for example :
- vsync ON uncapped: 98ms
* VS *
- vsync ON capped: 48ms (50ms reduction, which is 3 frames less latency; 16.7 * 3)
- vsync OFF capped: 34ms
- gsync capped: 39ms
It works fine for me.
This information comes from the experienced user "RealNC" on the GURU forum.
I strongly modified his post to make it easier to exploit.
Thanks to him.
