I am currently testing different settings in NVidiaControlPanel, VCStarterConfig and GameSetup to achieve that the game is displayed as smooth as possible.
Try 1
Framelimit in NVidiaControlPanel: 60
LowLatencyMode in NVidiaControlPanel: Ultra
VSync in NVidiaControlPanel: Disabled
Framelimit in VCStarterConfig: 60
Triplebuffer in GameSetup: Disabled
Result: Screentearing still happens and there are tiny lags here and there.
Try 2
Framelimit in NVidiaControlPanel: Disabled
LowLatencyMode in NVidiaControlPanel: Disabled
VSync in NVidiaControlPanel: Fast
Framelimit in VCStarterConfig: Disabled
Triplebuffer in GameSetup: Enabled
Result: Now there are still tiny lags here and there, but at least the screen-tearing is fixed.
Now I wonder what that "stuck-bug" is, that the VCStarterConfig says it fixes, when the FPS-Limit is enabled.
Who can tell me? Will I still run into problems, when i use VSync instead? What settings do you suggest?
I'm certainly not an expert but I'm going to try to explain you my way to obtain the smoothest display in Vietcong. Maybe you will find useful informations to try on your side!
Usually, I've noticed that most people are used to disable VSync to reach the highest fps as possible in game. Indeed, at 300 fps (for example), the rendering will be very good but also if it's at 100 fps... If our pc is able to maintain those 300 fps in all conditions, it won't be a problem. However, if it's not the case and if we have drops from 300 fps to 100 fps, then our eyes will perceive "stutterings" in the rendering, even though those fps are still very good!
Personally, I prefer to maintain the 60 fps in all conditions and to reach this goal, I prefer to set my VSync "on", to always display the frames my monitor can in 1 second (60Hz > 60 images > 60 fps)... Knowing that 60 fps is already a very good result for a smooth rendering. My cpu and graphic card are certainly able to deliver a lot more fps but I prefer to use my "extra power" to increase the other settings instead, like the anti-aliasing, the anisotropic filtering or the triplebuffer... As long as the 60 fps are constant, the rendering will always be very smooth and the stutterings will start only once I go under that number...
Note that if you own a 144Hz monitor, then I advise to never go under the 144 fps, that should be the "constant number", to always have the smoothest rendering... Better cpu and GC are suited in that case of course.
Notes:
-With my old NVidia drivers (2003) and VCStarter 1.70 BETA, my fps were always very good. But with my more recent NVidia drivers, the fps became suddenly very bad (drops from 60 > 30 fps!), with VCStarter 1.70 and the same settings! I had to reverse back to the 1.65 version to have better fps again but the result is still inferior to the one I had with my old drivers from 2003, unfortunately. So, the version of your NVidia drivers and VCStarter will also have a big impact on your fps in game.
-I also noticed that with some VC maps, badly optimized (or even not at all), a kind of stuttering is present even though the fps are still constant and high in game...
Some of my settings:
My monitor is running at 60Hz
Anti-aliasing setting: 8x
Anti-aliasing transparency: 8x
Anisotropic filtering: 16x
Texture filtering - quality: High quality
My VSync is set on: Fast
Triple buffer memory is: Activated
Framelimit in VCStarterConfig: 60
The whole problem is a little more complicated. Practically no one knows that VCstarter 1.65.1 and 1.70beta increase fps only in less demanding scenes. In very demanding scenes, on the contrary, VCstarter 1.65.1 and 1.70beta reduce fps (in the extreme case by more than 30%). Switching of the direct3d "vertex processing" from "software" to "hardware" wasn't most probably a good idea (sorry, Brchi, I know you meant it well). Unfortunately, it cannot be in VCstarter turned off. It is possible to use the older version 1.51, but this version doesn't include correction for 16:10 and 16:9 monitors.
Screens are from the [AT]Cutthroat_Base+ map. You can perform your own test, but I assume that the result of the comparison "VCstarter 1.65.1 vs. VCstarter 1.70beta vs. no VCstarter" will be similar. If you don't have this map, it is available here:
The whole problem is a little more complicated. Practically no one knows that VCstarter 1.65.1 and 1.70beta increase fps only in less demanding scenes. In very demanding scenes, on the contrary, VCstarter 1.65.1 and 1.70beta reduce fps (in the extreme case by more than 30%)...
It's interesting because with my older drivers (2003) and VCStarter 1.70, my fps were very good in the "demanding scenes" also and certainly better than with the previous versions... I could play at a constant 60 fps with the players LODs disabled on my VC_Storm_II (defending) map from the bunker. My guess is, it's maybe because now, VCStarter 1.70 should be updated for the more recent drivers...(?)
Please, try to perform the same test on the same map ([AT]Cutthroat_Base+) and on the same coordinates (see my screenshots). I performed the test on the two various PC configurations and the results were the same -> VCstarter 1.65.1 and 1.70beta reduce fps in very demanding scenes (in the extreme case by more than 30%). But of course, if you have an extremely powerful graphics card, you can't notice any reduction in fps.
Oh, but I trust in your tests, don't worry...
It's just that I was pointing out that with my same cpu, graphic card and VCStarter 1.70, my fps were very good with my old GC drivers only (on the same demanding map)...
1) CPU + GC + VCStarer 1.70 + VC_Storm_II + old drivers > good fps!
2) CPU + GC + VCStarer 1.70 + VC_Storm_II + recent drivers > bad fps!
So I really don't think VCStarter is the culprit of the fps drops but it's more due to the version(s) of our drivers...
And to illustrate my thought, take a look at the very interesting tests Nosek made years ago in this post... And many others noticed better fps with VCStarter 1.70 also!
Evidently we don't understand each other. OK, my mistake, my English is not good, but I will try to describe the whole problem in a different way. My term "very demanding scene" is a specific place on the map where fps values are extremely low. The VC_Storm_II map (which you mention) doesn't contain any such places. A ten-year-old graphics card is enough to play this map without any problems, and the fps values will always be good (even without VCStarter). Testing of VCStarter on similar maps is a waste of time. The tests must be performed mainly on "very demanding scenes", because this is where low-performance graphics cards have problems with fps. And I doubt that Nosek focused on such places in his tests. That's why I focused on such maps, where I knew about places with low fps. One such place is the respawning area of the [AT]Cutthroat_Base+ map. At this place I have much higher fps values without VCStarter than with it. Unfortunately, everyone is afraid to repeat my test on this map. Yes, I understand, "it's better not to know." What if it was confirmed that this problem really existed, right?
By the way, a lot of players prefer to use VCStarter in the older version 1.5.1, because with versions 1.65.1 and 1.7beta they have problems with occasional screen-tearing. They don't have this problem with version 1.5.1 because it doesn't contain the problematic "fps enhancer".
x 1
x 1
x 1