![]() |
SUBSIM: The Web's #1 resource for all submarine & naval simulations since 1997 |
|
![]() |
#1 |
Undetectable
![]() Join Date: Sep 1999
Location: Colorado
Posts: 1,221
Downloads: 132
Uploads: 0
|
![]() Perhaps because most of us just turn it up as high as we can until it affects frame rates. My slider is maxed out. Whether that is 16x or 32x to me is just semantics. |
![]() |
![]() |
![]() |
#2 |
Sparky
![]() Join Date: Jan 2010
Posts: 152
Downloads: 6
Uploads: 0
|
![]()
OK I have looked into it and (as some of you have noticed) it actually is:
0x 1x 2x 4x 8x 16x (max 16x) Where 0x and 1x mean NO antialiasing. That must be a bug which relates to querying the video card's AA settings, where 0x and 1x are listed as valid options. About the AF filter, the reason why you're not seeing any performance changes is that the texture filtering is not the bottleneck, so therefore no performance can be gained/lost from changing it. |
![]() |
![]() |
![]() |
#3 |
Chief
![]() Join Date: Oct 2010
Location: Highlands UK
Posts: 323
Downloads: 36
Uploads: 0
|
![]()
Nice one, that clears that up then!!
That said, if Max is 16XAA, then when i turned it down a notch to 8XAA i saw no difference in FPS. NOw 16X to 8XAA should make a huge difference!!....AA is normally a FPS killer! |
![]() |
![]() |
![]() |
#4 | |
Sparky
![]() Join Date: Jan 2010
Posts: 152
Downloads: 6
Uploads: 0
|
![]() Quote:
SH5 is also very CPU hungry, so if you have less than necessary CPU performance, you may put in whatever video card and settings you may like, performance is not gonna improve. |
|
![]() |
![]() |
![]() |
#5 | |
Chief
![]() Join Date: Oct 2010
Location: Highlands UK
Posts: 323
Downloads: 36
Uploads: 0
|
![]() Quote:
What you saying is right for lower resolutions, where the CPU will restrict the G-card flying at full rate, but even the most modern technoligical graphics intence game these day's is always restricted by the GPU at resolutions of 1600x1200 and above. Of course im not talking extreme mis-matches i.e GTX580 paired with a P4 1.8mhz, but guessing you have paired your graphics card with a similar quality CPU, then the GPU will be the difference at said resoloutions. For me i run at 1920x1200 res with GTX570, which will allways fall to the graphics card....my CPU is a Q9550 @ 4ghz, there will be no bottleneck from that CPU speed for any single card available today!! The I7 CPUS will yeid perhaps 2FPS difference in games, where only running SLI shows there true colours. So while i wont dispute your suggestion of 16xAA being the max setting, i would still expect to see a healthy difference in both FPS and V-ram used when lowering it a notch to 8x, which i dont. |
|
![]() |
![]() |
![]() |
#6 | |
Sparky
![]() Join Date: Jan 2010
Posts: 152
Downloads: 6
Uploads: 0
|
![]() Quote:
![]() As I said before, even if you have more V-ram, it won't increase FPS if that is not the bottle neck. As for the CPU limit, even an I7 can be the bottleneck....in SH5's case, it's not that much about processing power than it is about bandwidth transfer. The CPU is stalling the game while sending geometry data to the video card. That is a matter of lacking optimization from the game's part. It is not just a simple matter of "I have the greatest CPU on the market, it sure isn't a limiting factor because I paid a lot of money on it". There are many many factors that go into a game's performance. |
|
![]() |
![]() |
![]() |
#7 | |
Chief
![]() Join Date: Oct 2010
Location: Highlands UK
Posts: 323
Downloads: 36
Uploads: 0
|
![]() Quote:
The game is completely smooth, but once in a while i get the stuttering effect (when using external cam), this is almost certainly from running out of V-ram and my PC switching to system ram. I measure my V-ram by using the overclocking tool MSI Afterburner, which apart from measuring temps also records used V-ram including min-max amounts. With max AA in-game slider, it was measuring a peak of 1279mb (bear in mind i have 1280 on the GTX570. I personally think its switching to system memory at that point causing the stutering, and MSI Afterburner is just recording my ram used as 1279 instead of 1280.! When lowering the AA slider one less notch, to what you are suggesting is now 8xAA, im getting maybe 5FPS more frames less stutering and a V-ram peak off 1269. A difference, but not to the amount that i would expect from 16x to 8xAA....its more akin to a change from 4x to 2xAA. As for the CPU limiting factor, sure it could be the games not well optimized where the CPU is concerned, but it is known that SH5 runs much faster for ATI cards than Nvidia, with owners of the much slower than GTX570 OC'd 5870 getting FPS in excess of mine. Thus again it points to the graphics card. Also its worth noting, that if the CPU is limiting SH5 as much as you say, then it would be the only game im aware of that this is the case at these resolutions...the other game i play is the Totalwar series which is both CPU and GPU hungry in vast amounts....think of rendering individual movements on 10,000 seperate men at once, while providing physix for cannon shot explotions etc all at the same time; yet an I7 @3800 yeilds 2-3FPS extra frames above 1600x1200 than a E8500@3000!!! Seriously, there is truth in what you are saying, its not all black and white, but years of scanning the effects and reading all the benchmarks of CPU / GPU in games has shown me that at 1600x1200 + resolution, the graphics card will always yeild the most limiting fps difference and the difference between a respectible main-stream CPU and the best is very small in all but CPU benchmarks. |
|
![]() |
![]() |
![]() |
Thread Tools | |
Display Modes | |
|
|