8800gts 512 oblivion problems

dostanio

Distinguished
Oct 29, 2007
236
0
18,680
Hello guys,

I am playing oblivion all maxed at 1680x1050, now am gettin extremely low FPS at times, as low as 11! Mostly outdoors, and this goes down to 8! When am by an oblivion gate! Am very puzzled by this because the rest of my system is fine!

e6750!
2gb ram
32 bit vista

All help would be appreciated!

THanks
 

njalterio

Distinguished
Jan 14, 2008
780
0
18,990
This actually seems about right for that video card. Check out toms video card charts to verify. Also they are using the higher end 8800 GTS then you are as well. I would knock down the anti aliasing and that resolution to 1280 by 1024 and you would get much better performance.

http://www.tomshardware.com/charts/desktop-vga-charts/oblivion-the-elder-scrolls-4-outdoor,570.html?p=1591%2C1598%2C1601%2C1589%2C1592%2C1615%2C1616%2C1572%2C1590%2C1636%2C1568%2C1632%2C1639%2C1587%2C1627%2C1646%2C1611%2C1626%2C1623%2C1621%2C1603%2C1618%2C1617
 

njalterio

Distinguished
Jan 14, 2008
780
0
18,990
Yes it was rated for 23 FPS, however that number that is available on Tom's charts is an average. The OP mentioned that his gets AS LOW as 11/8....so the average would reasonably be around 20.

And like I pointed out initially, Toms hardware is testing it with the better version of the 8800 GTS.
 

brendano257

Distinguished
Apr 18, 2008
899
0
18,990
Make sure in the Nvidia control panel you are not assigning high AA+AF to Oblivion, you can set it game by game if needed. I've gotten better results with my 8800 GT playing at 1600X1200.
 

itheral

Distinguished
Oct 15, 2006
129
0
18,680


Leave it to the person who feels its necessary to buy an 1100w PSU to get his info wrong.. :sarcastic:
 

pauldh

Illustrious

Yeah, you can't go by one review though. Firingsquad tests a pretty demanding foliage section of the game and shows with fsaa an overclocked 8800GTS G92 can't touch a stock 8800GTX, never mind an ultra. IMO, while the GTS is an excellent Oblivion card, the 8800Ultra is without doubt the better Oblivion card. Of course, SLI kicks butt in the GPU demanding foliage. Like FS, I see near 100% SLI scaling in parts of Oblivion.
http://www.firingsquad.com/hardware/nvidia_geforce_8800_gts_512mb/page6.asp
http://www.firingsquad.com/hardware/nvidia_geforce_9800_gx2_performance_preview/page8.asp

 

pauldh

Illustrious
Yeah, stranger, 19x12 with an X1900 does seem like you would need to tweak the settings. I know my X1950XT and 8800GTS 320MB could not come close to maxing 1650x1050 with 4xaa/16xaf. Most of the game sure, but not off pathways in the woods/foliage. All it took was a few tweaks of shadows, shadow filtering, and min grass height to get it playable with 2xaa/16xaf.

I'm surprised the OP sees 11 fps with the G92 GTS even if he has aa/af cranked and is running a high res texture mod. Something seems wrong there. But the game does stress both the GPU and CPU quite a bit, and a massive battle near an Oblivion gate is one of those low fps areas.
 

evilshuriken

Distinguished
Jun 14, 2006
399
0
18,780

I don't think it would be a good idea to run a wide-screen monitor at that resolution...
 
Remember that Oblivion was 2006s Crysis, and its still a very demanding game. If youre using a SLI/CF setup, youre ok with everything maxxed, or maybe an Ultra at no higher than 16x10, but going higher res, or using a nice but not killer card, youre stretching it at 16x10 in some parts of the game. The Ultra with its 768mb ram and wider bus helps
 

randomizer

Champion
Moderator
My 9600GT can run 8xAA 16xAF @1680x1050 if I turn external shadows off. Besides, you can't see the shadows much anyway. I still prefer no AA for better framerates though. The oblivion gates are pretty taxing that's for sure.
 

dostanio

Distinguished
Oct 29, 2007
236
0
18,680
ooh sorry for not replying back guys!


okee, my psu is an enermax liberty 620.

Oke guys this is getting very frustrating!, I am playing 1680*1050,

and this time i have maxed EVERYTHING, forced AA through nvidea software, AF! plus HDR. am getting about 10FPS, outdoor.

Something is not right?

One of the main reasons i went for a 8800GTS 512, was to max out oblivion!

If i had known so i would of gone, for the 3870X2 or 9800GX2
 

tsd16

Distinguished
Jan 11, 2008
333
0
18,790
when you guys use that high a resolution do you seriously see a difference when you enable AA?

Outdoor is still taxing on modern systems. I suppose with your settings that may be about right, but i play 1024x1280 no AA and everything runs fine. try cutting out the AA, I mean seriously is it really necessary or do you just like to know its on?
 


Seriously like everone is saying you have to turn down some settings. I have an old AGP rig with a X1650XT and a 3000+ and i can get 55 FPS at 1024X1280. thats max obviously but it dosent go below 35 ever.
Its just a matter of tweaking the settings, that link jaydeejohn gave you is very good. turning down the shadows and reflections wont make that much of a differance visually but will make a good deal of differance to the performance.
Mactronix
 

pauldh

Illustrious

AT 16x10 yes I see a difference. Admittedly, I am a bit of an FSAA junkie who fixates on the jaggies and allows that to distract from gameplay. But it's a mental thing too as I notice jaggies for sure, but it really only distracts my attention on my main gaming machine. If I am on a budget rig and expect jaggies, they jump out at me the same, but it doesn't bother me as much.
 

pauldh

Illustrious


Sounds to me like you bought an 8600GTS. :)

Just kidding. What level of AA are you forcing? To be honest, true max details 4xaa/16xaf will be a stretch for your card. it should be playable, but you will expereice some low fps. I was an oblivion nut and have benched it to death. In some PM discussions of results with an 8800GTX owner I was told even the 8800GTX can struggle at 16x10 truedly maxed details and eye candy. This can be made worse with mods.

Can you capture a screenie with fps posted as an example? Also maybe a screenie of the map location so others can try the same game area. I'll try a single and dual 8800GT in that same area.

I think you should try 16x10 max details with no aa/af forced. See what framerates you are getting at both those settings. Enable 16xAF and test again. I can tell you with a stock clocked 8800GS, at 1600x1200 4xaa/16xaf, I can reduce the fps in waste high foliage to about 18 fps average, and not during combat even. Also, with an X2 5600+ at stock, I can reduce it to 12 fps minimums in cpu limited areas of the game, again not even during combat. Your system is better than that by far, but you may still be asking too much from it to not find some areas of slowdown. If it is widespread low fps, then something is wrong for sure. Do your fps start high and then suddenly go real low like a VRAM leak? At one time that was a problem with some NV cards/drivers in Oblivion.
 

dagger

Splendid
Mar 23, 2008
5,624
0
25,780



I don't trust those benchmarks. They show significant performance gap between different brands of the same card. Identical cards running identical stock clock rate should perform identically. There's no such thing as one brand outperforming another.

Another reason is other benchmarks do not duplicate those results, they duplicate the results from anandtech benchmark. Oblivion is often used for benchmarks. If no one duplicates a benchmark, it shouldn't be trusted. It's the internet, with a lot of people claiming a lot of things.

As for the low performance problem, are you using 170s beta Nvidia drivers? If so, try rolling back and see if it runs smoothly.
 

pauldh

Illustrious
SHow me the results that duplicate Anands results. And what results vary so much by brand with identical clocks?


I've been around a while and I'll tell ya I'd trust FS over Anand for many reasons that have been discussed here through the years. Also, Firingsquad had one of the most complete Oblivion performance comparisons on the net. They know that game, have benched/explore many parts of their game, and their foliage tests do a good job of stressing the GPU. Show me a site that runs 10 Fraps run of their Oblivion benchmark to assure an accurate average: http://www.firingsquad.com/hardware/oblivion_high-end_performance/default.asp


But yes, comparing results from multiple sites to forma an opionion, and weeding out flawed testing is important.