Another Site reviwing with 3DMARK and 53.03

Brent said:
Post the link here if you find it, and i'll give it another shot.

BTW using Cat 4.2.
Found it! I think it must be the drivers than, I reverted to 3.8 and AA worked fine forced thru control panel. I haven't tried it since putting the 4.2s on, although I remember hearing the 3.9s AA worked with too.

Here about 2/3's of the way down in one of my posts which will basically end up pointing you to this one here where kyleb pointed out that AA worked with the 3.8 set. (The post above mine, I suck at direct linking posts late at night and it is WAY past my bedtime!)

I'll try it in the morning with the 4.2s and see how it goes. Hope that's enough until after I get a bit of a nap in. :)


Quick note on AF from here
digitalwanderer said:
Just found this and wanted to stick it somewhere I wouldn't lose it 'til I had time to play with it. (Me and the kids are home in a snowstorm while the wife is at work, I'm hopping! :oops: )

In system.cfg of root game dir find the line at the top of fil that says D3D texture level= TRILINEAR, change trilinear to ANISTROPIC. Now go about 3/4 way down the page untill you find texture_anistropic_level=1, change the 1 to 4, 8, or 16. Quit and save. When you restart the game the texture detail level in the game menu should now read custom, instead of medium. To verify that I was indeed running in Anistropic and had an increase in image quality I did a before and after screenshot. The results were NIGHT and DAY. The textures are waaaaay more sharp and my fps show the change too, as they dropped about 10 fps. This also seemed to fix the stuttering/ chop I was getting when engagine enenmies too. I think it is due to the ATI cards being able to run better with the anos then with trilinear filtering in the game even though Anos is more intensive. Must be a software to driver thing, and ATI is known for it's powerful anos algorithms.
So with the Cat. 3.39 I am able to fix the harsh rock shadow problem, have 6x anos, and with the system.cfg tweak, have 16x Anistropic and have the detail texture in the menu read custom as result. I was also able to change the render mode that's in the game menu by editing the system.cfg. It is basically a shader effect. There is 4 settings, normal (default), paradise, cold, and cartoon.

From http://forums.ubi.com/messages/message_view-topic.asp?name=us_farcry_general&id=zyzoz
 
Cant wait to read/see your examples Brent. I still get comparible IQ on both cards with 53.03 (forcing 2.0 or not). Have you experienced the same issues with ingame settings for lighting changing between benchmark runs?
 
Pete said:
What's stopping you? Please, post. :)

Ok...

***Disclaimer***
This is not a direct comparison between the two cards, but just my experiences. I am not a reviewest for any site, nor do I pretend to be.
***Disclaimer***

I used 1024x768 res because I believe thats what most people will use. And to be honest, at 1280x1024 the 5900 was just too slow, often times in the low teens. I took the screen shots with 4.1 drivers for my 9800XT. I uninstalled them, and ran the cat uninstall file. I then shut down the PC, and changed the cards. Installed the 53.03 drivers, and set AA/AF to app preference, and all other quality settings to high. The same was done for the 9800XT. Both drivers were set to the same thing, and all game settings were left the same.

This is just to try and judge image quality, and give a rough estimate on how the cards do in the game. I realize its not a real 5950, and I think the extra ram will help it over the 5900NU I used, although probably not by much. Usually you dont see much of a difference from 128meg vs. 256meg untill you reach high res's, like 1280x1024 and especially 1600x1200 with lots of AA/AF.

There was no 3rd party tweaking software of any type. Just the drivers set to highest quality. I tried to enable PS 2.0 for the FX with Rivatuner, but I couldnt find the settings, so it was not used.


System I used was:
3.0c@3.9
IC7 MAX3
1gig HyperX DDR 520@1:1
9800XT @stock clock
5900NU @475/475
Audigy 2 ZS
Raptor 36gig
PC Power&Cooling 510 ATX Deluxe

This is on Farcry beta version 2.

Game settings:
Settings.jpg

Settings2.jpg


5900NU:
1.jpg


9800XT:
8.jpg


5900NU:
3.jpg


9800XT:
10.jpg


5900NU:
2.jpg


9800XT:
9.jpg
 
5900NU:
12.jpg


9800XT:
2.jpg


5900NU:
13.jpg


9800XT:
3.jpg


Thats all the pics I have. I cant see much of a difference at all, except the crain has much better AA on the cables for the XT. The only huge difference is the FPS between the two. But again, this is not a comparison as its not a true 5950 Ultra, so its not really fair. I am not sure how much a difference the extra 128megs of ram would have for the 5900, but probably not that much. Not sure how much of an impact PS 2.0 would have on frames either. But it can be argued that since its hard to tell which show is which thats its not needed.

I cant explain the Monkey Bay shots. The 9800XT was 2fps lower in both shots. The FPS were jumping around a little hovering in that area. One thing I did notice was that Monkey Bay by FAR was the hardest on my system. I should have taken some more shots on that map it seems, different areas and such, to see if the trend continued.

Brent, I didnt know that about the screenies, nice info. These were taken the day before the 4.2's came out, which was a few days before the 56.55's came out. Thats why they were not used. I dont know if it would change or not.

edit, AA works fine in the second beta. Just select app pref and then select it in game, works fine.
 
Brent said:
i've taken 2 water shots, 2 land shots, and 5 shader shots to put in my review for comparison, ABIT 9800XT (cat 4.2) vs. 5950U (forceware 56.56)

Does this mean we'll be seeing a review with unreleased nVidia drivers again, or will this set have an official WHQL release by the time the review is published?
 
Brent,

thanks for taking the time to check! I think Farcry can become like SS was, meaning it has a boat load of options that can driver a reviewer nuts as they have soo many of them to check :) Yet they are cool as you can do a lot with them.


FallGuy,

thanks for the work. Do you think it could be a diff in the Multiplayer vrs Single player? Again the screen shots of the single player is night and day. Also this is just a beta/demo so whos to say all of this wont change in the final game.
 
Veridian3 said:
Hanners said:
will this set have an official WHQL release by the time the review is published?

Now that would be telling.... :LOL:
I would be inclined to believe so, although the timing seems odd. Warp2search is claiming that they should be out today, Gainward had them up on their FTP, and you don't think that leak was intentional?

Perhaps this is also why Terry Makedon seems annoyed that he has to wait to release the ATI drivers with similar features until the R420 launch. I know that ATI has a department that keeps tabs on what other companies in the industry are doing (and if I remember correctly, it's called the department of competitive engineering--HOW BIZARRE IS THAT), so I imagine that ATI knows exactly what's going on at NVIDIA (and, more than likely, the reverse is true as well, although NVIDIA is pretty famous for being porous when it comes to information).

Guess we'll see.
 
Hanners said:
Brent said:
i've taken 2 water shots, 2 land shots, and 5 shader shots to put in my review for comparison, ABIT 9800XT (cat 4.2) vs. 5950U (forceware 56.56)

Does this mean we'll be seeing a review with unreleased nVidia drivers again, or will this set have an official WHQL release by the time the review is published?

we don't do reviews with un-released or non whql drivers anymore

56.56 is WHQL'd and their add-in board partners will be shipping their cards with the 56.56 driver
 
jb said:
Brent,

thanks for taking the time to check! I think Farcry can become like SS was, meaning it has a boat load of options that can driver a reviewer nuts as they have soo many of them to check :) Yet they are cool as you can do a lot with them.


FallGuy,

thanks for the work. Do you think it could be a diff in the Multiplayer vrs Single player? Again the screen shots of the single player is night and day. Also this is just a beta/demo so whos to say all of this wont change in the final game.

Yeah, FarCry is awesome, it has a LOT of options, i mean a LOT

check this out: http://s88289070.onlinehome.us/games/console.txt that is the complete list of commands for the demo, i mean wow
 
jb said:
FallGuy,

thanks for the work. Do you think it could be a diff in the Multiplayer vrs Single player? Again the screen shots of the single player is night and day. Also this is just a beta/demo so whos to say all of this wont change in the final game.

That is of multiplayer. I started up a server, and took the pics.

Here are some screenies in the different render modes;

Render mode: Normal
Normal_2.jpg


Render mode: Paradise
Paradise_2.jpg


Render mode: Cold
Cold_2.jpg


Render mode: Cartoon
Cartoon_2.jpg


The Cartoon one almost looks cel shaded.
 
yep, according to the command list to enable those modes the cartoon look is described as having a cel-shaded look similar to XIII
 
Brent, have you ever considered writing up a "how-to" type article for your benchmarking methods? Maybe with Tips n' Tricks, scripts, demos, stuff like that?
 
Ratchet said:
Brent, have you ever considered writing up a "how-to" type article for your benchmarking methods? Maybe with Tips n' Tricks, scripts, demos, stuff like that?

never thought about writting such an article

sounds like a neat idea, sounds fun actually

but any recorded demo's we use i can't release publically, we keep them private, actually i'm the only one that has them

but i could make some recorded demo's for public use, i have thought about that

just depends on what we got goin on, so far it has been pretty busy, one card after another, and it is fixin to get real busy with next gen card stuff
 
Back
Top