xfinrodx ...I'll give them a go-round I think.  I haven't used NHancer too much, mostly just for driver-maxing guild wars and adding AA to Mass Effect 2.

Thank you muchly for the write-up!

You're quite welcome.
 
There are still some games that won't co-operate, in particular Batman: Arkham Asylum. Doesn't matter how you try it - enhanced or over ride - as soon as you apply anything other than MSAA it's instant slide show.
 
But some games will surprise you.
 
Everyone complains that AA doesn't work right in Crysis. As long as you use in-game, that's a fairly accurate assessment. The game's engine is optimized for Edge AA because of the amount of vegetation in most scenes. Trying to apply AA across the screen is very taxing, and you'd think moving from 2xAA Edge defaults to 16x AA would make a difference that would be worth the cost in performance, but have a look:
 
2xEdgeAA
 
 
 
and switching modes to 16xAA we see
 
 
 
Now these are .jpg files this time around for two reasons. First is that the Crysis Benchmark Tool is the only way I know of to get exact matches on the frames for comparisons. The idea is to load each into its own browser tab or to use the Fwd/Rev keys on something like Photo Viewer to swap quickly between the two. When you do that with these two shots, you'll see where the difference lies and just what's up for the almost 20% performance loss.
 
Keep in mind that Anti-Aliasing is more than just about cleaning up edges - its main intent is to sharpen the image and increase the resolution. See very much of that happening? Although I do see a slightly higher level of detail, there's still a lot of the vegetation, for instance the trees surrounding the radio tower, that are only being approximated with "swatches" of green rather than rendered groups of leaves and shrubbery.
 
Here's the screencap of the performance costs. The left two are default and 16x, obviously. The one one the far right with COU-Z above it is the one set up with nHancer to over ride the game and apply 32xQ across the screen
 
 
 
Now, just judging by the speed alone, which would you rather run. And with this game, don't even think about trying the in-game 16xQ.
 
And now the second reason I used .jpgs for this.
 
At this point, keep in mind that .jpgs are compressed files and don't restore to full "fidelity". Even with that handicap, I think you might see a bit of difference with the 32xQ that nHancer gives you...
 
 

Es de nada, amigo...

BR fro jaafaman please!

Cheers!

IC

AlexisROCould you please post a screen of the settings you have in nHancer for Crysis? :D

Also i remember you showing a screen in which the texture of the mountain would be a lot more sharper/ clear. What settings have you used then?

You have a very good memory and a keen eye.
 
I do return every few months or so to kind of "repost" this information, usually when there's a big hullaballoo over a driver but sometimes to just bring the info back to the top of the thread stack. The one you're remembering is most likely about the time the 191.07s came out and everyone with an NF200 went running around fearing the end of graphics.
 
I wish I still had those shots, but i tend to refresh those types of things every so often. I kinda "fudged" a bit on that one, cause the default and 2xaa shots were from an earlier driver, and it had an 8x and 16xAA shot as well from 191.07. But they were the 32-bit engine, which I usually run for speed trials.
 
I purposely put those in-game 32-bit shots against a 64-bit engine shot to not only show that the engine speed had improved considerably, but that 64-bit was a lot better visually as well. I didn't hide that fact - it's all in the engine's display data within the shot - but I didn't highlight it either. But even at the same degree of processing, put the x64 against the x32 shot and it's no contest.
 
In other words, I placed the best against the worst to increase the perceived value of the process even further. And it's the starkness of that contrast that caught most folks attention that time. These are all straight x64s this time around.
 
No attempt at subterfuge or anything - just in a hurry to show folks that the drivers not only worked without destroying your system, but they looked very good on top of that.
 
About the same way I used modded drivers as an intro this time. It was the prod to show how good a modded driver could be that got me moving...
 
The CBT shot just shows how I set it up for a photo capture. They're always run seperately from the speed runs when I do this.
 
edit -
 
I also remember that the 32-bit shots were the nVidia driver and the 64-bit shot from TweakForce drivers. Mods weren't the focus then - the solidity of 191.07 was. So I kinda "failed to mention" that as well at the time...
 

Well, there are surprisingly few trouble threads on this driverset, aren't there? Even over on the nVidia Forums. And I've not run into anything myself. So...
 
Do you want a fast bench or a pretty one, and do you have a preference in benches?...
 

Perfect timing! I just finished such a group (check the time on Crysis).
 
I ran the rig up to speed so folks could see how well the driver's performance scales as well. With the exception of the two FutureMarks, everything's maxed out with nHancer. Vantage is run at High presets, and 3DMark06 is at full 1080p. Cryostasis will not allow a full override on the AA settings, so it's set to Enhance the application.
 
I turned VSync off on the games so y'all could see the full range. These are the tougher ones that do respond well to using nHancer (and have a bench with usable splash screen). FutureMark's included for its familiarity...
 
 

 
edit -
 
I'm not sure what you mean here, though:
 
"I'm curious why you didn't include Crysis in the nHancer group as you did with the Nvidia and Tweakforce."
 
unless you missed the sentence that sits directly above the screencap of Crysis Benchmark Tools' 3 settings for the run...

Have you looked at Post 36 yet   /\  /\ 

I used to have a Voodoo laptop that I bought back in '03 (from the now HP gaming studio) and they used Omega Drivers for their PCs... talked to a rep who admitted this to me. They ran better than the stock drivers for sure.

I'm curious as to how these W7 drivers will compare to their stock equivalents.

kscott190 Are these good for folding on?

I can't answer that one. I personally don't Fold and the guy I know that folds by the farmload and has folks like Mountain Mods custom fab his cases only uses release drivers...