X800 & 6800 Filtering Quality: NV Wins... I think?

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
This article offers professional proof that what I've said below about filtering quality was true. Look in the second page and read the discriptions of the videos.

This thread was originally started to prove to this thread that there is a difference between NV's Quality and High Quality Image Settings. I also compared filtering quality between the X800 & 6800 afterwards. There is also a section near the end that compares High Performance to Quality for 6800 by demand.

NV'S QUALITY VS HIGH QUALITY
NV states the difference between Quality and High Quality in this 10MB PDF:

Quality is the default setting that results in optimal image quality for your applications

High Quality results in the best image quality for your applications. This setting is not necessary for average users who run game applications. It is designed for more advanced users to generate images that do not take advantage of the programming capability of the texture filtering hardware.
Graph of Differences.

There is and there isn't a difference between those settings, and any difference will be much more noticeable when you are in motion. If you don't enable Anisotropic Filtering, you will see almost no difference. There may be some subtle differences that you will most likely never catch such as in Call of Duty Multiplyaer in the level Ship. In this level you will notice it from a distance with the shadow of the railing on the deck of the ship. At times, Quality may use Bilinear instead of Trilinear where Trilinear is set in game (observed in FarCry 1.3, HL2 w/o AF, Splinter Cell 2 v1.31 w/o AF) while High Quality will use Trilinear. Bilinear instead of Trilinear may also occur using High Quality. In these cases, you can Force Mipmap Trilinear in the drivers.

If you turn on Anisotropic Filtering, the differences become more pronounced, however, you will be able to see it on only certain types of textures. Sometimes you won't be able to see it at all, sometimes you'll be able to see it a bit in motion, and sometimes it will stick out like a black eye. Take Call of Duty for instance. The textures on the wall are nice, but never as sharp as some floor textures with small patterns. With Quality enabled you will notice texture aliasing which creates a moire effect on top of those sharp textures. It'll be visible in the pictures below. This is also noticeable on the floorboards on the ship in the level Ship. If the floorboards are vertical on the screen, motion will present nothing. If the floorboards are horizontal on the screen, motion will present texture aliasing. If the floorboards are diagnal on the screen, motion will present texture aliasing and moire effects.

From what I've seen when in motion in HL2 with Anisotropic Filtering: High Quality presents a slight moire effect; Quality without optimizations present a more pronounced moire effect; Quality with optimizations produce a very decernable moire effect.

HOW TO PROPERLY CHECK DIFFERENCES
Right click on the links and Save Target(s) to a folder. Now open one of the pictures with Windows Picture and Fax Viewer and keep clicking next and it will cycle through the pictures.

Quality VS High Quality
This is the level Rocket in Call of Duty Multiplayer. In the Quality picture, there is texture aliasing near those gray barrels followed by a Bilinear transition. In the HQ picture, there is a Bilinear transition between the gray and brown barrels.

Quality VS High Quality
In the Quality picture, you see some texture aliasing on the floor near the door followed by a moire effect in the distance. The High Quality picture shows Bilinear transitions. One is between the door and the gray barrels.

For those who complained about that 2nd set of pictures not be taken in the exact same position, these are corrected pictures:
Quality VS High Quality
These pictures, though, are using 71.84 BETA drivers. These exact same shot pictures are difficult to take, being that I have to shut down the game, change the driver settings and then get back into the game. I will try to keep them the same, but if I can't, I can guarrantee that any artifacts seen or not seen would be there independent of position. I cannot guarrantee the FPS in the upper left corner however.

These pictures were taken using the following system:
AMD Athlon XP 2700+
2x512MB Corsair Value Select PC2700
A7N8X-X
eVGA 6600GT - Forceware 66.93; 71.84 BETA used in second set and from this point forward.

Drivers left at default except Image Settings was changed when needed(Quality to High Quality), Vsync was disabled and Anisotropic Filtering was set to 8x.


FOR OPTIMAL IMAGE QUALITY SETTINGS
If you don't plan on using Anisotropic Filtering, leave driver settings at default, except Force Mipmaps to Trilinear. If you plan on using Anisotropic Filtering, also change Image Settings to High Quality and Negative LOD Bias(71.84 and up) to Clamp.


MORE PICTURES
Quality VS High Quality
This picture is from the Lockdown level in HL2 Deathmatch. In-game settings are at Highest and Anisotropic Filtering is enabled in driver only. Quality shows moire effects within the shadows.

Quality VS High Quality
This picture is from the beginning of FarCry. In-game settings are at High and Anisotropic Filtering enabled in driver only. Quality show a Bilinear transition.

Nv optimization
This picture was taken with the new 71.89 drivers. It uses 8xS Anti-aliasing as well as 16x Anisotropic Filtering. Everything is explained in the picture.


ATI VS NV
(The following paragraph is pre-X800 Speculation) Coming from a 9700 Pro, I noticed this loss in image quality in the 6600GT under the image setting Quality and with Anisotropic Filtering. I found that High Quality was much more comparable to my 9700 Pro and was angered that I based my purchase on benchmarks that didn't seem properly documented. The X800 didn't have a loss in image quality from what I've seen and read, how could people miss something this significant. I probably would have never bought the card if I new that it gained those framerates by lowering it's image quality. Websites use default driver settings when they benchmark. Being that Quality is default for NV, benchmark numbers reflect a loss in image quality with Anisotropic Filtering enabled and a boost in FPS up to 34%, which is unfair to the competition that I thought had image quality more comparable to High Quality. Here in this set of pics, 9800 VS the 6600, it will show that ATI had better image quality. Read on after the pics.

NV VS ATI
This is the level Rocket in Call of Duty Multiplayer. ATI produces a cleaner picture here akin to NV's High Quality found here.

NV VS ATI VS
This is the level Ship in Call of Duty Multiplayer. Here it is hard to notice the difference. I should have taken the picture with the floorboards diagnally on the screen. The difference that you notice in the NV screen is texture aliasing, which is unnoticeable when motionless.

These ATI pictures were taken using the following HP system:
AMD Athlon 64 3200+
512 PC3200
9800PRO 256MB - Catalyst 5.1 or 5.2 with CCC. Drivers left at default except Vsync was disabled and Anisotropic Filtering was set to 8x.

These pics clearly show that the 9800 has better default image quality than NV resulting in unfair benchmarks between them at least. I had few ATI card sources at the time so I couldn't do too many pics. Even now that I have an X800, I have very few games left. Now to show the true event - to show that there is a loss in image quality between 2004 gen ATI & NV. Or so I thought. See, the 9800 doesn't have a feature found in the X800 called Adaptive Trilinear, which is supposed to adaptively select the best filtering to give the best performance while retaining optimal image quality. I took that into consideration and assumed that the 9800 would show the same image quality as the X800 considering it said that it would only do it if it didn't lower image quality. But, I fell the to the clutches of corporate bullsht.

X800 default VS NV default VS NV High
Sorry about that NV High pic not being in the same spot, but everything said here remains true. As you can see, it seems that the X800 and it's Adaptive Trilinear are at work at not preserving the excellent image quality that it's previous generation displayed so proudly. We keep digging into the crap pool with image quality it seems. Anyway, ATIs default image quality is better than NV's default by a decent amount. ATI's default has all the aliasing of NV's, but very little of the moire effect which is good. And, again NV's High Quality displays excellent image quality - much better than ATIs default quality as it gets rid of most of the aliasing. But wait, what if we turn off optimizations on ATI cards.

X800 default VS X800 AI OFF
As you can see here, disabling ATI's AI is a stupid move. It does increase the image quality by making the moire effect a bit less pronounced, but not significant enough that it would matter, except in FPS. The texture aliasing is still very much existant. I had to look at the pics to find the difference, so it seems as if ATI hasn't disabled all of their optimizations.

So, if you want the best image quality possible, NV will provide that for you, no doubt. However, when you look at benchmarks you should keep in mind that ATI does have better image quality being that it's comparing default ATI to default NV. You judge by the pics to see what matters to you.

The X800 pictures were taken using the following system:
AMD Athlon XP 2700+
2x512MB Corsair Value Select PC2700
A7N8X-X
BBA X800 Pro - Cat 5.6 - everything default.


NV: HIGH PERFORMANCE VS QUALITY
By reader request: This comparison was done with the beta drivers supplied by the BF2 Demo, 77.30.

High Performance VS Quality
This is the Lockdown level in HL2 Deathmatch. As you can see, with High Performance, the texture aliasing and moire effect is twice as pronounced compared to Quality. If you look at the tiles at the back wall, it looks like extra detail, but it's really just a lot of aliasing.

And I thought Quality was bad. On all other textures, where the loss in image quality is unnoticeable with the Quality setting, High Performance shows loss in the form of texture shimmers. You can't notice them with pics, but you can in motion.


THANKS
Jeff7181 for explaining that there is a difference between Quality without optimizations and High Quality.
 

yhelothar

Lifer
Dec 11, 2002
18,408
39
91
So you get 20% performance increase, equivalent to a one step higher resolution in performance hit. Sounds good to me.
Fanboys just love to bicker... :roll:
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
Well, see this is why I have the problem with benching at default quality in Nvidia cards.

 

kylebisme

Diamond Member
Mar 25, 2000
9,396
0
0
Halo really shows how crap some of the anistropy optimzations can be, especialy on rolling terain like seen on Blood Gultch. This goes for both Nvidia's "quality" and ATI's "Catalist AI", though I have to say Nvidia's is worse.
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
This is sort of misleading since the shots are not taken in the exact same spots, especially in the 2nd set of shots. That WILL effect where those transitions appear on the floor in relation to objects in view.
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
Well, I can't fix the picture now - I would have to reinstall older drivers, but I assure you that it is the same if they were taken at the exact same place.
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
Also, you can turn off the Anisotropic and Trilinear optimizations in the control panel. I do it, and I use Quality mode. I cannot see ANY difference between Quality and High Quality with those optimizations turned off, but there is a difference in frame rates.

So I think your not REALLY talking about Quality vs. High Quality image settings. What you're noticing is the effects of the optimizations.
 

kylebisme

Diamond Member
Mar 25, 2000
9,396
0
0
You won't see any difference in benchmarks either between "high quality" and qaulity with the optimzations off, the high quality setting is simply a way to dissable the optional optimzations all at once.
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
Originally posted by: TheSnowman
You won't see any difference in benchmarks either between "high quality" and qaulity with the optimzations off, the high quality setting is simply a way to dissable the optional optimzations all at once.

Sure I will... I just ran the Stress Test in CS:S on Quality with all optimizations off and got 148 FPS, then ran it on High Quality with all optimizations off and got 130 FPS.
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
Originally posted by: ifesfor
DO ati have this option too? I need to tell my friend who got an ati card.

They have different quality settings, but I don't think they allow you to turn off anisotropic optimizations.
 
Jun 14, 2003
10,442
0
0
i use quality mode for everything, used to use high, but realised that the guy was right, you just cant tell the difference between HQ and Q (for the games i play), well i cant, and for that reason why not thave the extra performance if the change in output between the two is negligable
 
Jun 14, 2003
10,442
0
0
oh and i take your screen shots with a grain of salt, since you are in different position in all of them, if u wanna do a comparison you need to be in the exact same posiition to make it count
 

BenSkywalker

Diamond Member
Oct 9, 1999
9,140
67
91
If anyone can oppose that ATI has better default driver image quality with 8x Anisotropic Filtering enabled in an ATI pic, that would be great since I don't have an ATI card. But I'm looking into buying a cheap 9600 to prove this, but I don't know how long that will take.

You will certainly see more contrast with an ATi part- their texture filtering accuracy is so low it gives the appearance of increased detail although it is aliasing artifacts more then anything. I've been running my R9800Pro for over a year now and ATi's texture filtering is extremely poor at best. Not saying that the current nV parts are a lot better, they have been on a downward trend since the last part that actually did real anisotropic filtering(the NV2x core parts).
 

kylebisme

Diamond Member
Mar 25, 2000
9,396
0
0
Can you provide some in game screenshots or videos to show the difference, Ben? I have both the latest gen from ATI and Nvidia and I don't see what you are talking about.


Originally posted by: Jeff7181
Originally posted by: TheSnowman
You won't see any difference in benchmarks either between "high quality" and qaulity with the optimzations off, the high quality setting is simply a way to dissable the optional optimzations all at once.

Sure I will... I just ran the Stress Test in CS:S on Quality with all optimizations off and got 148 FPS, then ran it on High Quality with all optimizations off and got 130 FPS.

Humm, I just ran the same tests and came out under 1fps apart, but granted I don't have the best CPU paired with my 6800gt. Could someone else try this?
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
oh and i take your screen shots with a grain of salt, since you are in different position in all of them, if u wanna do a comparison you need to be in the exact same posiition to make it count
The first set are in the exact same position. Although the second set aren't I assure you that the same thing is still there. You're just in denial. Anyway, you can go test it for yourself. Got COD. Go to the rocket area with Quality enabled and Anisotropic Filtering at 8x and go see for yourself. And then change it to High Quality and notice the clean beautiful surface that you paid much money for. If you read the above carefully it only appears with defined pattern so, it'd be really hard to find it some games that don't have textures like those, but many games also do have textures like those, maybe not everywhere, but they do.

You will certainly see more contrast with an ATi part- their texture filtering accuracy is so low it gives the appearance of increased detail although it is aliasing artifacts more then anything. I've been running my R9800Pro for over a year now and ATi's texture filtering is extremely poor at best. Not saying that the current nV parts are a lot better, they have been on a downward trend since the last part that actually did real anisotropic filtering(the NV2x core parts).
I can attest to that. But I still believe that ATI has better default IQ than. But if you set Nvidia to High Quality, it will have less, unnoticeable, textures shimmering and better overall image quality, if Trilinear is working.

hey, where's "refresh rate overide" option in nvidia's newest driver (71.84)?
It's missing.
 

THUGSROOK

Elite Member
Feb 3, 2001
11,847
0
0
for me its quite the opposite....

i get a higher framerate using HQ off off off then i do with Q off off off.
Q on on on is the fastest tho.
maybe its just a CSS thing?

 

kylebisme

Diamond Member
Mar 25, 2000
9,396
0
0
Originally posted by: THUGSROOK
for me its quite the opposite....

i get a higher framerate using HQ off off off then i do with Q off off off.
Q on on on is the fastest tho.
maybe its just a CSS thing?


Interesting, how much difference was there? Was it withing a reasonable margin of error?
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |