AMD & NV image quality comparison

[antihelten]

Really doubt that it's a Titan X issue.

http://www.techpowerup.com/reviews/MSI/GTX_980_Ti_Gaming/10.html

It'll be visible in those benchmarks.

Why would it be visible in those benches? Assuming TPU runs with default settings, who's to say the results wouldn't decrease when the setting is changed to quality?

 

[amenx]

What is this then:

Isn't that "Let the 3D Application Decide" on by default? Thats how I remember it when I last used Nvidia card.

I think this is what he meant:

Defaults for above are are Quality for texture filtering and AF= Application controlled. For as long as I can remember I have put them on AFx16 and high quality for texture filtering in 'global settings'. Each game has its own profile in NV CP with similar settings options as well. Again, I remember this did have an effect on 3dmark scores, HQ texture filtering more significant impact than AF. Will do some 3dmark runs later on to check how much.

 

[Tapoer]

For clarification, all I did was alter the colour slightly on the Fury X to suit my preference and when switching over to the TX, I didn't touch anything (it was only for a couple of hours in comparison tests).

Points to note are:

• I used DDU before installing the Fury X
• I used DDU before installing the TX
• I use a capture card (avermedia LGP lite) for both runs
• Both these runs were the exact same quality settings
• nothing is touched in the CCC or NVCP
• All settings are exactly the same for both runs

That is how he did it.
Maybe DDU messed up something, or did not make a 100% driver removal.
Or a clean installation of a recent Nvidia driver can have a bug that need a previous drivers installed to be fixed?

 

[sontin]

Or - he has just no clue what he is doing.

I found out that i need to restart Battlefield 4 for any changes in the "texture filtering" setting.

 

[amenx]

That is how he did it.
Maybe DDU messed up something, or did not make a 100% driver removal.
Or a clean installation of a recent Nvidia driver can have a bug that need a previous drivers installed to be fixed?

I believe GTA 5 has an issue with Nvidias AF. Some have been able to 'fix' it by turning it off in game and forcing it to AFx16 in NV CP. But that trick seems to no longer work after last patch for the game. Its a game + Nvidia issue rather than an overall Nvidia vs AMD IQ issue.

https://support.rockstargames.com/hc/communities/public/questions/203289108-Blurry-textures-on-PC

https://www.reddit.com/r/GrandTheftAutoV_PC/comments/33hjte/blurry_textures_just_a_few_meters_away/

 

[Tapoer]

Or - he has just no clue what he is doing.

I found out that i need to restart Battlefield 4 for any changes in the "texture filtering" setting.

No, he never changed anything on BF4, he put it on ultra and never change it.

He is not the only one that this happens, some others that used AMD and now Nvidia are complained about the loss of IQ at least in some games.

This might only happen when there was an AMD card in the system, this needs to be further investigated.

EDIT: How many reviews are done just by swapping a card with a driver cleaner?
EDIT2: Probably, most of them, otherwise the power consumption figures could not be compared, specially with overclocked CPU.

 

[3DVagabond]

Or it was a full moon? Wow! I can't believe the denial in this thread!

 

[jj109]

Or it was a full moon? Wow! I can't believe the denial in this thread!

So I take that you concede your tinfoil BS? Cool. Good to know that you're 0/3.

 

[amenx]

Or it was a full moon? Wow! I can't believe the denial in this thread!

'Denial' works both ways.

 

[sontin]

No, he never changed anything on BF4, he put it on ultra and never change it.

He is not the only one that this happens, some others that used AMD and now Nvidia are complained about the loss of IQ at least in some games.

This might only happen when there was an AMD card in the system, this needs to be further investigated.

EDIT: How many reviews are done just by swapping a card with a driver cleaner?
EDIT2: Probably, most of them, otherwise the power consumption figures could not be compared, specially with overclocked CPU.

And he has never shown his control panel and the BF4 profile.
He has the problem and nobody else.

 

[Tapoer]

And he has never shown his control panel and the BF4 profile.

He said many times that it was at default, he did not change anything drivers wise, be it control panel or profile.

He has the problem and nobody else.

You cannot prove that, once again there are people switching from AMD to Nvidia that noticed a lower IQ on some games.

But sure, some/most people with Nvidia have not that problem in BF4, that doesn't make "only him have a problem".

 

[sontin]

Sure i can say it: He is the only with this problem in Battlefield 4. Show me someone else who has this problem.

And yes, he is saying all this stuff. Doesnt change the fact that he hasnt provided a picture from the control panel and the profile.

 

[ShintaiDK]

Incidentally, I can only reproduce Gregster's "defaults" by forcing AF to 0x in the control panel.

Sigh.

Also explains why his Titan X is 30-50% faster than Fury X.

At best he is an ignorant. At worst he doctored the results. We know he already played with different colour settings.

 

[Tapoer]

Sure i can say it: He is the only with this problem in Battlefield 4. Show me someone else who has this problem.

You are the one with the argument, you have to be the one with the proves.
Your logic is very flawed...

We know he already played with different colour settings.

Only in AMD's CCC, he changed nothing else.

 

[sontin]

I think you dont understand how this works... In this thread there are three different people who cant reproduce the problem.

On the other hand you havent proved the opposite of our findings. It is your turn now.

 

[Tapoer]

I think you dont understand how this works... In this thread there are three different people who cant reproduce the problem.

On the other hand you havent proved the opposite of our findings. It is your turn now.

I do understand logic better than you, that I know.

I already said that many if not most don't seem to have this problem in BF4, but there is no valid ways to disprove that what happened with Gregster doesn't happened with anybody else swapping cards as he did.

They are not completely reproducing the problem, how many of those people without BF4 problems had a AMD card on their system, used DDU and then used an Nvidia card and installed the latest drivers?

I didn't make any argument that needed proves, I'm just speculating on the evidences. I only said that you cannot prove your argument and you are proving me to be true.

 

[Flapdrol1337]

I've been playing around with battlefield 4 and the nvidia control panel.

The only way you can get results as bad as those screens is if you overrride anisotropic filtering, looking at the images I think it's completely off.

If you go to the nvidia control panel, adust image settings with preview, then do the slider all the way to the left (performance) it actually turns anisotropic filtering off, so that's probably how it happened, or it's a bug in the game/new driver, I'm running the 353.06

Driver defaults, even setting the texture filtering optimisations to highest performance actually produce good image quality, you have to override anisotropic filtering for the bad image.

There's no way nvidia would turn off anisotropic filtering for a modest performance gain, either it's user error or a bug.

 

[sontin]

I do understand logic better than you, that I know.

I already said that many if not most don't seem to have this problem in BF4, but there is no valid ways to disprove that what happened with Gregster doesn't happened with anybody else swapping cards as he did.

They are not completely reproducing the problem, how many of those people without BF4 problems had a AMD card on their system, used DDU and then used an Nvidia card and installed the latest drivers?

I didn't make any argument that needed proves, I'm just speculating on the evidences. I only said that you cannot prove your argument and you are proving me to be true.

So, you cant show another person with this problem?

Like Flapdrol1337 i can reproduce this behaviour when i override the setting in the BF4 profile.

So either you find another source for your claims or you get a picture from the BF4 profile of the driver. Until then there doesnt exist a problem. And yes it is just so simple.

 

[amenx]

Just did a few 3dmark 11 runs:

http://www.3dmark.com/3dm11/10012862 11466 graphics score: 14376 - default settings

http://www.3dmark.com/3dm11/10012905 11195 graphics score: 13802 - AFx16

http://www.3dmark.com/3dm11/10012998 11009 graphics score: 13485 - tex filtering = HQ

http://www.3dmark.com/3dm11/10012969 10995 graphics score: 13485 - AFx16 + tex filtering = HQ

last repeated:

http://www.3dmark.com/3dm11/10013031 11011 graphics score: 13431 - AFx16 + tex filtering = HQ

Notable points:

1- AFx16 and HQ tex filtering do not appear to have combined impact effect on score.

2- These tests seem a bit different to last time I run them (couple years ago?) where impact of HQ tex filtering seemed greater on score (but different card then).

3- Best to check in game FPS performance to see if roughly in line. Otherwise the 3dm results probably meaningless.

First thing I do after driver installs always switch to AFx16 and HQ tex filtering. All my games look very good (to me) except ones with poorly implemented bloom. I do not have GTA 5 or BF4 though so no 1st hand experience there. Although I do think with some tinkering around I can go a long way to address any of the IQ issues reported in these games.

IQ to me is of paramount importance... FAR FAR more than raw FPS. If I find a sliver of evidence that AMD has better IQ that I cannot match with Nvidia.. bye bye Nvidia.

 

[amenx]

The difference on Titan X, 1080p, in BF4 is ~10%.

https://www.youtube.com/watch?v=a2IIM9fncqc

Might test in other games, dont have BF4. Wish some reputable site can do more extensive testing on variety of games.

 

[Deders]

That is how he did it.
Maybe DDU messed up something, or did not make a 100% driver removal.
Or a clean installation of a recent Nvidia driver can have a bug that need a previous drivers installed to be fixed?

This is a good point. some games will detect what kind of card you are using when you install the game. A friend of mine had Crysis on his PC, switched from an AMD card to Nvidia and the Z-buffering was so messed up.

 

[Tapoer]

So, you cant show another person with this problem?

Like Flapdrol1337 i can reproduce this behaviour when i override the setting in the BF4 profile.

And you can't show me another person swapping cards the same way as he did that have no problems.

I already said that many don't have this problem, you are not reading what I wrote.

So either you find another source for your claims or you get a picture from the BF4 profile of the driver. Until then there doesnt exist a problem. And yes it is just so simple.

As he said, it is default, if after using DDU the default settings doesn't give the best results than it might be up to Nvidia to fix this bug, if you think he changed something and don't believe him that's up to you, but at this point you might very well believe on anything that you like.

I can also close my ears and my eyes and say that there is no problem on this world, or you are just a bot because I saw no pictures of you. And yes it is just so simple.

 

[ShintaiDK]

Tapoer, seems like you want to be a problem despite facts showing otherwise.

There isnt any kind of saving grace for the Fury lurking.

 

[Tapoer]

Tapoer, seems like you want to be a problem despite facts showing otherwise.

There isnt any kind of saving grace for the Fury lurking.

I want the true, and the true is that Grester have a problem in BF4, and as him some other swapping AMD for Nvidia are also reporting problems.

I'm no fanboy or cheerleader, I don't care about Fury, is way out of my income, I care about PC gaming, I care about people doing wise GPU choices based on correct reviews.

I care about finding out exactly how big is this problem and what causes it, once again I care about he true.

What do you care?

 

[maddie]

Tapoer, seems like you want to be a problem despite facts showing otherwise.

There isnt any kind of saving grace for the Fury lurking.

Shintai. You really should learn and understand what the word 'fact' means. The problem exists. As to the reason, that is where proper analysis is important.