Bad image quality for NVIDIA GPUs at default settings

[mroofie]

Another link read the comments please

http://forums.anandtech.com/showthread.php?t=2437903&page=3

Birdie needs to do some research

 

[Xzibit]

Another link read the comments please

http://forums.anandtech.com/showthread.php?t=2437903&page=3

Birdie needs to do some research

That person is mistaking in-game setting with driver UI settings I pointed out.

As far as color goes you can see the confusion in this posted screen grab

Somehow Limited range appeared to be more detailed and prefered to a person.

The color can be correct for both but DEFAULTS are in question.

More people have to do DEFAULT only test in various games to see if there is a difference in how drivers handle the IQ. For this to go anywhere.

If hes capturing video with-in AMD Fury X, CCC has a separate video color enhancements which are on by default. He'll have to turn that off or capture it from output so it wont have an effect on video comparisons.

 

[mirakul]

Found something interesting there
http://forums.anandtech.com/showthread.php?t=2437903&page=3

Originally Posted by Atomic Playboy
GTA V has an issue with the ingame Anisotropic Filtering on Nvidia cards; basically, it doesn't work right. Disable AF in the ingame menu and force 16x AF through the Nvidia control panel (make sure to choose "Override application setting"). Also set the filtering quality to "High Quality." That should fix the texture filtering issues (as shown in this comparison shot).http://international.download.nvidi...ering-interactive-comparison-1-on-vs-off.html

So this means every GTA V bench for nVidia cards are inaccurate in term of AF setting?

 

[the54thvoid]

I love my Nvidia image quality, so much I bought shares in the company and smoked my AMD shares.

But seriously, I'm all for tech discussion, and praise to @Xzibit for being very neutral with good, honest info but this thread is still toeing the line of doom.
I guarantee right now people are googling "Nvidia bad image quality" and just posting it here without any recourse to technical dissection or logical thought.
Think I'll go start a thread on why Fury X has such poor frame rate latency in many games.

Edit: it does you know. Though not as bad as poor old Kepler

 

[AsRock]

That person is mistaking in-game setting with driver UI settings I pointed out.

As far as color goes you can see the confusion in this posted screen grab

Somehow Limited range appeared to be more detailed and prefered to a person.

The color can be correct for both but DEFAULTS are in question.

More people have to do DEFAULT only test in various games to see if there is a difference in how drivers handle the IQ. For this to go anywhere.

If hes capturing video with-in AMD Fury X, CCC has a separate video color enhancements which are on by default. He'll have to turn that off or capture it from output so it wont have an effect on video comparisons.

Why you posing this crap still ?, 1st off the picture's are different AGAIN just like the ones from Sleeping Dogs were it could be just a angle or even bug in game or drivers, or in this case some one fucked with the color settings lol.

 

[Caring1]

Isn't the purpose of this Comments and feedback section for topics relative to TPU, not general hardware or software?
If the OP wanted to whine about graphics cards and settings then he could have said what he wanted in one of the many existing threads!

 

[LightningJR]

This all started at wccftech (the conspiracy) why am I not surprised
After the gtx 970 fiasco every fanboy out there is looking for problems where there aren't none (espically now after the fury x got ripped)

Thing is this was an issue way back in the day. Both parties were reducing the quality of graphics to get better 3DMark scores. It would not surprise me to see it happen again.

I haven't seen any real or reputable sources doing these tests to really believe any of this. Until then or I do the tests myself (I have no AMD card other than a 4830) I am calling it fud. But I can't say it's not possible.

 

[rtwjunkie]

Personally, my view if you really want to compare picture quality, the settings should be done IN GAME settings. Also, you MUST have the exact same scene at the same time of day in game. Only with all this can you see if there are any real differences in the cards' ability to render a quality image, and only then can you compare frame rates.

Interestingly, I believe that is how W1zzard does HIS testing.

 

[mirakul]

Can anyone check whether the default setting of image quality in nV control panel is 'quality' override or 'let the application decide'? Because it's high chance that every reviewer don't even touch there and just let it to default.

 

[blued]

Not so obvious. G-Sync Monitors and Limited Range RGB 16-235 on NVIDIA GPUs

The question was as to defaults and when your using a lesser color signal you save on bandwidth.

Limited

Full

Nvidia fixed the issue many months ago with a driver update. You get full color 0-255 HDMI now. Really, you should know better than to post a year old link.

 

[the54thvoid]

Can anyone check whether the default setting of image quality in nV control panel is 'quality' override or 'let the application decide'? Because it's high chance that every reviewer don't even touch there and just let it to default.

I just checked mine - I'm perfectly happy with the settings. They're mostly allowing application control and all non app controlled are set to high. A few features are set to off but these aren't ones I even care about. I also have the option of doing whatever the hell I want.

@W1zzard - what setting do you use. Just to shut this

Spoiler

muppet

up.

Nvidia fixed the issue many months ago with a driver update. You get full color 0-255 HDMI now. Really, you should know better than to post a year old link.

Can you source that? otherwise @Xzibit will tear you a new one - he's quite good at finding things.

 

[blued]

Some people claim that NVIDIA GPUs produce worse image quality than AMD GPUs using the same in-game settings which gives NVIDIA an edge.

One example:

Another example:

I wonder if W1zzard could write an article comparing image quality in BF4 and GTA5 (these are the games where the difference is the most striking) for top NVIDIA (980Ti/Titan X) and AMD GPUs (R9 390X/Fury X).

Thank you.

Where did you find the second pic? Also at AT forums? Quite funny because I also posted it there as an example of a disingenuous argument vs a comparison pic with my Nvidia card when I originally argued this a couple years ago. Anyway here is same scene/location with my Nvidia card:

Just incredible how far people are willing to go to conceal or omit info or facts that may debunk their argument.

 

[blued]

I just checked mine - I'm perfectly happy with the settings. They're mostly allowing application control and all non app controlled are set to high. A few features are set to off but these aren't ones I even care about. I also have the option of doing whatever the hell I want.

@W1zzard - what setting do you use. Just to shut this

Spoiler

muppet

up.



Can you source that? otherwise @Xzibit will tear you a new one - he's quite good at finding things.

Jeezuz... how about from people how actually own Nvidia cards who may be more informed than those who dont? In case you dont believe them, well, here:

https://pcmonitors.info/articles/correcting-hdmi-colour-on-nvidia-and-amd-gpus/

 

[blued]

Can you source that? otherwise @Xzibit will tear you a new one - he's quite good at finding things.

 

[blued]

Btw.. I mentioned this debate at AT forums where someone posted a set of pics comparing 660ti vs 7950.. well to his surprise, I posted my Nvidia screen shots at same scenes/locations in Sleeping Dogs. I can only imagine the embarrassment of those trying to ride on that argument.

See the differences:

His furnished pics.

My own pics.

 

[mirakul]

I just checked mine - I'm perfectly happy with the settings. They're mostly allowing application control and all non app controlled are set to high. A few features are set to off but these aren't ones I even care about. I also have the option of doing whatever the hell I want.

Did I ask about your settings? I'm asking the DEFAULT settings. You didn't bother to answer my question btw.

 

[okidna]

Can anyone check whether the default setting of image quality in nV control panel is 'quality' override or 'let the application decide'? Because it's high chance that every reviewer don't even touch there and just let it to default.

On 353.30, the differences between Default settings and "Quality" is :

Default : Anisotropic Filtering - Application-controlled
Quality : Anisotropic Filtering - 8X

Default : Antialiasing Mode - Application-controlled
Quality : Antialiasing Mode - Override any application setting

Default : Antialiasing Setting - Application-controlled
Quality : Antialiasing Setting - 4X

That's it.

 

[blued]

Nvidia fixed the issue many months ago with a driver update. You get full color 0-255 HDMI now. Really, you should know better than to post a year old link.

I should also add that this would only have been an issue for those who did not have DVI or DP as an option. Monitors with only HDMI must be rare to come by.

 

[Steevo]

Found something interesting there
http://forums.anandtech.com/showthread.php?t=2437903&page=3

So this means every GTA V bench for nVidia cards are inaccurate in term of AF setting?

My 7970 is the same, AF in game causes corruption or distortion effects, forcing it through the driver fixes the issue, as well surface format optimization causes transparent effects to be corrupted or distorted.

 

[birdie]

Another link read the comments please

http://forums.anandtech.com/showthread.php?t=2437903&page=3

Birdie needs to do some research

Yeah, really, ... or maybe in the same thread people are proving my point:

http://forums.anandtech.com/showpost.php?p=37532964&postcount=67

Still poor lod distance for Titan X, and poor AF.

Which means NVIDIA is cheating and their products might actually be slower than what reviews and reviewers are saying. Anyways I've moved to respective threads at anandtech and overclockers forums since people over there are actually discussing the issue instead of calling names or calling one another "fanboy".

 

[the54thvoid]

Which means NVIDIA is cheating and their products might actually be slower than what reviews and reviewers are saying. Anyways I've moved to respective threads at anandtech and overclockers forums since people over there are actually discussing the issue instead of calling names or calling one another "fanboy".

Or when a supposition is not fully supported, OP ups sticks and goes somewhere else where he might find back up.
Really, there is no major issue here at all. The default settings are almost all high quality or off, for some more exotic FX. But these settings are easily over ridden in NCP.
For this topic to be relevant you actually do NEED to know what the reviews used. Spouting assumptions about settings that you can't possibly know without that knowledge is just plain daft.
Ask the reviewers, don't create flame threads for your delectation. It's like me creating a thread about what boxer shorts you wear without knowing if you even wear them.
Find the proof, then we can discuss it. The null hypothesis is "Reviewers selecting default NCP options result in no fps increases above equivalent AMD CCC settings."
So, do reviewers use default? Can you control variables by matching identical AMD settings? Is there an unfair advantage from this two known factors?

Without knowing the variables used in reviews, all you get is people without any actual empirical evidence spouting unsubstantiated opinions.

End of.

 

[rtwjunkie]

So, now you are outright accusing, instead of inquiring. We all suspected when you presented two Sleeping Dogs screenshots that looked 5 miles apart in quality.

As to moving your discussion on overclockers and anandtech, enjoy their less informed opinions and total head nodding to your flame attempts.

Mods, on that note, can we please close this?

 

[mroofie]

Yeah, really, ... or maybe in the same thread people are proving my point:

http://forums.anandtech.com/showpost.php?p=37532964&postcount=67


Which means NVIDIA is cheating and their products might actually be slower than what reviews and reviewers are saying. Anyways I've moved to respective threads at anandtech and overclockers forums since people over there are actually discussing the issue instead of calling names or calling one another "fanboy".

Look at blued's comment aka his pictures

So, now you are outright accusing, instead of inquiring. We all suspected when you presented two Sleeping Dogs screenshots that looked 5 miles apart in quality.

As to moving your discussion on overclockers and anandtech, enjoy their less informed opinions and total head nodding to your flame attempts.

Mods, on that note, can we please close this?

Finally someone else with a brain too

 

[mroofie]

So, now you are outright accusing, instead of inquiring. We all suspected when you presented two Sleeping Dogs screenshots that looked 5 miles apart in quality.

As to moving your discussion on overclockers and anandtech, enjoy their less informed opinions and total head nodding to your flame attempts.

Mods, on that note, can we please close this?

I have seem some comments on both of those sites shocking!

Viva retarded comments


(Ps i needed to use retarded don't kill me )

Why you posing this crap still ?, 1st off the picture's are different AGAIN just like the ones from Sleeping Dogs were it could be just a angle or even bug in game or drivers, or in this case some one fucked with the color settings lol.[/QUOTE

Epic comment is epic

 

[erocker]

Since thread/discussion is based on misinformation, continuing this discussion seems to be a moot point.