Image quality Who is better

wand3r3r · Apr 7, 2013

Further investigation could be useful.

See if it's NV cutting corners or not. I wouldn't be surprised either way.

Lots of NV defense and skepticism. Would you be so defensive when looking at another GPU company?

BallaTheFeared · Apr 7, 2013

wand3r3r said:
Further investigation could be useful.

See if it's NV cutting corners or not. I wouldn't be surprised either way.

Lots of NV defense and skepticism. Would you be so defensive when looking at another GPU company?

Indeed it should be looked into, perhaps Nvidia is cheating with lighting, perhaps it's a LoD issue 2GB vs 3GB, perhaps it's AMD cheating with DOF, or perhaps it's nothing.

Lonbjerg · Apr 7, 2013

Will Robinson said:
With comments like that it's obvious your funny old CRT monitor has a higher IQ than you.:whiste:

So you got nothing but namecalling...not surprising.

Leadbox · Apr 7, 2013

BallaTheFeared said:
Indeed it should be looked into, perhaps Nvidia is cheating with lighting, perhaps it's a LoD issue 2GB vs 3GB, perhaps it's AMD cheating with DOF, or perhaps it's nothing.

Can you expand on this DOF cheating, not seeing it

BallaTheFeared · Apr 7, 2013

Wouldn't not seeing it be the point?

Leadbox · Apr 7, 2013

BallaTheFeared said:
Wouldn't not seeing it be the point?

Would turning lighting effects off in the name of DOF help or hinder performance?

BallaTheFeared · Apr 7, 2013

Huh?

DOF is like smearing vaseline on a camera lens. You're simply distorting image quality at the cost of performance. It's for a more cinematic experience, like TXAA.

That said, I don't see missing lighting with the Nvidia cards, what I see is lower image quality and a lot less saturation from the lights. In dimly lighted areas this seems to cause a loss of definition, though I can't say if it's actually a loss, or if it something you wouldn't see IRL. With AMD every light is extremely vibrant, even in dark areas it's like the sun is out.

But there are some low light areas that seem to have a lower LOD, either through DOF distortion, or simply because its not bright enough to actually see details.

It really depends on what the artist of the game was going for, with AMD it looks like a vibrant city with high power bulbs everywhere, while on the Nvidia system it seems more of an earthly tone, and the lights themselves aren't nearly as powerful.

Leadbox · Apr 7, 2013

BallaTheFeared said:
Huh?

DOF is like smearing vaseline on a camera lens. You're simply distorting image quality at the cost of performance. It's for a more cinematic experience, like TXAA.

That said, I don't see missing lighting with the Nvidia cards, what I see is lower image quality and a lot less saturation from the lights. In dimly lighted areas this seems to cause a loss of definition, though I can't say if it's actually a loss, or if it something you wouldn't see IRL. With AMD every light is extremely vibrant, even in dark areas it's like the sun is out.

But there are some low light areas that seem to have a lower LOD, either through DOF distortion, or simply because its not bright enough to actually see details.

It really depends on what the artist of the game was going for, with AMD it looks like a vibrant city with high power bulbs everywhere, while on the Nvidia system it seems more of an earthly tone, and the lights themselves aren't nearly as powerful.

Interesting, so using your analogy, you reckon AMD is not "smearing enough vaseline" on the camera and thats why it looks livelier?

BallaTheFeared · Apr 7, 2013

Could be, hard to say. I don't have the game or a Kepler card to test with. We do know AMD has brighter colors and more saturation, if that's enough to do what's happening I don't know, but that doesn't explain why distant objects are perfectly clear whereas with Kepler they're distorted.

Just tossing ideas out.

amenx · Apr 7, 2013

My own screenshots of same places in other pics posted, at different times of night/day. This with my 660ti:

p.s. the wavyness of the store shutter door in 2nd pick is optical illusion. Click to enlarge pic, it goes away.

Skurge · Apr 7, 2013

bystander36 said:
SLI might not be perfect, but what are you talking about?
http://www.pcper.com/reviews/Graphi...ils-Capture-based-Graphics-Performance-Test-4

Keep in mind, drivers issues do happen, but the crossfire issue is more of a general one.

Crossfire is pretty bad right now in standard form. I was talking about single GPUs. Fermi was pretty bad across the board throughout its lifetime and got worse as time went by. Kepler had issues in bf3 for a very long time. It was no smoother than SI until the 660ti tech report review. And today SI is more consistent than Kepler.

As for crossfire. AMD made a conscious decision to sacrifice smoothness for lower latency. It might have been the wrong one. They will be giving users a choice for smooth or responsive whearass you have no such choice with SLI.

So yes driver issues do happen. BF3 for Kepler was. The month or 2 where SI was stuttery in some games was a driver issue.
If Fermi was a driver issue, then its never been fixed. Yet nobody complained. Crossfire was, well I'm not sure what to make of that. Just the wrong choice?

Although that makes me wonder. When people recommend an FPS cap or vsync for crossfire, people complain about the latency it adds, but nobody questions how much latency nvidia's frame metering adds.

BallaTheFeared · Apr 7, 2013

Since when did Fermi have latency issues, outside texture trashing that occurred later when they were trying to compare cards with the same performance but 750mb more vram?

CF has always been broken, it's not an issue with just 7xxx series.

It's kind of sad when you can look at one company, who has accross the board problems and make excuses based on one game the competition didn't do great in.

"Oh yeah, but that one time, in that one game, they had a problem too!!!"

Leadbox · Apr 7, 2013

@amenx

Thx for the pics, it looks like the lighting issue is nothing more than a time of day thing as evidenced in the first 2 screenshots

wand3r3r · Apr 7, 2013

Leadbox said:
@amenx

Thx for the pics, it looks like the lighting issue is nothing more than a time of day thing as evidenced in the first 2 screenshots

I don't buy that without comparable AMD screenshots. The first two pics are clearly different times of the day. They don't prove anything without AMD shots at the same place/time.

In the NV vs AMD shots earlier in the thread, it's generally very similar in overall darkness but different areas are lighter and clear in AMD whereas they are darkened with little or no visible details with the NV card.

Skurge · Apr 7, 2013

BallaTheFeared said:
Since when did Fermi have latency issues, outside texture trashing that occurred later when they were trying to compare cards with the same performance but 750mb more vram?

CF has always been broken, it's not an issue with just 7xxx series.

It's kind of sad when you can look at one company, who has accross the board problems and make excuses based on one game the competition didn't do great in.

"Oh yeah, but that one time, in that one game, they had a problem too!!!"

cmdrdredd · Apr 7, 2013

Anyone remember the Far Cry 3 comparison where ambient occlusion was discussed between AMD and Nvidia hardware and although some claimed that Nvidia's cards would overdarken some spots, in those same screenshots the AMD cards had clear loss of texture details. Where graffiti was clearly seen on the Nvidia shot and the AMD shot had it smeared in with the wood panel it was on. Also on the brick structures on the roof, cracks were defined on the Nvidia shot and almost invisible on the AMD shot. Note the fleshtone difference in the screenshot below as well. The Nvidia shot clearly has more red in there.

This too could be attributed to the dynamic weather used in the game (time of day, overcast sky is clearely visible on one shot vs the other). We also don't know the details of the system used to acquire these shots. Was there a different setting used somewhere that didn't give a level playing field in the drivers? Too many variables to draw conclusions. The positioning is hardly exact as well. Just look at the proximity of the tree in relation to the player, and the grass. Is the FOV different or is the player actually standing closer to the structure and that is what changes the details from being clearly defined to slightly smeared?

I guess my point is that we cannot really know what is causing these differences for sure and we don't know what the original intent of the artists and programmers was for that particular scene with that particular weather and time of day. Was it intended to be like one shot or like the other? Is the color off on purpose or is it incorrectly rendered by one card?

Skurge · Apr 7, 2013

cmdrdredd said:
Anyone remember the Far Cry 3 comparison where ambient occlusion was discussed between AMD and Nvidia hardware and although some claimed that Nvidia's cards would overdarken some spots, in those same screenshots the AMD cards had clear loss of texture details. Where graffiti was clearly seen on the Nvidia shot and the AMD shot had it smeared in with the wood panel it was on. Also on the brick structures on the roof, cracks were defined on the Nvidia shot and almost invisible on the AMD shot. Note the fleshtone difference in the screenshot below as well. The Nvidia shot clearly has more red in there.

This too could be attributed to the dynamic weather used in the game (time of day, overcast sky is clearely visible on one shot vs the other). We also don't know the details of the system used to acquire these shots. Was there a different setting used somewhere that didn't give a level playing field in the drivers? Too many variables to draw conclusions. The positioning is hardly exact as well. Just look at the proximity of the tree in relation to the player, and the grass. Is the FOV different or is the player actually standing closer to the structure and that is what changes the details from being clearly defined to slightly smeared?

I guess my point is that we cannot really know what is causing these differences for sure and we don't know what the original intent of the artists and programmers was for that particular scene with that particular weather and time of day. Was it intended to be like one shot or like the other? Is the color off on purpose or is it incorrectly rendered by one card?

You have a very good point there.

BallaTheFeared · Apr 7, 2013

Skurge said:
Oh that was it, they had insufficient memory? Makes me wonder how you managed to play any games on your 470s for so long (plus the added latency of SLI)

6xxx crossfire was fine. No worse than the GTX5xx SLI.

By using proper settings of course, but also having 250mb more memory than the 1GB cards you're comparing.

The slight increase in latency was offset by the fact it wasn't a stuttering mess like the Radeon cards. It was however, a trivial increase next to the power of the vsync warriors.

We can argue until we're blue in the face, however the perception of Radeon has gotten worse this generation, not better.

I don't really see this conversation going anywhere, anyone who thinks CF is on the level of SLI (which btw, is far from perfect) after all of this needs to take another look.

Fastx · Apr 7, 2013

Amenx thanks for posting the screen shot which was interesting for me seeing the 3rd one down again only this time without smearing and loss of detail. I just still don't believe time of day was the reason for this but now maybe a setting more in that first video & screen shot but it was good to see your screen shot of that same location (& no problem this time) without the loss of detail and smearing! I didn't see this problem anywhere in the other videos FC3/BF3 so I think it was setting imo but I could be wrong but is what I believe now so far.

aaksheytalwar · Apr 7, 2013

Amd is too dark, I prefer nvidia colors which are brighter yet lighter.

UaVaj · Apr 7, 2013

amenx said:
My own screenshots of same places in other pics posted, at different times of night/day. This with my 660ti:

p.s. the wavyness of the store shutter door in 2nd pick is optical illusion. Click to enlarge pic, it goes away.

TnL is definitely turned on. thank for verifying that other review was on AMD's payroll.

Cloudfire777 · Apr 7, 2013

Probably just sloppy work by the reviewer.

PCPer isn`t on AMD`s payroll. They have been slamming AMD with negative articles about the CF stuttering for a while now. :biggrin:

UaVaj · Apr 7, 2013

amd's payroll was just a figure of speed. it is just sloppy work.

Skurge · Apr 7, 2013

BallaTheFeared said:
By using proper settings of course, but also having 250mb more memory than the 1GB cards you're comparing.

The slight increase in latency was offset by the fact it wasn't a stuttering mess like the Radeon cards. It was however, a trivial increase next to the power of the vsync warriors.

We can argue until we're blue in the face, however the perception of Radeon has gotten worse this generation, not better.

I don't really see this conversation going anywhere, anyone who thinks CF is on the level of SLI (which btw, is far from perfect) after all of this needs to take another look.

Don't get me wrong. I would not recommend CFX right now to anyone who isn't willing to fool around with radeonpro. but 6xxx CFX was just as good as SLI. The 570 had the same issue and the 650Ti/Boost also has the issue. 680/670 vs 7970/7950 is game dependent on the stutter. 7xxx was perfectly smooth throughout it's lifetime apart from a month or 2 where they had issues. nV cards have been having issues on and off for a few years now, but nobody calls them out on it.

I figured you to be a little more impartial than that.

bystander36 · Apr 7, 2013

Skurge said:
Don't get me wrong. I would not recommend CFX right now to anyone who isn't willing to fool around with radeonpro. but 6xxx CFX was just as good as SLI. The 570 had the same issue and the 650Ti/Boost also has the issue. 680/670 vs 7970/7950 is game dependent on the stutter. 7xxx was perfectly smooth throughout it's lifetime apart from a month or 2 where they had issues. nV cards have been having issues on and off for a few years now, but nobody calls them out on it.

I figured you to be a little more impartial than that.

I don't think that was true either: http://www.tomshardware.com/reviews/radeon-geforce-stutter-crossfire,2995-5.html

Image quality Who is better

Diamond Member

Diamond Member

Diamond Member

Senior member

Diamond Member

Senior member

Diamond Member

Senior member

Diamond Member

Diamond Member

Diamond Member

Diamond Member

Senior member

Diamond Member

Diamond Member

Lifer

Diamond Member

Diamond Member

Senior member

Diamond Member

Golden Member

Golden Member

Golden Member

Diamond Member

Diamond Member