FCAT: The Evolution of Frame Interval Benchmarking

[Grooveriding]

The interesting thing is

Frame time variance is relatively high for both graphics setups on this game. The GeForce cards seemed choppier, based on my experience, but that's a subjective call.

The spikiness from AMD's Radeon cards can be demonstrated by looking at frame time variance. At the 95th percentile, we can see that frames aren't being delivered as consistently. I did notice the game felt a little laggier, too.

I saw that as well and it feels in harmony with some of what Ryan shared in AMD addressing latency and intervals; the games play a role and each vendor has to account for them in drivers.

Like you see in your graphs it goes back and forth between them depending on the game.

 

[tviceman]

Awesome tool by Nvidia. Should bring some good information to light when various websites do their testing. Thumbs up to the green team!

 

[f1sherman]

more good news

 

[Will Robinson]

Hey wait...weren't Keysplayr and Lonbjerg so sure that the Never Settle drivers had traded off smoothness for FPS?

railven
Platinum Member

Join Date: Mar 2010
Posts: 2,338

Okay, so there are a few threads about this issue (seems a lot of sites dropped their articles on the same date), but, dunno where to post anymore haha.

Either way from the PCPer article based on all the hooplah that occured here, i was not honestly expecting this conclusion:

Quote:
Finally, one of the newest games to our test suite, Sleeping Dogs, the AMD Radeon HD 7970 holds a sizeable advantage across the board of the three tested resolutions. The margins are 34% at 1920x1080, 37% at 2560x1440 and 23% when using triple displays.

While some people might have assumed that this new testing methodology would paint a prettier picture of NVIDIA’s current GPU lineup across the board (due to its involvement in some tools), with single card configurations nothing much is changing in how we view these comparisons. The Radeon HD 7970 GHz Edition and its 3GB frame buffer is still a faster graphics card than a stock GeForce GTX 680 2GB GPU. In my testing there was only a couple of instances in which the experience on the GTX 680 was faster or smoother than the HD 7970 at 1920x1080, 2560x1440 or even 5760x1080.
So much for AMD cheating with the Never Settle driver package. People were waiting for this article to set things straight, and I hope they accept this as their answer.
__________________

Guess the silence will be deafening from them now...

 

[omeds]

Can anyone else spot the problems with this picture?

That is not how frames are drawn on the screen at all.

Yes they are, if thats the timings at which they are recieved.

 

[Keysplayr]

Hey wait...weren't Keysplayr and Lonbjerg so sure that the Never Settle drivers had traded off smoothness for FPS?

Guess the silence will be deafening from them now...

I know I suspected it. Still do in some cases. The games that were focused on with AMD's first hotfix (3 or 4 games) did show an 8% decrease in one game for less frame latency. Another game dropped only 1 or two frames, and another not affected at all.

Will, the callouts have to stop. Go get your jollies some other way.
You feed off this conflict like a leech.

 

[VulgarDisplay]

Yes they are, if thats the timings at which they are recieved.

The picture illustrates the timings that a game is outputting, but the "Displayed" side is incorrect because frames are drawn from the top to the bottom of the screen. So the light blue frame should be at the bottom of display refresh one, and the red frame should be at the top of display refresh one. In the next display refresh frame the purple frame should be on the top, and the remainder of the red frame should be on the bottom. In display refresh 3 the sliver of purple frame should be at the bottom, the yellow frame should be drawn down to a little over halfway, and the black frame should be starting at the top.

The picture from the article gets the point across that pushing frames out of sync with your monitor is bad, but it makes it look far to tidy by just lining the frames up like that. There should be a jumbled mess on the display refresh side. I'd try and whip something up to show what I'm talking about, but I don't have photoshop installed on this computer.

 

[omeds]

No it shouldn't, thats exactly how it will look. I have no idea what you are on about tbh, that looks like a good representation to me, I think you are just not understanding it correctly?


Refresh 1

1111111111111
1111111111111
1111111111111
2222222222222
2222222222222
2222222222222

Refresh 2

2222222222222
2222222222222
2222222222222
3333333333333
3333333333333
3333333333333

Refresh 3

3333333333333
3333333333333
3333333333333
4444444444444
4444444444444
4444444444444

and so on

 

[VulgarDisplay]

This is what it should look like. Frames are drawn from the top of the screen to the bottom, and that does not change.



Uploaded with ImageShack.us

Now with 100% more MSpaint butchering.

 

[omeds]

Yes frames are drawn from top to bottom, so why have you gone from bottom to top?? It was right the first time. The original diagram even shows you the location of tear lines.

 

[VulgarDisplay]

Yes frames are drawn from top to bottom, so why have you gone from bottom to top?? It was right the first time. The original diagram even shows you the location of tear lines.

Replace game frame with a picture. You're telling me that the top of that picture in game frame 2 appears at bottom of what is put out on your monitor? Then the bottom of Frame two appears at the top of the monitor at refresh 3?

That is not how it works, each refresh rate is a picture that is "flashed" at your eyes, so you need to stretch the game frame out to fill the whole refresh rate. If the picture gets cut off halfway through one refresh, the bottom half of that frame is still drawn in the next refresh rate. What we are seeing here is the next frame coming in during the refresh rate instead of in perfect sync with it which is why you see the next frame appearing at the top where it started being drawn.

 

[Lonbjerg]

Hey wait...weren't Keysplayr and Lonbjerg so sure that the Never Settle drivers had traded off smoothness for FPS?

Guess the silence will be deafening from them now...

Link or you are reported for lying...Again.

 

[VulgarDisplay]

Link or you are reported for lying...Again.

Nice red herring.

 

[omeds]

Replace game frame with a picture. You're telling me that the top of that picture in game frame 2 appears at bottom of what is put out on your monitor? Then the bottom of Frame two appears at the top of the monitor at refresh 3?

That is not how it works, each refresh rate is a picture that is "flashed" at your eyes, so you need to stretch the game frame out to fill the whole refresh rate. If the picture gets cut off halfway through one refresh, the bottom half of that frame is still drawn in the next refresh rate. What we are seeing here is the next frame coming in during the refresh rate instead of in perfect sync with it which is why you see the next frame appearing at the top where it started being drawn.

No I think you misunderstand how it works. It's not a film reel.

 

[VulgarDisplay]

No I think you misunderstand how it works. It's not a film reel.

Which is how you are looking at it. I know it's not, but you don't based upon your explanations.

Game Frame on the right hand side is not the actual size of the frame being output. It's the timing of when the game is sending a frame. The frame is the actual size of the window you see in Refresh rate. If you look closely I have crudely matched the size of the things I have drawn over to match the amount of game frame time that is being show per refresh. I just rearranged them so they would be in the correct order in each refresh window.

 

[Lonbjerg]

Nice red herring.

You do see the Paradox here right?
Of to reporting.

 

[VulgarDisplay]

You do see the Paradox here right?
Of to reporting.

Yes, you're the paradox because you defy common sense.

 

[omeds]

Which is how you are looking at it. I know it's not, but you don't based upon your explanations.

Game Frame on the right hand side is not the actual size of the frame being output. It's the timing of when the game is sending a frame. The frame is the actual size of the window you see in Refresh rate. If you look closely I have crudely matched the size of the things I have drawn over to match the amount of game frame time that is being show per refresh. I just rearranged them so they would be in the correct order in each refresh window.

No I'm not at all. You seem to misunderstand that graph represents time, not the physical frames. The left is the timing of frames, the right is the timing of refresh cycles. The lines from left to right are the tearing lines as they land mid refresh, when changing frames.

 

[Lonbjerg]

Yes, you're the paradox because you defy common sense.

Take it too PM...I got called out by name by Will...so I respond.
What you are doing are nothing but threadcapping.
Reported.

 

[MrK6]

I know I suspected it. Still do in some cases. The games that were focused on with AMD's first hotfix (3 or 4 games) did show an 8% decrease in one game for less frame latency. Another game dropped only 1 or two frames, and another not affected at all.

Will, the callouts have to stop. Go get your jollies some other way.
You feed off this conflict like a leech.

:biggrin:
Oh man, pot, meet kettle.

So after all the chest thumping and team promoting we still come back to the fact that AFR is has faults that no one has yet to overcome. Good job, fanboys.

 

[USER8000]

It's the same:
http://www.pcper.com/reviews/Graphi...ils-Capture-based-Graphics-Performance-Testin

http://blogs.nvidia.com/2013/03/with-our-new-tool-what-you-see-is-what-you-get/

They linked back to what PC Perspective have done as a usage of FCAT.

It appears PC Perspective got access to FCAT weeks before other websites.

However,in their articles they seemed to have indicated it is an independent effort on their part:

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-New-Graphics-Performance-Metric

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-Part-2-Finding-and-Defining-Stutter

Were they under NDA??

 

[blackened23]

Take it too PM...I got called out by name by Will...so I respond.
What you are doing are nothing but threadcapping.
Reported.

LOL

Lonbjerg is certainly above threadcrapping. His opinions are completely fair and objective and always stated in a calm and concise manner. If only everyone should emulate you.

 

[Imouto]

http://blogs.nvidia.com/2013/03/with-our-new-tool-what-you-see-is-what-you-get/

They linked back to what PC Perspective have done as a usage of FCAT.

It appears PC Perspective got access to FCAT weeks before other websites.

However,in their articles they seemed to have indicated it is an independent effort on their part:

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-New-Graphics-Performance-Metric

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-Part-2-Finding-and-Defining-Stutter

Were they under NDA??

PCPer guy just likes to brag. If none of the tech sites that reviewed FCAT mentioned PCPer and even Nvidia itself just linked their site without giving them any credit... the whole issue is crystal clear.

Will be obvious when this tool will be publicly available since everyone will be able to read the GPL manifest.

 

[AdamK47]

LOL

Lonbjerg is certainly above threadcrapping. His opinions are completely fair and objective and always stated in a calm and concise manner. If only everyone should emulate you.

Lonbjerg posts invariably associate with threadcrapping. Where Lonbjerg posts, threadcrapping follows.

 

[Ryan Smith]

http://blogs.nvidia.com/2013/03/with-our-new-tool-what-you-see-is-what-you-get/

They linked back to what PC Perspective have done as a usage of FCAT.

It appears PC Perspective got access to FCAT weeks before other websites.

However,in their articles they seemed to have indicated it is an independent effort on their part:

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-New-Graphics-Performance-Metric

http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-Part-2-Finding-and-Defining-Stutter

Were they under NDA??

PCPer was beta testing FCAT a few months before its wider release. They already had the equipment, so when NVIDIA was ready to try it outside the company it was a good match for them.