(Youtube) Frame Rating and FCAT Explanation and Discussion with NVIDIA *PCPER*

[BallaTheFeared]

http://www.youtube.com/watch?v=2cH_ozvn0gA

 

[Imouto]

Dear lord, this again?

 

[Keysplayr]

Dear lord, this again?

What do you mean, "again" ?
This is ongoing. Probably will be for a while.
If you're here to crap, sorry, we are all out of toilet paper. Just don't participate if this subject is a sore one for you.
TIA.

 

[BallaTheFeared]

Dear lord, this again?

Yes more knowledge and insight, haters of such beware.

 

[Keysplayr]

AMD really did improve with their 7990 II prototype drivers. Nice work. They still have a ways to go though. Nvidia has a bit of work to do also with some inconsistencies with the 690.

 

[wand3r3r]

It'll be ongoing until the situation is reversed, suddenly it'll be irrelevant.
/ironic sarcasm

Let the show continue...

 

[Keysplayr]

It'll be ongoing until the situation is reversed, suddenly it'll be irrelevant.
/ironic sarcasm

Let the show continue...

I highly doubt that. Lets be a bit more realistic shall we instead of the bitter snipey remarks. K?

 

[Black Octagon]

Great video (it would have been even better with AMD in the room, but this was about an NVIDIA tool after all). Thanks for sharing

 

[f1sherman]

wow an hour and a half of uber geekness

how did I miss this

 

[Black Octagon]

wow an hour and a half of uber geekness

how did I miss this

You didn't though...

 

[WiseUp216]

http://www.youtube.com/watch?v=2cH_ozvn0gA

Thanks for sharing this.

I tend to avoid most of the threads in VC&G, so something like this is helpful to those of us who don't read about it everyday.

 

[omeds]

Excellent bit in there about input lag, SLI, and frame metering. Turns out SLI with frame metering actually has less input lag than a single GPU.

Also, confirms what I have always noticed, using a frame cap actually increases input lag.

 

[Keysplayr]

And everyone's solution for V-sync input lag is said to be triple buffering. Is this a load of BS? Does it do anything to stop the "back-pressure" as Tom Petersen likened it to?

 

[omeds]

There is still input lag with triple buffering, or 59fps caps.

 

[Erenhardt]

There is still input lag with triple buffering, or 59fps caps.

technically... there is always input lag.
Is there anyone here who lives of competitive gaming? I can understand that he may have problems with input lag, but other than that... Don't make me lough.
This is stupid. You need special tools to find out if there is stuttering or not? You kidding? If you can't see it, what is the difference? Or is it the same thing as with FPS counter in the corner of a screen. Every game was smooth and playable until you didn't put that counter showing 30fps. Suddenly you feel like watching google picasa web album.

 

[omeds]

I can feel input lag with vsync, using TB or 59fps cap *shrug*

30fps is smooth and playable to you??

 

[socialdarwinist]

I can feel input lag with vsync, using TB or 59fps cap *shrug*

30fps is smooth and playable to you??

For me it's game to game. Some can seem relatively smooth at 30fps, but I don't have a 120hz monitor running 120fps consistently. I suspect if I got used to that, 30fps would seem like molasses. All relative.

 

[Keysplayr]

technically... there is always input lag.
Is there anyone here who lives of competitive gaming? I can understand that he may have problems with input lag, but other than that... Don't make me lough.
This is stupid. You need special tools to find out if there is stuttering or not? You kidding? If you can't see it, what is the difference? Or is it the same thing as with FPS counter in the corner of a screen. Every game was smooth and playable until you didn't put that counter showing 30fps. Suddenly you feel like watching google picasa web album.

Oye..... But it CAN be seen and experienced. Why else have these discussions if there wasn't any downside to it and all was peachy? Use that gray matter between yer ears.

 

[bystander36]

This latency causes me to get sick, so it is quite important to me. I find 60 FPS to be a minimum for an enjoyable experience, if the game is 1st person and mouse controlled. At 3rd person, I can tolerate a fair bit less.

 

[bystander36]

Dear lord, this again?

FCAT reviews are going to be the new normal. It may not be used on every site, but those with the tool will be using it often. Get used to it.

 

[BrightCandle]

Frame times and FCAT put absolute numbers onto what is a subjective measure. We still need to spend some time working out what acceptable patterns are as well as absolute swing maximums but now we have the data that sort of analysis can be done along with the videos.

One of the key pieces of information that comes out of this talk is NVidias range of 2ms - 10ms, where at 2ms no one notices the variation and 10ms almost everyone sees it. That tallies with the 8ms I have determined from all the traces we have seen in these forums. So we can probably put a slightly tighter bound on the noticeable variation than NVidia but its good to see their numbers tallying. The fact they have a lower bound is also great news, that gives us something to work with not only to say that above this amount is unacceptable but that below this point it clearly doesn't matter. Everything else in between a number of people but not all people will perceive.

One of the insights I have from writing a tool that measures the reason for frame times is that its almost always the GPU and its driver causing the stuttering, that I didn't expect. I also figured these variations would come from the CPU but they don't. Once I deliver this tool for public consumption this analysis is going to take another step forward, suddenly we'll be able to say whether its the game or the GPU and its drivers that are at fault.

 

[Revolution 11]

Ignore the naysayers, this video was very informative and educational on the subject of frame latency, variation, etc. I didn't get all the attention about this but now I do. Thank you, Nvidia and PC Perspective, for the good work.

 

[Erenhardt]

One of the insights I have from writing a tool that measures the reason for frame times is that its almost always the GPU and its driver causing the stuttering, that I didn't expect. I also figured these variations would come from the CPU but they don't. Once I deliver this tool for public consumption this analysis is going to take another step forward, suddenly we'll be able to say whether its the game or the GPU and its drivers that are at fault.

Measuring the reasons... sounds legit

First and last reason why all this FCAT is BS was in the video(@0:40:30). People think they found a solution for stuttering in frame limiter or vsync. And what do they say?"No, my tool measures that there is still stuttering. Don't let your eyes fool you!".

But I don't care since I don't use multi GPU setup.

 

[Keysplayr]

Remain in the dark if you wish. Lord knows its your choice albeit an unfortunate one.
I really don't care if you continue as there isn't anything you can do to stop this new form of gpu testing no matter how much you kick and scream. So have at it, it affects nothing. Well, except maybe how you are perceived.

 

[Rvenger]

The video wasn't really bias at all. It was a very informative video and kudos to Nvidia for developing metering software that shows basically the reality of what we see. The software works for both camps and you can clearly see that AMD is fixing their issues too.