Asking Ryan Smith of AT if a special examination could be done?

Page 18 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
Microstutter in FC3:
http://www.youtube.com/watch?v=AV_4etDNi8E&feature=player_embedded#

To me that looks worse on the 690 than the 7970. Doesn't the 690 have some hardware solution to the MS problem? Why is it present there?

EDIT: Sorry, was posted in the other thread too.

Is that true micro stutter or just illustrating times when the fps drops really super low and you notice the change? Also I can see stuttering on both setups when you present it in slow motion
 

thilanliyan

Lifer
Jun 21, 2005
11,912
2,130
126
Is that true micro stutter or just illustrating times when the fps drops really super low and you notice the change? Also I can see stuttering on both setups when you present it in slow motion

From what I read in the other thread...it could be an engine issue too.

Yes I can see it in both setups as well, but it is much more noticeable in the 690 setup.
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
Microstutter in FC3:
http://www.youtube.com/watch?v=AV_4etDNi8E&feature=player_embedded#

To me that looks worse on the 690 than the 7970. Doesn't the 690 have some hardware solution to the MS problem? Why is it present there?

EDIT: Sorry, was posted in the other thread too.

There are a few things that make me wonder about this video such as reports that the Dunia engine stutters below 60 fps no matter what type of setup you are using. Single GPU or dual GPU, and regardless of brand that engine gets funky below 60fps and has since Far Cry 2. Also, what type of camera are they using to record the screens? Maybe it's not a high enough speed camera and it's amplifying what is perceived to the viewer. Also, youtube in itself is garbage and has a 30fps cap.

I like what they did in their comparison, and I think doing that with a high speed camera is really the only way to truly test this.
 

blastingcap

Diamond Member
Sep 16, 2010
6,654
5
76
Play back hi-speed video in slow motion obviously... but someone's gotta do it first. There may be other ways but a hi-speed cam would be effective even if it would take forever to tabulate results frame by frame.
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
Play back hi-speed video in slow motion obviously... but someone's gotta do it first. There may be other ways but a hi-speed cam would be effective even if it would take forever to tabulate results frame by frame.

Would you even have to get as scientific as to go frame by frame to compare if you just watch the video's side by side and compare? Maybe even cutting the video's in half a stitching them in the middle would make it easier to see when one side jumps. That way people could use their own eyes to judge the slow motion video.
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
Microstutter in FC3:
http://www.youtube.com/watch?v=AV_4etDNi8E&feature=player_embedded#

To me that looks worse on the 690 than the 7970. Doesn't the 690 have some hardware solution to the MS problem? Why is it present there?

EDIT: Sorry, was posted in the other thread too.

http://forums.ubi.com/showthread.ph...inates-the-stuttering-Please-fix-UBI-!!/page5

Here's the UBI thread regarding the stuttering in FC3. Why I don't think this video is a good example of sli/xfire stuttering.
 

wand3r3r

Diamond Member
May 16, 2008
3,180
0
0
Would you even have to get as scientific as to go frame by frame to compare if you just watch the video's side by side and compare? Maybe even cutting the video's in half a stitching them in the middle would make it easier to see when one side jumps. That way people could use their own eyes to judge the slow motion video.

Maybe they can give us a download of the full speed video and we can try play them side by side ourselves if there is even video software that can run at least at the monitor refresh rate 60/120 fps.

Then we could at least see the subjective experience, although I don't know if this would even be beneficial, it'd still be interesting.

I don't propose this as the solution, merely an addition to any inquiry or research that will be conducted.
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
Maybe they can give us a download of the full speed video and we can try play them side by side ourselves if there is even video software that can run at least at the monitor refresh rate 60/120 fps.

Then we could at least see the subjective experience, although I don't know if this would even be beneficial, it'd still be interesting.

I don't propose this as the solution, merely an addition to any inquiry or research that will be conducted.

Actually, not telling us what brand was what video and then having a vote would be pretty cool too. No bias that way.
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
From what I read in the other thread...it could be an engine issue too.

Yes I can see it in both setups as well, but it is much more noticeable in the 690 setup.

As I said it could illustrate the lower fps because we know that Nvidia cards are not getting the same framerates. I am not entirely convinced this is microstutter.
 

lavaheadache

Diamond Member
Jan 28, 2005
6,893
14
81
the trouble with the game is, as it has already been stated, as soon as frame rates dip below 60 fps there is noticeable hitching regardless of whether or not you are using Amd or nVidia hardware.
 

VulgarDisplay

Diamond Member
Apr 3, 2009
6,193
2
76
the trouble with the game is, as it has already been stated, as soon as frame rates dip below 60 fps there is noticeable hitching regardless of whether or not you are using Amd or nVidia hardware.

Some think it's tied to the repeat rate of your keyboard. I haven't tested it yet, but I turned my repeat rate down to the lowest setting and I'll give it a shot later tonight to see if it fixes my problems.
 

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0
Measurable numbers mean nothing if we cannot see/perceive the difference. Agreed?

Some peole do see this...hence why the debat over microstutter started....some people even deny to this day.
I'm susceptible to microstutter...but because you are not...I cannot get any data...wtf?!
 

ICDP

Senior member
Nov 15, 2012
707
0
0
It's the FC3 graphics engine that's causing the stuttering. I get an average of around 35-38 FPS at 2560x1600 in FC3 on a single GPU and the stutter is perceptable. The gameplay is smooth because the stuttering for me is only noticable if I look for it by circling around with my view locked on the ground.

Compared to the video of multi GPU my single card is as smooth as butter. Now I know why I keep fighting the urge to go multi GPU.
 
Last edited:

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
Looks like Brent has a few things to say about Radeon Pro.

http://hardforum.com/showthread.php?t=1732720&page=2

Posts 24 & 25.

You also might see a few familiar posters there.

From: Brent Justice [mailto:brent@hardocp.com]
Sent: Thursday, December 06, 2012 10:09 AM
To: 'Kyle Bennett'
Subject: RE: Radeon Pro

Program was more complicated than I thought, there isn't just one setting that will fix things, it's combination of settings to try various ways to possibly reduce the stuttering. The "default" way, for me, did not work. The vsync control dynamic did not work, simply forcing vsync on with and without tripple buffering, did not work, however, there was one setting that did work, and it is rather simple, and you may not even need Radeon Pro for it, and that is to 1.) cap your framerate artificially lower than your average framerate or 2.) use a double vsync option, which caps the vsync to 30 fps.

So what I found is that basically the only way to really smooth out the framerates, make it appear smoother, is to cap the framerate 10-20 fps below the average you are getting in a game. Let's say you are getting average 50-55 fps in far cry 3. If you cap the framerate at 40 fps, or anything below 40, it smooths out, less stuttering. But simply setting Vsync on won't help, cause that sets vsync to 60 fps, that's above the average. What you have to do is cap the framerate, or if the game has the option (in this case far cry 3 does with vsync 2 frames option) then it caps the fps to 30 and it smooth.

If your average fps is over 60 fps, then vsync on should work most of the time, because it will cap the fps to 60 hz/fps.

So it all comes down to what framerates are you are getting in-game, and then basically set your fps cap 10-20 fps lower, and the lower framerate, is smoother. Obviously this is going to retard performance overall, and you aren't going to get the full performance out of crossfire because you are capping your fps well below what it can actually achieve.

I found no way to maintain the full performance of crossfire and at the same time smooth the stuttering out, which SLI does natively.

However, I do need to spend more time with Radeon Pro and test it in more scenerios, such as eyefinity and other games. There's so many different ways to configure things that something may work in some game, but I don't know. But as far as Far Cry 3, I was able to keep the stuttering smoothed out by capping the game to 40fps or 30fps.

What I can do, is in our Far Cry 3 full evaluation article, in the CrossFire section where I will show performance and experience, I can include all of this information for Far Cry 3 and show screenshots in Radeon Pro how I accomplished the smoothness, or what was required. But like I said, I could also do this without Radeon Pro by using the "Vsync 2 frames" option in Far Cry 3 itself, it just caps the fps to a solid 30 fps, and that's really the gist of it. I did not find an overall Radeon pro setting so far that just works in smoothing out things.

----------------------
Regards,
Brent Justice
HardOCP.com Video Cards

I think we may have to take Radeon Pro and smoothing out stuttering in crossfire on a game by game basis, there isn't an overall sweeping option that fixes the issue, and gives us full performance. The only option i've seen so far, is to just cap the framerate 10-20 fps below the average you are getting in the game. Obviously the average framerate will vary depending on what in-game quality settings/AA settings you have enabled and what performance level you are happy playing at. Then just cap the fps 10-20 fps below that, and its smooth. It's hardly a fix though, more like a band aid. I'll include all the information in the Far Cry 3 article in how to smooth that game out with CrosFire, but it's going to mean keeping the game at 30 or 40 fps.

----------------------
Regards,
Brent Justice
HardOCP.com Video Cards


From the posts you referenced he only tried it in one game, FC3. A game that by others accounts suffers from stuttering not because of hardware. It doesn't seem very conclusive.
 

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0
From the posts you referenced he only tried it in one game, FC3. A game that by others accounts suffers from stuttering not because of hardware. It doesn't seem very conclusive.

Hence why Keys asked for a more indepth look from AT....you are slow today arn't you?
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
Hence why Keys asked for a more indepth look from AT....you are slow today arn't you?

Don't be ignorant. You don't know me.

I'm just stating that it's one game that's known to have stuttering problems anyway. So, what does it prove? Nothing. It was a waste of time for Brent to test with it, a waste of time for him to have posted it, a waste of time linking to it, and a waste of time reading it.
 

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0
Don't be ignorant. You don't know me.

I'm just stating that it's one game that's known to have stuttering problems anyway. So, what does it prove? Nothing. It was a waste of time for Brent to test with it, a waste of time for him to have posted it, a waste of time linking to it, and a waste of time reading it.

You consider factfinding a waste of time....gotcha.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
Don't be ignorant. You don't know me.

I'm just stating that it's one game that's known to have stuttering problems anyway. So, what does it prove? Nothing. It was a waste of time for Brent to test with it, a waste of time for him to have posted it, a waste of time linking to it, and a waste of time reading it.

And a waste of time reading your posts showing that all you want is for this to go away. It isn't. Okay? If you do not support an investigation by not only Anandtech, but as many tech review sites as possible, then you actually have no reason to post in this thread. After all, this thread is a request for just such a thing. You could start another thread about opposing it though. Fair is fair.
 

pandemonium

Golden Member
Mar 17, 2011
1,777
76
91
I'm all for a verifiable metric to weigh frametimes versus framerates and think this has a valid place in reviewers responsibilities.

I think to eleviate any confusion about those weights, the metric needs to be deviation of frametime against ideal framerates. By this, I mean that if a card is producing 60 FPS, then the deviation from that frametime (being 16.7ms) should be <= a certain amount showing microstutter is absent. Say, 99%? However - as example - the 99th percentile metric that techreport gives isn't applicable across all cards in comparison; it's unfortunately hardly relevant and doesn't explain much without looking at the data in detail. Sure, a setup with frametimes that sit more outside the 99th percentile is going to have more perceptual microstutter, but the question is, how much exactly and how do we compare that to everything else?

To explain further, in order to accurately gauge microstutter accross multiple configurations, each setup needs to be tested for their FPS. Once that is done, frametimes need to be extrapolated against the FPS adjacently. That means, don't just take the average FPS and weigh it against the frametimes it should be. Instead use the frametimes it actually has on each frame, in accordance to the frame rate being produced. Then take how many of those were abhorrent and label those outside the 99th percentile. This will actually plot how much microstutter a setup has, giving a more in depth and accurate representation of what many people observe as microstutter.

*-*
The whole argument of, "if you can't see it, then it's irrelevant" needs to be shot down and never come again. If you couldn't see the FPS being produced and were perfectly happy with 15FPS, then there wouldn't be much purpose to review graphics cards other than their boxes are pretty and the HSF is cool looking! The standards have been raised...if microstutter doesn't bother you, then you're not perceptive enough to see it, you're in denial, your eyes and brain aren't trained to recognize it, or you just don't care. Either way, you're out of this discussion.
*-*
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
And a waste of time reading your posts showing that all you want is for this to go away. It isn't. Okay? If you do not support an investigation by not only Anandtech, but as many tech review sites as possible, then you actually have no reason to post in this thread. After all, this thread is a request for just such a thing. You could start another thread about opposing it though. Fair is fair.

I'm not opposing investigation. That's a game specific problem, stuttering in FC3. It's not a relevant test for the software or comparing architectures. If you just want to study stuttering in FC3, you start a thread. Or, better yet join one already in progress.

Absolutely nothing eliminates the stuttering. Please fix UBI !!
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
I'm not opposing investigation. That's a game specific problem, stuttering in FC3. It's not a relevant test for the software or comparing architectures. If you just want to study stuttering in FC3, you start a thread. Or, better yet join one already in progress.

Absolutely nothing eliminates the stuttering. Please fix UBI !!

So you're going to maintain that between TR and H, the only game they reported on for this issue is FC3?
Because this thread isn't just about far cry 3 although I see some attempted steerage in that direction. Well, you can stop now.
You all keep saying you don't oppose this, but take a stroll back through this thread. I'm talking a dumptruck full of charmin couldn't clean this up.
 
Last edited:

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
So you're going to maintain that between TR and H, the only game they reported on for this issue is FC3?
Because this thread isn't just about far cry 3 although I see some attempted steerage in that direction. Well, you can stop now.
You all keep saying you don't oppose this, but take a stroll back through this thread. I'm talking a dumptruck full of charmin couldn't clean this up.

No. Your post I responded to was about FC3. Stop trying to make it seem like I'm saying something I'm not. The post did nothing to address the topic of the thread. It's a game specific issue. That thread on UBI forum was started by someone who owns an nVidia card.
 

Keysplayr

Elite Member
Jan 16, 2003
21,209
50
91
No. Your post I responded to was about FC3. Stop trying to make it seem like I'm saying something I'm not. The post did nothing to address the topic of the thread. It's a game specific issue. That thread on UBI forum was started by someone who owns an nVidia card.

My apologies. So what is the overall message you are trying to send here?
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |