A Short study in microstutter

[Rikard]

Try using the RadeonPro benchmarking tool. If you're using Windows 8, I know that Fraps doesn't play well with certain configurations.

Regarding OCs, check that Powertune is set to +20%.

Thanks for the advice, but I am running Win7 and I have Powertune +20% already. I still do not know what went on with that one Heaven+Fraps benchmark, but I have not be able to reproduce it...

Anyway, here is Hitman Absolution (Ultra, 1080p):
Min 44, Max 88, Average 66

No sign of any microstuttering there.

 

[njdevilsfan87]

If you had microstutter youd have a complete pause in gameplay followed by normal frame rates. You had had a driver problem.

Wrong.

In layman's terms, you have microstutter when your frame rate counter shows 40fps but it feels like 25. And it will show up as consistent release spikes on the graphs like the one in the post above.

 

[BrightCandle]

Microstutter has always been described in terms of sli/xfire because no one noticed it on single cards. Now we do notice it that definition is pointing at the cause, not the effect.

Microstutter IMO is simply periodic patterns of differing frame times. When those swing differences are sufficiently large (>5ms) some people perceive the uneven delivery of frames. We call that microstutter.

Stuttering is just infrequently very wildly different frame times caused by one off events. We all feel those jolts but they are a different problem.

Microstutter is often alternate frames, but sometimes its groups of 3 or more. Neither less you can see a periodic pattern when you graph it that shows swings in highs and lows of frame times in very short periods of time. The ideal 60 fps chart would be a straight line at 16.6ms as a comparison.

I am interested in when people perceive it. I obviously perceive it at quite low amounts of stutter as does Dan but clearly many people are happy at much higher than that.

 

[BoFox]

If you had microstutter youd have a complete pause in gameplay followed by normal frame rates. You had had a driver problem.

No, that's not microstutter. A complete pause is more like a stutter or a hitch, not a MICROstutter. Microstuttering is much more constant, in making 60fps CONSISTENTLY feel like 30-40fps, due to the "micro"stutter in between every other frame.

EDIT - A guy beat me to it 2 posts above.. I just skimmed the thread and overlooked that post, ARGH!

 

[Final8ty]

No, that's not microstutter. A complete pause is more like a stutter or a hitch, not a MICROstutter. Microstuttering is much more constant, in making 60fps CONSISTENTLY feel like 30-40fps, due to the "micro"stutter in between every other frame.

EDIT - A guy beat me to it 2 posts above.. I just skimmed the thread and overlooked that post, ARGH!

I have had 120fps that felt like 20fps in EVE-Online once, for some reason it would not Vsync to 60 but locked itself to 120fps.

 

[BoFox]

I have had 120fps that felt like 20fps in EVE-Online once, for some reason it would not Vsync to 60 but locked itself to 120fps.

Haven't played that game myself, but it must've been the game's internal rendering engine that ran things at only 20fps - maybe a glitch or something.. kind of like Bioshock's characters moving around at only 30fps even though the game environment moves around at 60+fps. Maybe just some strange messed up stuff!

 

[Final8ty]

Haven't played that game myself, but it must've been the game's internal rendering engine that ran things at only 20fps - maybe a glitch or something.. kind of like Bioshock's characters moving around at only 30fps even though the game environment moves around at 60+fps. Maybe just some strange messed up stuff!

Well EVE-Online does get allot of patches and sometimes they break thing but they fixed that one quickly.

 

[BD231]

The definition above doesn't differentiate between plain stuttering and "micro-stuttering" - both have a large instantaneous change in framerate relative to actual framerate (that is, the 2nd derivative of the framerate is large). However, in pretty much all tech discussions (let's exclude this month from consideration), micro-stuttering refers to the specific phenomenon caused by SLI/XFire solutions, while stuttering is a phenomenon affected by other factors (drivers, disk access, CPU use, bad program, etc)

This thread paints a different picture: http://hardforum.com/showthread.php?t=1317582



From there, it seems that microstuttering is more of a synchronization issue - note that the changes in framerate are large, but periodic. Normal stuttering does not exhibit any predictable periodicity. Neither should the observations by TR be described as "microstuttering".

This guy has it right, unless you guys just feel like changing the term to fit your driver problem, which I guess you do.

 

[BD231]

No, that's not microstutter. A complete pause is more like a stutter or a hitch, not a MICROstutter. Microstuttering is much more constant, in making 60fps CONSISTENTLY feel like 30-40fps, due to the "micro"stutter in between every other frame.

EDIT - A guy beat me to it 2 posts above.. I just skimmed the thread and overlooked that post, ARGH!

Yeah, no it's not. You're just keying in on words.

 

[Dankk]

This guy has it right, unless you guys just feel like changing the term to fit your driver problem, which I guess you do.

Okay. Help me understand:

Basically, some members here are saying that, by definition, micro-stutter is a symptom caused by multi-GPU configurations, and nothing else.

Why does micro-stuttering have to be exclusive to multi-GPU setups? Because as far as I'm concerned, it's not exclusive at all. Single-GPU micro-stuttering is a very real, hotly-discussed topic on this forum right now, judging by the several other threads we have on it.

The cause may be different, but the effect is exactly the same: Minor, inconsistent frame delivery that happens in intervals of less than a second, and thus not registered by regular FPS measurement. In other words: Micro-stuttering.

Microstutter IMO is simply periodic patterns of differing frame times. When those swing differences are sufficiently large (>5ms) some people perceive the uneven delivery of frames. We call that microstutter.

Seems pretty simple to me. Please tell me: How, by definition, does it have to be caused by multiple GPU's? Micro-stuttering is a symptom of multi-GPUs, but who says a symptom can't have multiple causes?

I'd hope you agree with this logic:

Cause A > Symptom A.
Cause B > Symptom A.

Generally speaking, it's entirely possible that two different actions can have the same reaction; no? A single symptom can have one or more causes.

But according to others, micro-stuttering can only be caused by multiple GPUs, and nothing else.

So... what is it then? Look at those graphs posted above. What is it? If it's not micro-stuttering, then what is it?

Wait. I already know how you're going to answer:

It's a driver problem.

The driver was the cause. The micro-stutter was the symptom. Multiple GPU's wasn't the cause; in this case, it was something else entirely.

This whole time, you've been arguing "driver problems." You do realize that I already fixed that before this thread even started, don't you? Do you realize that those benchmarks I posted above were well after I fixed the problem with the clocks? Did you even read the whole thread?

Someone explain to me how micro-stuttering is denotatively exclusive to multi GPUs. I'm still not getting it.

Edit: And I'm sorry for sounding accusatory. I guess it's a bit silly to be arguing this much over semantics.

 

[BD231]

You're confused because some people around here seem to apply a hooked on phonics like mentality to the term microstutter, its tiny and it stutters ! The true topic of this thread is a study in poor frame time delivery which has been covered by credible sources.

 

[KingFatty]

Hey I wonder, are these subjects (e.g., microstutter in single and multi GPU systems) pretty ripe for academic papers? Are there computer engineering students today that are trying to figure out what topic to use for their senior papers? If I could go back in time I'd try to dig in and tackle this subject using lab equipment at the university.

Anyway I agree that the term "microstutter" may throw off some people because it's definition is being clarified. So many people associate microstutter with multi-GPU, when I don't think that's necessary. It's just a variation between frame times, that causes a stuttering effect.

 

[BoFox]

You're confused because some people around here seem to apply a hooked on phonics like mentality to the term microstutter, its tiny and it stutters ! The true topic of this thread is a study in poor frame time delivery which has been covered by credible sources.

Take it easy, LOL.

 

[njdevilsfan87]

Anyway I agree that the term "microstutter" may throw off some people because it's definition is being clarified. So many people associate microstutter with multi-GPU, when I don't think that's necessary. It's just a variation between frame times, that causes a stuttering effect.

It's because multi-GPU have a very consistent and noticeable problem with micro-stutter. Single GPU have it as well, but no where near the amount that multi-GPU setups have it. AMD has something weird going on with their current drivers causing Skyrim to experience noticeable micro-stutter on single GPU setups.

The occasional stutter caused for example by the game accessing something of a mechanical hard drive is not micro-sutter because the delay is too large (and it just becomes termed stutter without the micro part). You normally wouldn't notice micro-stutter (hence it being called micro) unless it's consistently happening like in the form of multi-GPU setups.

 

[Final8ty]

It's because multi-GPU have a very consistent and noticeable problem with micro-stutter. Single GPU have it as well, but no where near the amount that multi-GPU setups have it. AMD has something weird going on with their current drivers causing Skyrim to experience noticeable micro-stutter on single GPU setups.

The occasional stutter caused for example by the game accessing something of a mechanical hard drive is not micro-sutter because the delay is too large (and it just becomes termed stutter without the micro part). You normally wouldn't notice micro-stutter (hence it being called micro) unless it's consistently happening like in the form of multi-GPU setups.

Looking at the Skyrim video it looked like hitching,tearing and not Micro strutter as in fact without slow motion on that video you would not be able to see micro-stutter on Youtube which is 30fps.