How do you notice frame "time" issues?

From the pixels, bits, and shaders to the graphic cards that power them. Discuss the latest from AMD and NVIDIA here.

Moderators: morphine, SecretSquirrel

How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 9:33 am

Now that I've got a 7950, up from 6850 and 4850, my games are butter smooth, so how do I tell that I'm suffering from frame time problems.
With my older system and GPU's, it was easy to induce lag and noticeable stutter, but Crysis 1 and Metro2033 1920x1200 maxed are supa smooth.
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 9:59 am

It's simple, you just do. If you don't be happy because it means you are playing games that are not affected by bad video drivers.

If you are asking "How can you determine it's actually high frame times and not something else?". That can be done by checking for 2 things, average fps (it must be 60 or higher) and latency (if you play multiplyer, preferably under 40-30ms, the lower the better). If those check out, the "jerkiness" is probably caused by bad drivers which cause high frame time issues.

On multiplayer though, there can be another cause (though it's pretty rare these days) server hardware failure/malfunction. I say it's rare today because private servers are a rarity.
Last edited by Arclight on Tue Mar 19, 2013 2:52 pm, edited 1 time in total.
nVidia video drivers FAIL, click for more info
Disclaimer: All answers and suggestions are provided by an enthusiastic amateur and are therefore without warranty either explicit or implicit. Basically you use my suggestions at your own risk.
Arclight
Gerbil Elite
 
Posts: 695
Joined: Tue Feb 01, 2011 3:50 am

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 1:04 pm

Now that I've got a 7950, up from 6850 and 4850, my games are butter smooth, so how do I tell that I'm suffering from frame time problems.
if you don't see it you aren't suffering.

yet more blowback from the overhyped "framelatencygate" that redefined how the web does gfx benches because of results shown from altered video.
neg
clone
Gerbil Elite
 
Posts: 900
Joined: Mon Mar 12, 2007 10:40 am

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 2:45 pm

clone wrote:yet more blowback from the overhyped "framelatencygate" that redefined how the web does gfx benches because of results shown from altered video.

Are you really accusing TR (along with a slew of other tech sites) of doctoring their videos of the stuttering issues seen with many common games/GPUs? :-?
Z68XP-UD4 | 2700K @ 4.4 GHz | 16 GB | 770 | PCP&C Silencer 950 | XSPC RX360 | Heatkiller R3 | D5 + RP-452X2 | HAF 932 | 1 TB WD Black w/ SRT
Waco
Gerbil Elite
 
Posts: 743
Joined: Tue Jan 20, 2009 4:14 pm

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 3:17 pm

Waco wrote:
clone wrote:yet more blowback from the overhyped "framelatencygate" that redefined how the web does gfx benches because of results shown from altered video.

Are you really accusing TR (along with a slew of other tech sites) of doctoring their videos of the stuttering issues seen with many common games/GPUs? :-?

...only when it comes to AMD products :wink:

But I actually agree with the part about "if you don't notice it - you are not suffering". The pretty charts and graphs may show all kinds of abnormalities (which AMD can fix through drivers), BUT if you don't notice any during actual game play - these "issues" are irrelevant to all except the reviewing websites (which earn more Ad $$$ from such "breaking controversies") :wink:
My subscription allows you people to exist on this site and makes me a better human being than you'll ever be
JohnC
Gerbil Jedi
Gold subscriber
 
 
Posts: 1862
Joined: Fri Jan 28, 2011 2:08 pm
Location: NY/NJ/FL

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 3:25 pm

CaptTomato wrote:Now that I've got a 7950, up from 6850 and 4850, my games are butter smooth, so how do I tell that I'm suffering from frame time problems.
With my older system and GPU's, it was easy to induce lag and noticeable stutter, but Crysis 1 and Metro2033 1920x1200 maxed are supa smooth.


That's kind of like asking "How do I know if my leg hurts?"

If you aren't noticing, then there isn't really a problem. There might very well be some time frame issues taking place... but if you aren't noticing them...
cphite
Gerbil Elite
 
Posts: 543
Joined: Thu Apr 29, 2010 9:28 am

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 3:33 pm

So far with the 7950 and also with the jump from 4850-6850, I noticed that as long as FPS exceed 50, then it's extremely smooth.
I tried playing Crysis1 with the 6850 maxed+2xaa and even that induced stutter during firefights, but with 7950, Crysis is maxed+8xaa and silky smooth, in fact I was marveling at how slick the gun and game mechanics are.

It's interesting as there's often times where AMD has higher frames than Nvidia, and the frames are high in general, yet frametime suggests the parties over for AMD.

I also loaded Skyrim maxed, and that was as clear as day and smooth as well.
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 3:40 pm

cphite wrote:
If you aren't noticing, then there isn't really a problem


But does anyone notice, even if the frametimes are high...?
I remember reading that people move from crossfire to a single Nvidia and whilst FPS drop, gameplay is improved, and that seems believable as microstutter appears to be a CPU limitation of trying to run 2 GPU's, but it would also seem reasonable that for example, moving from 7850 XF to a single 7970 would be a smoother experience.
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 4:04 pm

CaptTomato wrote:
cphite wrote:
If you aren't noticing, then there isn't really a problem


But does anyone notice, even if the frametimes are high...?
I remember reading that people move from crossfire to a single Nvidia and whilst FPS drop, gameplay is improved, and that seems believable as microstutter appears to be a CPU limitation of trying to run 2 GPU's, but it would also seem reasonable that for example, moving from 7850 XF to a single 7970 would be a smoother experience.


Going from 6950 CF to a single GTX670, was like night and day. I'd assume that a single 7970 would be a similar performance experience, maybe even better depending, but the things were so damn loud and hot.

What I noticed most was an input-lag heavy experience in BF3 that seemed choppy even though FRAPS was running in the 60-70FPS range. Dropping below around 45FPS was a stutterfest. The GTX670 (and presumably any other single card that's about as fast at 2560x1600) is perfectly smooth at all of the above framerates. And I have much higher hopes that a second GTX670 will be smoother than my previous Crossfire experiment.
Canon 6D||[24-105/4L IS USM|100/2.8L Macro IS USM|70-300/4-5.6 IS USM|40/2.8 STM|50/1.4 USM|85/1.8 USM|Samyang/Bower 14/2.8 Full-Manual Rectilinear Wide-angle|
Canon EOS-M|11-22/4-5.6 IS STM|22/2 STM|EF-M 18-55/3.5-5.6 IS STM|
For sale!|24/2.8 IS USM
|
Airmantharp
Maximum Gerbil
 
Posts: 4962
Joined: Fri Oct 15, 2004 10:41 pm

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 4:10 pm

CaptTomato wrote:
cphite wrote:
If you aren't noticing, then there isn't really a problem


But does anyone notice, even if the frametimes are high...?

Some people may claim that they do... Some also claim that they notice teh difference between a $5 HDMI cable and a $5000 cable of same length and same specification compliance, if you know what I mean :wink:
My subscription allows you people to exist on this site and makes me a better human being than you'll ever be
JohnC
Gerbil Jedi
Gold subscriber
 
 
Posts: 1862
Joined: Fri Jan 28, 2011 2:08 pm
Location: NY/NJ/FL

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 4:11 pm

CaptTomato wrote:
cphite wrote:
If you aren't noticing, then there isn't really a problem


But does anyone notice, even if the frametimes are high...?
I remember reading that people move from crossfire to a single Nvidia and whilst FPS drop, gameplay is improved, and that seems believable as microstutter appears to be a CPU limitation of trying to run 2 GPU's, but it would also seem reasonable that for example, moving from 7850 XF to a single 7970 would be a smoother experience.

Yes.

Quick hypothetical here:

You have two GPUs. Both have similar processing power yet one of them has half as much texture memory. One constantly bumps up on that limit and induces occasional frames that take longer to render because textures must be swapped out in memory. Both get very similar FPS but one is clearly less ideal.

Now, take that and make one of the GPUs twice as fast...but leave the memory limitation in place. It has higher FPS but delivers a much less smooth experience overall.

That's where we stand. If you don't notice it, then you either don't have the issue or you just don't notice it. That doesn't mean it doesn't exist nor does it mean that other people aren't incredibly bothered by it.

It's no conspiracy.
Z68XP-UD4 | 2700K @ 4.4 GHz | 16 GB | 770 | PCP&C Silencer 950 | XSPC RX360 | Heatkiller R3 | D5 + RP-452X2 | HAF 932 | 1 TB WD Black w/ SRT
Waco
Gerbil Elite
 
Posts: 743
Joined: Tue Jan 20, 2009 4:14 pm

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 4:22 pm

cphite wrote:
CaptTomato wrote:Now that I've got a 7950, up from 6850 and 4850, my games are butter smooth, so how do I tell that I'm suffering from frame time problems.
With my older system and GPU's, it was easy to induce lag and noticeable stutter, but Crysis 1 and Metro2033 1920x1200 maxed are supa smooth.


That's kind of like asking "How do I know if my leg hurts?"


This topic reminded me of the old philosophical debate "If a tree falls in the forest and no one is around to hear it, does it make a sound?" :P

Waco wrote: Are you really accusing TR (along with a slew of other tech sites) of doctoring their videos of the stuttering issues seen with many common games/GPUs?


I'm not accusing TR of anything. It's their job to investigate these issues and they do it well. However, what are all of these videos? The only video I remember on TR was that one cherry picked from that particularly problematic open world Skyrim level (purportedly known to have issues when v-sync is not enabled). I don't think TR was trying to make AMD look bad, just that they were trying to show an obvious example of what latency problems actually looked like. But in the end, AMD ended up getting a bit of a raw deal with all the backlash. Hopefully it will prompt them to be more vigilant with their drivers and this will all be water under the bridge soon enough (water under a bridge in a forest that doesn't make a sound because no one is around to hear it, of course).
cynan
Gerbil Elite
Gold subscriber
 
 
Posts: 804
Joined: Thu Feb 05, 2004 2:30 pm

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 4:27 pm

cynan wrote:I'm not accusing TR of anything.

I never said you were, unless you have multiple accounts on here. :P

Besides - it's not just AMD GPUs that have these issues occasionally. The negative impression on AMD may have been an unintentional side effect of that particular video you mention though.
Z68XP-UD4 | 2700K @ 4.4 GHz | 16 GB | 770 | PCP&C Silencer 950 | XSPC RX360 | Heatkiller R3 | D5 + RP-452X2 | HAF 932 | 1 TB WD Black w/ SRT
Waco
Gerbil Elite
 
Posts: 743
Joined: Tue Jan 20, 2009 4:14 pm

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 4:44 pm

CaptTomato wrote:
cphite wrote:
If you aren't noticing, then there isn't really a problem


But does anyone notice, even if the frametimes are high...?
I remember reading that people move from crossfire to a single Nvidia and whilst FPS drop, gameplay is improved, and that seems believable as microstutter appears to be a CPU limitation of trying to run 2 GPU's, but it would also seem reasonable that for example, moving from 7850 XF to a single 7970 would be a smoother experience.

Yes, people notice. Some will be affected more than others.

The important thing to know is what actually happens at the screen when visual anomolies pop up. PC Perspective has some videos on this. If you have VSync off, you can get tearing, which is the result of having more than one frame displayed during a monitor refresh cycle.

Stuttering, on the other hand, can happen regardless of VSync. Stuttering looks like jumps in animation, which can be made worse if the animation skips the display of several frames when it recovers. With VSync off, it can make a tear worse.

And cynan, the videos Scott posted weren't supposed to be comprehensive, and weren't really meant to be harsh against AMD. In the wake of the "revisted" review, TR got boatloads of hate in the comments about how "biased they were toward Nvidia," or how the change to Windows 8 could have caused the issue, or this or that or the engle of the sun...yeah...

All the hate spinning around that ordeal was about people not knowing how to interperet data and looking for fights when only data was presented.
Being an adult doesn't mean you have to know what you're doing. It just means you have to look like you know what you're doing.
superjawes
Graphmaster Gerbil
Gold subscriber
 
 
Posts: 1027
Joined: Thu May 28, 2009 9:49 am

Re: How do you notice frame "time" issues?

Postposted on Tue Mar 19, 2013 11:27 pm

Waco wrote:Are you really accusing TR (along with a slew of other tech sites) of doctoring their videos of the stuttering issues seen with many common games/GPUs?
it's not an accusation, TR discussed quite openly that in order to effectively show the problem they had to record it using high speed video and then reduce the playback speed.....
to highlight exactly how open TechReport was they named the article wrote: "GeForce versus Radeon captured on high-speed video

Skyrim at 120 and 240 FPS"
the title of the article for heavens sake.... did you miss it?... no offense but next time before you try question, attack, dismiss or counter anything I say spend a moment using google to get the barest of facts in place.

to be clear I don't believe what TR did was wrong, they did gamers a service but as I mentioned inevitably what comes with even the best of intentions is..... the silly, the issue got hugely overblown, Waco you apparently didn't know the video was "doctored" given you tried to call my comments out on it, this despite TechReport telling you in the title of the article along with in the article that they were "doctoring" the video and now we've also got a new Radeon owner trying to figure out how to find out if he's got the issue.
neg
clone
Gerbil Elite
 
Posts: 900
Joined: Mon Mar 12, 2007 10:40 am

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 2:18 am

Frametime isn't the same as FPS, which is misleading if you look at the conversion table. It's the amount of time between when each frame is rendered. The only way to measure this is with fraps, it can spitout frametimes. Other then that there is no on screen indicator of it. FPS is a average and not a absolute. Objectively this is the only way to measure it right now...

Some games have in game frame time readouts, I've only seen one with this and that was NS2.

Subjectively it's pretty much what you pointed out. Jerkiness or stutters (for the real bad ones) would be a indicator of it.

If you haven't already, I'd suggest turning up the powertune option in the AMD control panel.
Bensam123
Gerbil Elite
 
Posts: 949
Joined: Wed May 29, 2002 12:19 pm

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 3:13 am

clone wrote:TR discussed quite openly that in order to effectively show the problem they had to record it using high speed video and then reduce the playback speed....................................................more ellipses please..............

Of course they had to. Most online video is limited to ridiculously bad framerates (by gamers' standards). In order to showcase an issue that's normally only "felt" by players, they needed a slow-motion recording to plainly reveal what we feel.


CaptTomato wrote:my games are butter smooth

If you can't feel a problem, then there is none.
Meadows
Grand Gerbil Poohbah
Silver subscriber
 
 
Posts: 3151
Joined: Mon Oct 08, 2007 1:10 pm
Location: Location: Location

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 5:41 am

clone wrote:the title of the article for heavens sake.... did you miss it?... no offense but next time before you try question, attack, dismiss or counter anything I say spend a moment using google to get the barest of facts in place.


???

So, it's a fact that slowing down a video is "doctoring it?" :o :o :o :o

So when Techreport zooms in on a screenshot to show jaggies in anti-aliasing comparisons, that's "doctoring" too?

Or, heck, I guess when prosecutors slow down surveillance video in courtrooms, that's "doctoring" evidence too? Funny how judges don't seem to agree... :roll: :roll: :roll:

Plus, you know, there is what Meadows said too.

But, fine, attack Waco and Techreport and probably me. :-?
Glorious
Darth Gerbil
Gold subscriber
 
 
Posts: 7836
Joined: Tue Aug 27, 2002 6:35 pm

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 7:46 am

The point I'm making is that I was under the impression that frametime results indicated that AMD had issues/more issues than Nvidia, however, in my experience, every time the framerates exceed 50FPS, it's a extremely smooth experience, IOW, high frames/second still seem to be the goal.

I'm sure you can appreciate that if I believed AMD had a problem I wouldn't have forked out $330 on my Saphire Vapour X 7950, ie, my experience with my 4850 and 6850 suggested that frames/second are key to smooth gameplay.

Bare in mind I'm only speaking from a single GPU perspective, and that I've always been under the impression that both SLI and XF produced microstutter.
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 7:50 am

Meadows wrote:If you can't feel a problem, then there is none.


But NO-ONE can deny that games stutter below 24fps, it's as plain as day, yet it seems that regardless of what the frametime tests indicate, my AMD's are smooth as long as framerates are high{50fps+}

Is it the case that frametimes are an objective measure that have subjective impacts, or no impact other than the data read out....?
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 8:05 am

Frame times and frame rates are very highly correlated. If you have a card with frame time problems that is able to also give very high frame rates on a given game, you won't notice any frame time problems (unless they happen to be extreme for very few frames.)

The value of frame times is that they are more precise (you can get frame rate from frame times by looking at the average frame time) and they allow you to see if there are specific hard to render sections of the game, sections that might create problems if your card isn't capable of overwhelming it.

The graphs of frames per second over a period of time that you sometimes see are essentially rudimentary frame time graphs, they just aren't being precise in what period they are measuring over.
peartart
Gerbil
 
Posts: 41
Joined: Wed Mar 21, 2012 3:01 pm

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 8:40 am

For the most part you aren't going to notice anything, because it doesn't exist. Especially now.

TR found a BUG in the game/driver that only appears under certain instances, and the games that they tested have now been fixed. AMD isn't as fast as Nvidia for game specific optimization, so it's quite easy to rig results by benchmarking just released or low priority games. If you aren't playing new releases day one, you'll never notice any issues.

It's basically the equivalent of benchmarking Tomb Raider BEFORE patch, and claiming there was a frame time issue. The issue with Tomb Raider was that Nvidia requires game specific tuning for every game, because they removed the hardware scheduling logic in Kepler, and Nvidia didn't have early access to the game to compile an optimized driver. Boo Hoo.

That's all these issues are. It's merely the side effect of not having specific optimizations for your game, and it disappears as the driver is updated. IMO, testing for this is "voodoo" benchmarking because you're not getting valid long term statistics. You're losing the forest for the trees, and skewing the metric towards who optimizes the driver better instead of who has the better hardware. AMD clearly has better hardware in a 7950 vs 660 Ti comparison, but Nvidia optimizes their driver more aggressively. Therefore, frame time statistics are always going to be biased toward Nvidia in new games.

Don't let anyone brainwash you into thinking frame times are more important than frame rates. Frame rates show what the hardware's capable of, frame times show driver optimizations.
l33t-g4m3r
Gerbil Jedi
Silver subscriber
 
 
Posts: 1972
Joined: Mon Dec 29, 2003 2:54 am

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 9:19 am

l33t-g4m3r wrote:Don't let anyone brainwash you into thinking frame times are more important than frame rates. Frame rates show what the hardware's capable of, frame times show driver optimizations.


Based on my experience over the last 5 yrs with AMD GPU, I fully agree.
Excellent post as well.
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 10:44 am

I wouldn't stress too much over frame times. When I read the TR GPU reviews, I pay most attention to the "Time Spent Beyond ___ms" graphs. Those graphs are a sum of all the "high latency frames" (stutters) in the test sequence. As you can see in all of TRs GPU reviews, the current-gen cards are showing MUCH less "stuttering" than previous-gen cards. (2x reduction in high latency frames here for example)

The point of minimizing frame latecy spikes is to produce a "smooth" gameplay no matter what your average framerates are. A card that cranks out a very consistent 35fps w/o latency spikes could make gameplay feel equally enjoyable or better than a card that's putting out 50fps but with many latency spikes.

Still, the main point, like you said, is that stuttering is subjective and TR is trying to illustrate that behavior with objective measurements. It's a very handy tool for the potential buyer, not so much for the owner.
i5-3570K, ASRock Z77 Pro4-m, Asus GTX660 TOP, 120 GB Vertex 3 Max IOPS, 2 TB Samsung EcoGreen F4, 8GB G-Skill @1.25V, Silverstone PS07B
DPete27
Gerbil Jedi
Silver subscriber
 
 
Posts: 1618
Joined: Wed Jan 26, 2011 12:50 pm
Location: Madison, Wisconsin

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 11:16 am

Just wanted to drop in and offer some thoughts.

First, I agree that if you aren't noticing any performance problems, you aren't having any frame time issues of note. What matters at the end of the day is how smooth you perceive the animation to be.

With that said--and this is well-trod ground by now--measuring frame latencies is the best way to detect noteworthy slowdowns. FPS averages can mask slowdowns into the multiple hundreds of milliseconds, and you WILL notice those problems when playing. The folks in this thread saying "frame rates are just as important" are simply wrong, for reasons we have explained time and again.

Next, the folks who say high frame latencies are always simply the result of driver issues are incorrect. GPU performance matters tremendously to frame rendering times. Slower GPUs tend to produce higher frame latencies and larger latency spikes than faster GPUs, given the same CPU, driver, and other components.

Also, different GPU architectures can produce different results--for instance, Fermi-based GPUs have had lots of latency spikes in BF3 for a long time, but Kepler GPUs do not.

Yes, drivers issues can contribute to high frame rendering times, and optimizations can help to reduce them, but that is only one part of a larger story. Game frame output is a very complex system involving a vast swath of hardware components--CPU, system memory, disk, PCIe transfers, GPU memory, etc.--and software components--the Windows Kernel, the game engine, DirectX, GPU driver compilers, and more. Many of these component parts, say DirectX HLSL compilation or internal GPU kernel execution, are incredibly complex in themselves. A delay in any one place can slow frame production.

Furthermore, GPUs for gaming are very much software-and-hardware combo products. Yes, drivers change, but it would be naive and mistaken to assume that any GPU with lower performance properly understood--that is, higher frame times and latency spikes--will magically catch or become superior to the competition with driver updates. The product that works best now must be assumed to be superior until proven otherwise.

We'll have more to say about all of these things in an article soon that offers more insights, but the OP would do well to ignore some of the mistaken perceptions being repeated above.
Scott Wasson - "Damage"
Editor - The Tech Report
Damage
TR Staff
Gold subscriber
 
 
Posts: 1671
Joined: Wed Dec 26, 2001 7:00 pm
Location: Lee's Summit, Missouri, USA

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 11:39 am

Meadows wrote:Of course they had to. Most online video is limited to ridiculously bad framerates (by gamers' standards). In order to showcase an issue that's normally only "felt" by players, they needed a slow-motion recording to plainly reveal what we feel.
never said otherwise.
Glorious wrote:So, it's a fact that slowing down a video is "doctoring it?
But, fine, attack Waco and Techreport and probably me.
"doctoring" is Waco's accusation not mine, if you spent 8 seconds looking back you'd notice I said altered video and more to the point I deliberately put "doctored" in highlights.

2nd glorious I'm not "attacking" anyone and more to the point you don't matter to me, literally, so no worries of that happening unless you feel a need to continue to insert yourself you'll quickly be forgotten.
Damage wrote:Just wanted to drop in and offer some thoughts.
while I'm not the op appreciate the response and look forward to the article.
Last edited by clone on Wed Mar 20, 2013 11:46 am, edited 2 times in total.
neg
clone
Gerbil Elite
 
Posts: 900
Joined: Mon Mar 12, 2007 10:40 am

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 11:42 am

Damage wrote: but the OP would do well to ignore some of the mistaken perceptions being repeated above.


I think my problem is that I was under the impression that even with high framerates, I "might" notice frametime issues as that's the impression TR articles gave me, however, {at least so far....more games will be tested over the coming weeks/months}, not one game with high framerates feels anything other than especially smooth.

Like anyone, I've experienced stutter+lag on many occasions, but only as a consequence of low framerates, for example with my old e8400/4gig ram and 6850, Crysis1 was borderline playable at 2xaa, but shockingly bad at 8xaa, but with my new system, 8xaa is extremely smooth, so smooth that I find it impossible to believe that there's any meaningful frametime problems, which there probably isn't, but if say Crysis3 stutters on higher settings, then I'd have to lower the graphical settings to achieve approx 60fps average, and if gameplay is smooth, then it would almost disqualify frametimes spikes being anything but related to framerates.

OTOH, if noticeable stutter is always a factor no matter the graphical settings, then it's fair to suggest that in this instance, AMD+Crysis 3 are a poor match, but if driver updates eliminate frametimes issues, then it would be reasonable to assume that poor SW optimization is the culprit.

This hardly means frametime evaluation is worthless, but I'm mainly curious about whether AMD has inherent HW deficits, and at this early stage, it appears that it doesn't in single card set ups that are component matched, ie, powerful CPU+ram etc.
CaptTomato
Gerbil First Class
 
Posts: 150
Joined: Mon Nov 22, 2010 6:12 am
Location: Australia

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 11:53 am

It's not like we haven't provided clear examples of high frame rates and poor playability.

Although almost everything struggles in this example, the 7970 GHz CF config here is a great case in point:

http://techreport.com/review/24381/nvid ... eviewed/12

58 FPS average, horrible experience.

We have a nice cache of GPU reviews showing similar issues with single-card configs. Long experience has taught me that big spikes in Fraps data tend to correspond with slowdowns while playing.

It's good you haven't had too many problems. Gamers do tend to tune their rigs for good performance in each game, after all, and we tune games to stress GPUs and bring out the differences between them, since that's our task. But FPS averages are almost meaningless in conveying the experience you'll get while playing.
Scott Wasson - "Damage"
Editor - The Tech Report
Damage
TR Staff
Gold subscriber
 
 
Posts: 1671
Joined: Wed Dec 26, 2001 7:00 pm
Location: Lee's Summit, Missouri, USA

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 12:38 pm

Damage wrote:Next, the folks who say high frame latencies are always simply the result of driver issues are incorrect.

I do agree with you, but only because you included the word always. ("only" would have been another good word choice) As your article pointed out, drivers certainly can play a big role in high frame latencies.
i5-3570K, ASRock Z77 Pro4-m, Asus GTX660 TOP, 120 GB Vertex 3 Max IOPS, 2 TB Samsung EcoGreen F4, 8GB G-Skill @1.25V, Silverstone PS07B
DPete27
Gerbil Jedi
Silver subscriber
 
 
Posts: 1618
Joined: Wed Jan 26, 2011 12:50 pm
Location: Madison, Wisconsin

Re: How do you notice frame "time" issues?

Postposted on Wed Mar 20, 2013 12:39 pm

clone wrote:"doctoring" is Waco's accusation not mine, if you spent 8 seconds looking back you'd notice I said altered video and more to the point I deliberately put "doctored" in highlights.


I'm sorry, are you relying on the completely non-existent semantic difference between "altered" and "doctored" :o

You yourself put "doctored" in highlights when you accused Waco of not knowing the Techreport did it! :roll:

Here, let me quote you:

clone wrote:Waco you apparently didn't know the video was "doctored" given you tried to call my comments out on it


My point is just that what you were saying is not only misleading, but inflammatory. Waco was saying the same thing, and you acted completely obnoxiously. If you don't want to be understood as falsely impugning our gracious hosts you ought to be a lot less reckless with the words you recklessly sling around and little more circumspect with people who honestly inquire if you really mean what you say. :roll:
Glorious
Darth Gerbil
Gold subscriber
 
 
Posts: 7836
Joined: Tue Aug 27, 2002 6:35 pm

Next

Return to Graphics

Who is online

Users browsing this forum: No registered users and 8 guests