[pcper] frame metering review 690 vs. 7970 CF vs. Titan

[Silverforce11]

Anything higher than 60fps is irrelevant? So all the benchmarks from all the years past never meant anything in terms of which cards were truly faster?
I NEVER use Vsync and rarely do I see any tearing. Maybe it's because I don't use and AMD card. In fact, I can't remember any instances where I see any tearing.

WTF is this crap?

You have a 60 hz monitor and dont see screen tear without vsync? What monitor is it, some kind of magical technology??

Otherwise stop talking out of ur ass, screen tearing is not like microstutter, everyone notices it.

 

[Rvenger]

WTF is this crap?

You have a 60 hz monitor and dont see screen tear without vsync? What monitor is it, some kind of magical technology??

Otherwise stop talking out of ur ass, screen tearing is not like microstutter, everyone notices it.

I get some nice tearing on my Titan when Vsync is disabled, just like my HD 7970 so I don't know what he is talking about.

 

[omeds]

What game can you run 250 FPS? COD4? LOL who plays that any more. CS:GO? most boring game ever.

BF3 will never hit 250fps because its CPU bound. Far Cry 3 wont hit 120 FPS without turning off MSAA. Crysis 3.... Lol good luck hitting 60fps consistant.

I already told you I play BF3 at a solid 125fps cap. Ultra settings, 4xMSAA + 4xtrSSAA with Mesh quality at medium. At both 1080p 120hz or 1440p 60hz.

So,
125 for BF3
250 for CoD series
125 for RO2
125 for MoH
250 for TF2
250 for CSS
125 for BC2

Those are the only games I play online, and I build my system to run those games at those frame rates.

And I get a solid 59 capped or 60fps vsync in Crysis3 with 4xMSAA btw.

Are you playing at 1280x768 just to get frame rates into the hundreds? On what games.

No, either 120hz 1080p or 60hz 1440p. Soon I will get a 1440p 120hz display so I can finally kick crappy 60hz to the curb.

1200p screens will never hit 120fps. 1440p wont either and 1600p LOL good luck with that.

There are 1440p displays that can be oc'd to 120hz and are in high demand, aswell as standard 1080p 120hz displays as you know.

You CANNOT see input lag on the above setup. I have tested for the last 4 hours.

I and many others can feel input lag at 59fps + vsync. Sure its better, but its still not ideal. 125 and 250 are feel more responsive, even than 59 fps without vsync. Thats the whole purpose of lucid virtu MVP, to increase mouse responsiveness to levels that are felt with higher fps.

I like to use 59+vsync in single player games, so I can crank up IQ such as SSAA & AO where fluid motion and responsiveness dont matter so much.


Dude you're wrong, stop emabarrasing yourself.

 

[Fx1]

I get some nice tearing on my Titan when Vsync is disabled, just like my HD 7970 so I don't know what he is talking about.

Grass is green , sky is blue, water is wet, V-Sync off means screen tearing.

 

[omeds]

WTF is this crap?

You have a 60 hz monitor and dont see screen tear without vsync? What monitor is it, some kind of magical technology??

Otherwise stop talking out of ur ass, screen tearing is not like microstutter, everyone notices it.

I get some nice tearing on my Titan when Vsync is disabled, just like my HD 7970 so I don't know what he is talking about.

Use consistent caps. All is explained by this guy who started the lightboost threads, he knows what hes talking about.

http://forums.anandtech.com/showpost.php?p=34693183&postcount=3

 

[railven]

Welps, not sure what is going on here to be frank, but got a few questions since I just re-joined the Multi-GPU crew:

#1: Is frame limiting the same as v-sync? As far as I know, it isn't but Omeds showed an image of him applying a 58 FPS Frame limit and kept calling it v-sync, so not sure if he was also applying v-sync with the limit or he was calling the limit v-sync. Either way, I can assure you guys that a limit + v-sync does not alleviate stuttering in WoW with a GTX 680, HD 7970 or CFX 7970. The stuttering is still there, of course more noticeable on the Radeon setups, but still there.

#2: I can vouch that a GTX 680 shows tons of tearing in various games without v-sync. I don't even understand how any one can claim it doesn't, unless that person isn't sensitive to tearing - which if they are, kudos! Keep doing what you aren't doing because once you see it, you'll never unsee it.

#3: Radeon Pro and this dynamic V-sync thing, so I got my Multi-GPU setup, I'm getting stupid high FPS in WoW, then it tanks <60 FPS in specific parts (must be bad coding or just stacked effects I just can't see, maybe an unseeable tessellated ocean, whatever) but my FPS doesn't lock in the 30s. I have tripple buffering on (frankly, don't really notice the input lag, notice more the frame drop lag though). What does Radeon Pro offer? I disable tripple buffering and turn on dynamic v-sync? I assume I will get tearing out the ying-yang below 60 FPS like Adaptive V-sync does for our GTX 680? Doesn't sound like a good trade off to me. I'd rather use full v-sync with tripple buffering and never experience tearing than tear-free above 60 FPS, because unfortunately WoW loves to just drop frames (yes, on both setups).

EDIT: #4: what's with this recommendation to buy a weaker GPU because you use v-sync? So I play game A, with an GTX 680 or HD 7970, get 60 FPS with v-sync and 40-50% GPU utilization, I can do the same with an HD 7870/GTX 660 Ti but get >90% GPU utilization. If I stop there, sure that argument holds water, but usually what I do (and I'm sure most people who buy higher tier hardware) is I start to apply more AA or use texture paks. Now I'm running the same game with 2x/4xSSAA and HQ-AF 16x getting >80% GPU utilization on my GTX680/HD7970 setup while the other GTX660Ti/HD7870 is already maxed out. At least that is how I think of an answer to that question.

Must read more, carry on guys!

 

[omeds]

#1: No I never said a frame cap is vsync. The idea is to use them both together, a 58/59 fps cap + vsync at the same time.

#2: Yes there is tearing without vsync. The appearance of this tearing can be greatly reduced by using frame caps of 125, 250, 333, 500 and so on. The higher the frame cap, the more effective it will be.

#3: Yes dynamic vsync simply disables vsync when your fps are below your refresh rate, like adaptive vsync. I dont like either option tbh because of the very reason you pointed out.

 

[Keysplayr]

WTF is this crap?

You have a 60 hz monitor and dont see screen tear without vsync? What monitor is it, some kind of magical technology??

Otherwise stop talking out of ur ass, screen tearing is not like microstutter, everyone notices it.

Heh. Watch it son. Get those emotions under control.

 

[felang]

Just have one question. What´s the difference between caping at 59fps + Vsync and just plain 60fps Vsync? I´ve read through the whole thread and that is not clear to me.

 

[omeds]

Just have one question. What´s the difference between caping at 59fps + Vsync and just plain 60fps Vsync? I´ve read through the whole thread and that is not clear to me.

It helps reduce input lag when using vsync. It also helps with microstutter that may still be present with vsync in troublesome games.

 

[AdamK47]

#1: No I never said a frame cap is vsync. The idea is to use them both together, a 58/59 fps cap + vsync at the same time.

#2: Yes there is tearing without vsync. The appearance of this tearing can be greatly reduced by using frame caps of 125, 250, 333, 500 and so on. The higher the frame cap, the more effective it will be.

#3: Yes dynamic vsync simply disables vsync when your fps are below your refresh rate, like adaptive vsync. I dont like either option tbh because of the very reason you pointed out.

I'm calling BS on point #2. What you see is purely placebo. A frame cap does not and will not decide how a frame goes into the frame buffer. It will still have that stitched together multiframe mess as it enteres. In fact, capping the framerate higher amplifies the problem. Instead of two frames stitched together, you get multiple as it is haphazardly thrown to the buffer.

 

[felang]

It helps reduce input lag when using vsync. It also helps with microstutter that may still be present with vsync in troublesome games.

Ah ok, so in other words mostly for multi GPU where you may have unacceptable levels of microstutter?

In BF3 for example I usually just cap at 62 fps (no vsync), no noticeable tearing to me (1440p) with a single GTX 680, no MSAA or FXAA, I use SweetFX SMAA injector.

 

[Fx1]

Ah ok, so in other words mostly for multi GPU where you may have unacceptable levels of microstutter?

In BF3 for example I usually just cap at 62 fps (no vsync), no noticeable tearing to me (1440p) with a single GTX 680, no MSAA or FXAA, I use SweetFX SMAA injector.

No its not just for CF.

I have 1 7970 and i tested it today.

i dont fully understand why but its something to do with 59.94hz refresh and windows setting as 59hz. 60 FPS Vsync has this weird input lag and lack of response. Soon as you cap it to 59 FPs you get no input lag. The input lag should be very small and no where near as bad as it is with a 60fps vsync.

 

[Fx1]

#1: No I never said a frame cap is vsync. The idea is to use them both together, a 58/59 fps cap + vsync at the same time.

#2: Yes there is tearing without vsync. The appearance of this tearing can be greatly reduced by using frame caps of 125, 250, 333, 500 and so on. The higher the frame cap, the more effective it will be.

#3: Yes dynamic vsync simply disables vsync when your fps are below your refresh rate, like adaptive vsync. I dont like either option tbh because of the very reason you pointed out.

no no No No NOOOOOOOOOOOO!!!!!

125,250,333 and 500 are ALL Quake 3 engine exploits and HAVE NOTHING TO DO WITH VISUALS!

This is classic gamer myth propagated by people exactly like you. These frame caps have nothing to do with any other games that dont use the quake 3 engine.

 

[blackened23]

#1: No I never said a frame cap is vsync. The idea is to use them both together, a 58/59 fps cap + vsync at the same time.

#2: Yes there is tearing without vsync. The appearance of this tearing can be greatly reduced by using frame caps of 125, 250, 333, 500 and so on. The higher the frame cap, the more effective it will be.

#3: Yes dynamic vsync simply disables vsync when your fps are below your refresh rate, like adaptive vsync. I dont like either option tbh because of the very reason you pointed out.

Point #2:

With all due respect, gimme a break. Framecaps do nothing to prevent tearing. If you don't use vsync on a LCD you will get tearing absolutely guaranteed. That breaks immersion far more than anything, and will make a game unplayable.

Even 120hz monitors have tearing at times without vsync have slight tearing, although is far less than what appears on typical 60hz LCDs.

 

[railven]

#1: No I never said a frame cap is vsync. The idea is to use them both together, a 58/59 fps cap + vsync at the same time.

#2: Yes there is tearing without vsync. The appearance of this tearing can be greatly reduced by using frame caps of 125, 250, 333, 500 and so on. The higher the frame cap, the more effective it will be.

#3: Yes dynamic vsync simply disables vsync when your fps are below your refresh rate, like adaptive vsync. I dont like either option tbh because of the very reason you pointed out.

Ahh, thanks for the clarification. I guess it just got lost in the bickering.

Thankfully I don't play online/competively so the input lag doesn't bother me much, but I will test this idea as I'm sure a few times I died in WoW because I didn't press a cool down in time haha.

 

[omeds]

no no No No NOOOOOOOOOOOO!!!!!

125,250,333 and 500 are ALL Quake 3 engine exploits and HAVE NOTHING TO DO WITH VISUALS!

This is classic gamer myth propagated by people exactly like you. These frame caps have nothing to do with any other games that dont use the quake 3 engine.

Wrong. Read this again:

http://forums.anandtech.com/showpost.php?p=34693183&postcount=3

Point #2:

With all due respect, gimme a break. Framecaps do nothing to prevent tearing. If you don't use vsync on a LCD you will get tearing absolutely guaranteed. That breaks immersion far more than anything, and will make a game unplayable.

Even 120hz monitors have tearing at times without vsync have slight tearing, although is far less than what appears on typical 60hz LCDs.

Higher fps with caps do help reduce the appearance of tearing. Higher fps reduce the offset of the next frame. I doubt you would notice a pixel or two offset while gaming, especially in a fast paced FPS for example.

 

[omeds]

Ah ok, so in other words mostly for multi GPU where you may have unacceptable levels of microstutter?

In BF3 for example I usually just cap at 62 fps (no vsync), no noticeable tearing to me (1440p) with a single GTX 680, no MSAA or FXAA, I use SweetFX SMAA injector.

62 is a good cap to use if you can't maintain 125, I'd just continue to use that if youre happy with it tbh as it doesn't have the drawbacks of vsync.

 

[omeds]

I'm calling BS on point #2. What you see is purely placebo. A frame cap does not and will not decide how a frame goes into the frame buffer. It will still have that stitched together multiframe mess as it enteres. In fact, capping the framerate higher amplifies the problem. Instead of two frames stitched together, you get multiple as it is haphazardly thrown to the buffer.

Read this: http://forums.anandtech.com/showpost.php?p=34693183&postcount=3

Its not new, gamers have been doing it for ~15 years. Its not haphazardly thrown into the buffer if you use a cap, which helps them to be delivered evenly. The higher the frame rate, the smaller the offset between the frames along the tear lines.

 

[felang]

No its not just for CF.

I have 1 7970 and i tested it today.

i dont fully understand why but its something to do with 59.94hz refresh and windows setting as 59hz. 60 FPS Vsync has this weird input lag and lack of response. Soon as you cap it to 59 FPs you get no input lag. The input lag should be very small and no where near as bad as it is with a 60fps vsync.

Wouldn't caping at 59 hz and setting vsync at 60 Hz cause framerate to lock at 45 or 30 fps?

 

[Fx1]

Read this: http://forums.anandtech.com/showpost.php?p=34693183&postcount=3

Its not new, gamers have been doing it for ~15 years. Its not haphazardly thrown into the buffer if you use a cap, which helps them to be delivered evenly. The higher the frame rate, the smaller the offset between the frames along the tear lines.

For a start your game wont run at 250 fps or 500 fps. Not unless your gaming on old games. FACT

You missing the whole point. IF my car had wings it could fly!! If you could run BF3 at 500 fps you wouldnt see the tearing? lol.

Fact is that with Vsync at 59 fps you get all the benefits and none of the tearing. So Forget pie in the sky 250 fps.

 

[omeds]

For a start your game wont run at 250 fps or 500 fps. Not unless your gaming on old games. FACT

You missing the whole point. IF my car had wings it could fly!! If you could run BF3 at 500 fps you wouldnt see the tearing? lol.

Fact is that with Vsync at 59 fps you get all the benefits and none of the tearing. So Forget pie in the sky 250 fps.

Thats right, so the next best option is 125. Failing that many people use 62. You use the highest cap you can maintain, obviously the higher the better, from 62 (UE3 games use this), to 125, some people like 200 (BF3 also has a cap at 200 btw), 250, 317 or 333, 500, or simply uncap it if you have uber high fps.

No the fact is 59+vsync has drawbacks that many wish to avoid. Its fine for single player or slower paced games though imo.

 

[Fx1]

Thats right, so the next best option is 125. Failing that many people use 62. You use the highest cap you can maintain, from 62 (UE3 games use this), to 125, some people like 200 (BF3 also has a cap at 200 btw), 250, 317 or 333, 500, or simply uncap it if you have uber high fps.

No the fact is 59+vsync has drawbacks that many wish to avoid. Its fine for single player or slower paced games though imo.

What drawbacks?

 

[railven]

For a start your game wont run at 250 fps or 500 fps. Not unless your gaming on old games. FACT

You missing the whole point. IF my car had wings it could fly!! If you could run BF3 at 500 fps you wouldnt see the tearing? lol.

Fact is that with Vsync at 59 fps you get all the benefits and none of the tearing. So Forget pie in the sky 250 fps.

I think you are missing his point. He isn't saying you WONT get tearing, he's saying due to the frame interval being less the shift in the frame would be smaller, thus the tear not as visible, if at all to you.

I, however, would have to test it myself to see just how much of the tear is visible, but say a game renders 100 FPS with unrestricted options, limiting to 60 vs limiting to 125, when it tears, the tears in 60 would be far more noticeable than the tear when limited to 125.



So, I'm not sure what option I should try to reduce the microstutter in WoW as much as possible while not tearing to all hell. I guess I'll try the 125 option for Dynamic V-sync and see what happens.

 

[Fx1]

I think you are missing his point. He isn't saying you WONT get tearing, he's saying due to the frame interval being less the shift in the frame would be smaller, thus the tear not as visible, if at all to you.

I, however, would have to test it myself to see just how much of the tear is visible, but say a game renders 100 FPS with unrestricted options, limiting to 60 vs limiting to 125, when it tears, the tears in 60 would be far more noticeable than the tear when limited to 125.



So, I'm not sure what option I should try to reduce the microstutter in WoW as much as possible while not tearing to all hell. I guess I'll try the 125 option for Dynamic V-sync and see what happens.

You will always notice the tears. The are mostly visible on edges of walls and straight lines. People will debate all day about the benefit of FXAA and MSAA whilst living with screen tearing.