News Nvidia Explains Why it Thinks High Frame Rates Matter in Competitive Games

Yeah, I'll buy into this when they can prove that people can, first of all, SEE things quickly enough that 100fps vs 240fps or more will make any sort of difference, and second, that they can physically respond that quickly.

They're trying to sell more powerful video cards, I get it - but this seems, to me, to be feeding into a line of BS about the "need' for faster frame rates that can't be backed.

240fps = 0.00417 seconds per frame.
 
Pumping unnecessarily high frames generates a lot of heat and wastes power. It's the computing equivalent of 1960's muscle cars... an illusion of superior performance wasting irreplaceable energy and filling the air with CO2 while bolstering fragile egos.

Thank you, Nvidia, for such forward thinking.
 

Mottamort

Distinguished
Oct 23, 2008
201
0
18,680
0
Yeah, I'll buy into this when they can prove that people can, first of all, SEE things quickly enough that 100fps vs 240fps or more will make any sort of difference, and second, that they can physically respond that quickly.

They're trying to sell more powerful video cards, I get it - but this seems, to me, to be feeding into a line of BS about the "need' for faster frame rates that can't be backed.

240fps = 0.00417 seconds per frame.
Actually Linus Tech Tips did quite an in-depth video with different e-sport "pro" players with different setups using 60/144/240hz setups with different tests.
Results were pretty conclusive...higher fps gives better results across the board.

Next thing you'll say is that the eye can't see faster than 24fps :p
 
Actually Linus Tech Tips did quite an in-depth video with different e-sport "pro" players with different setups using 60/144/240hz setups with different tests.
Results were pretty conclusive...higher fps gives better results across the board.
Link? Because that sounds pretty questionable at best. So, if they put pro players in front of monitors with randomly assigned refresh rates, these players could tell when it was 60, 75, 120, 144, 165, and 240? I doubt it.

What was their methodology? Do they have a write-up? I'd rather read it than slog through a video.

I'll believe it when there's a repeatable, provable process.

Next thing you'll say is that the eye can't see faster than 24fps :p
You'd better take your crystal ball in to the shop - it's giving you bad info.


Pumping unnecessarily high frames generates a lot of heat and wastes power. It's the computing equivalent of 1960's muscle cars... an illusion of superior performance wasting irreplaceable energy and filling the air with CO2 while bolstering fragile egos.
To be fair, you most certainly could feel the difference with the ridiculous gobs of torque those things produced. Of course, the tires, brakes, and suspensions of the day weren't up to that kind of abuse.

Naturally, with the progress of technology, those old cars are laughably inefficient for fuel usage, as well as the amount of power they produce per amount of displacement, compared to today's standards.

There were some odd outliers, in terms of efficiency, though, strangely enough.
 
Last edited:

nikolajj

Honorable
Feb 27, 2013
122
2
10,685
0
This article attracted a lot of deniers.
More frames = more good. (diminishing return apply ofc.)

Passive media like TV and movies is different in that you do not have a user performance to be affected. Heigher FPS is still more true to life, but people are so used to 24 FPS that they are thrown off by new standards like HFR. "Looks too realistic" and "It is like being there for real" is thing that I for SOME reason have seen thrown around as negatives. Unbelievable!

The Linus Tech Tips video mentioned:
View: https://www.youtube.com/watch?v=OX31kZbAXsA
 
Reactions: TJ Hooker

Gurg

Distinguished
Mar 13, 2013
515
61
19,070
1
You don't interact with a TV in the same manner you interact with a monitor when gaming.
Both are 4K 3840x2160 @ 60Hz both computer and TV are connected to ethernet although a displayport 60Hz connection from PC to monitor, my PC has a six core processor and GPU capable of 60fps at 4K While my TV has a "Quad-core CPU/Multi-core GPU for responsive streaming – experience ultra-smooth streaming of 4K video at up to 60 fps, instant search results and fast and fluid responsiveness". Monitor is 28" vs 55" for TV although when factoring in distance from screen both have similar effective viewing size.

I admit I haven't tried gaming on my TV as I'm very often using both PC and TV at the same time. When there are competing ball games on I often fire up my laptop to carry a second game. Can't really watch all three for TV broadcasts simultaneously as ATT Now limits me to two devices at one time.

Just like "King V" I also would like to see reliable high level research evaluating gaming performance differences over 60fps by experienced high level gamers to see the point of diminishing returns.
 
Yeah, I'll buy into this when they can prove that people can, first of all, SEE things quickly enough that 100fps vs 240fps or more will make any sort of difference, and second, that they can physically respond that quickly.

They're trying to sell more powerful video cards, I get it - but this seems, to me, to be feeding into a line of BS about the "need' for faster frame rates that can't be backed.

240fps = 0.00417 seconds per frame.
This is such a difficult thing to prove. I liken it to audio. I grew up with studio grade equipment in my home and I know I am more sensitive to different equipment than most people. I have got people to try different setups and usually they will say they can’t tell the difference but I can. I’m not as sensitive as a professional musician but also I definitely notice a lot more than most people. I really would not be surprised if very experienced and skilled players can tell the difference that the vast majority cannot.
 
Reactions: Gurg and TJ Hooker
Dec 4, 2019
1
0
10
0
Yeah, I'll buy into this when they can prove that people can, first of all, SEE things quickly enough that 100fps vs 240fps or more will make any sort of difference, and second, that they can physically respond that quickly.

They're trying to sell more powerful video cards, I get it - but this seems, to me, to be feeding into a line of BS about the "need' for faster frame rates that can't be backed.

240fps = 0.00417 seconds per frame.
The graphics card will not, unless you use v-sync, pump out 60fps with even intervals, necessarily, but 60hz will always be 60hz. If you get two images with a different interval, one of them is lost. Losing an image at 144hz or 240hz isn't that big of a deal, but in FPS games, losing an image at 60hz could decide who dies and who lives. As for that, I do notice when my 144hz is set to 60hz. They say you don't notice it when you go from 60hz to 120hz+, but when you go back, you are in for trouble.

EDIT: Also, we did up our 60hz CRTs to 100hz back in 2003 as well.
EDIT2: Try enabling v-sync in an FPS game and see. You should feel the difference. If you don't, your reaction time is wack.
 

Gurg

Distinguished
Mar 13, 2013
515
61
19,070
1
If we accept this article as fact then no serious gamer should use less than a 9700k OC to 5.1ghz or a 9600k OC to 5.0ghz with a 2080ti. The OC 9700k has a 13.5% frame rate @ 1080 (144.2 vs 127.2 avg FPS) advantage over any OC AMD CPU while the OC 9600k has a 7.4% (136.4 avg FPS) advantage. And you better have a decent AIO water cooler equivalent to at least the Corsair H115i to cool the CPU, at least 16gbs of very fast DDR4 and an M.2 Samsung 970+ equivalent drive. Anyone playing with less is apparently handicapping themselves.
https://www.tomshardware.com/reviews/amd-ryzen-5-3600x-review,6245-11.html

"Clock speed is more important than core number: Higher CPU clock speeds translate to snappier performance in simple, common tasks such as gaming "
https://www.tomshardware.com/reviews/best-cpus,3986.html

The next question to be researched appears to be at what point is there a trade off between higher FPS, refresh rates, video detail (1080/1440/2160) and finally audio accuracy.

The good news is that there are lots of holiday sales on PC components going on now. Forget buying your spouse, kids or family any presents.
 

Born2Hula

Prominent
Sep 27, 2019
14
0
510
0
So, I did some quick maths and, if nvidia is to be believed when they imply the difference in framerates is multiplied 4 times over, then the improvement from 60fps to 240 fps will shave about 5 milliseconds off your total response time (of, very roughly, 200-300 milliseconds depending on age, setup etc), so that's not nothing.

It's worth pointing out that the difference is only 2 milliseconds if you're at 100fps to begin with, and going from 240 to 480 fps will only shave off another quarter millisecond. Also this all assumes that nVidia isnt overstating the difference (which they probably are to some extent), in which case the true value is only 1.25 milliseconds saved from going from 60 to 240.
 

nikolajj

Honorable
Feb 27, 2013
122
2
10,685
0
So, I did some quick maths and, if nvidia is to be believed when they imply the difference in framerates is multiplied 4 times over, then the improvement from 60fps to 240 fps will shave about 5 milliseconds off your total response time (of, very roughly, 200-300 milliseconds depending on age, setup etc), so that's not nothing.

It's worth pointing out that the difference is only 2 milliseconds if you're at 100fps to begin with, and going from 240 to 480 fps will only shave off another quarter millisecond. Also this all assumes that nVidia isnt overstating the difference (which they probably are to some extent), in which case the true value is only 1.25 milliseconds saved from going from 60 to 240.
Your math is off.
1 sec = 1000 ms
60 hz = 1000/60 = 16.6666
240 hz = 1000/240 = 4.1666

16.6666 - 4.1666 = 12.5 milliseconds saved from going from 60 to 240
16.6666 / 4.1666 = 4 times greater delay with 60 hz
 

Born2Hula

Prominent
Sep 27, 2019
14
0
510
0
Your math is off.
1 sec = 1000 ms
60 hz = 1000/60 = 16.6666
240 hz = 1000/240 = 4.1666

16.6666 - 4.1666 = 12.5 milliseconds saved from going from 60 to 240
16.6666 / 4.1666 = 4 times greater delay with 60 hz

Hmmm, you're right, I'm out by a factor of 10. This just makes me less inclined to believe nVidia however, the difference is just too large. At those sorts of values we wouldn't be arguing this - it would be immediately obvious.
 
King V and Gurg... did you event watch the video?
If you have access to a 120+ hz monitor, try this:
At work, so, no. Also, I'm really not at all willing to slog through a half hour of video. Is their analysis, in detail written up somewhere such that I (and others) can read it?

If not, then if they couldn't be bothered to detail out their process in writing, then why bother?

https://www.livescience.com/42666-human-brain-sees-images-record-speed.html

13 milliseconds . . . whereas 240Hz is a new frame every 4.17 milliseconds. So, it would take a human 3 frames to respond to that.

Sounds an awful lot like the record for human vision would thus be reacting to a speed at best of 240Hz x (4.17 ÷ 13) = 76.985 frames/second. That's less than my rough guesstimate of "maybe 100 fps tops."
 
Last edited:

Gurg

Distinguished
Mar 13, 2013
515
61
19,070
1

Born2Hula

Prominent
Sep 27, 2019
14
0
510
0
An article on PcGamer (no link because I'm at work) says the K/D ratio of pro gamers in their test improved 90(!!!!)% when they moved from 60 to 240. I'm pressing X to doubt on that. That's going from a noob to a pro gamer with the flick of a switch territory.
 

Born2Hula

Prominent
Sep 27, 2019
14
0
510
0
At work, so, no. Also, I'm really not at all willing to slog through a half hour of video. Is their analysis, in detail written up somewhere such that I (and others) can read it?

If not, then if they couldn't be bothered to detail out their process in writing, then why bother?

https://www.livescience.com/42666-human-brain-sees-images-record-speed.html

13 milliseconds . . . whereas 240Hz is a new frame every 4.17 milliseconds. So, it would take a human 3 frames to respond to that.

Sounds an awful lot like the record for human vision would thus be reacting to a speed at best of 240Hz x (4.17 ÷ 13) = 76.985 frames/second. That's less than my rough guesstimate of "maybe 100 fps tops."
There's two different ways of looking at it - taking the fastest thing you COULD react to, in which case I think you're right at 77fps, but also looking at your reaction time end-to-end (like I did) where there is potentially no limit to how fast something can be. I'll be honest, I came to this expecting to debunk the whole shebang but I'm starting to think they might be on to something... at least up to point. 240Hz is probably overkill, but maybe 60 isn't as good as a lot of us assume it to be.
 

DSzymborski

Polypheme
Moderator
There's two different ways of looking at it - taking the fastest thing you COULD react to, in which case I think you're right at 77fps, but also looking at your reaction time end-to-end (like I did) where there is potentially no limit to how fast something can be. I'll be honest, I came to this expecting to debunk the whole shebang but I'm starting to think they might be on to something... at least up to point. 240Hz is probably overkill, but maybe 60 isn't as good as a lot of us assume it to be.
I hardly want to delve too much into this thread, but it's important to remember that the study listed above had nothing to do with motion, but identifying what's in a single frame, with no benefit of persistence of vision, which is present in any kind of gaming scenario.

Perception isn't something that just crosses over a line and disappears; we're able to perceive different kinds of things at different "framerates." For example, humans can perceive flicker artifacts at 500 Hz. There are different degrees of perception and it depends on issues like the lighting, the type of content, the variance from other content, how things are being interacted with, and even the color. Remember, this is a brain processing problem; people don't explicitly see in framerate.
 
Reactions: Born2Hula
There's two different ways of looking at it - taking the fastest thing you COULD react to, in which case I think you're right at 77fps, but also looking at your reaction time end-to-end (like I did) where there is potentially no limit to how fast something can be. I'll be honest, I came to this expecting to debunk the whole shebang but I'm starting to think they might be on to something... at least up to point. 240Hz is probably overkill, but maybe 60 isn't as good as a lot of us assume it to be.
Oh, I have no doubt that 60 is not the limit, though it's quite good.

Also, as @DSzymborski noted, there are different types of perception.

Then as well, there's perception, and there's also the ability to process enough of it to make a choice, then beyond that, the ability to physically react.


Also, note that a jump from 60 to 240 directly is going to be noticable. But, I also suspect that, say, 60 to 100 would be far more noticable than, say 144 to 240.
 
Reactions: Born2Hula

ASK THE COMMUNITY