Blur Busters Forums

jorimt wrote: ↑

09 May 2020, 08:54

Avean wrote: ↑

09 May 2020, 03:51

Its basicly this part im questioning. I am all for high refreshrates and the minimum of input lag as possible, but its more the advertisement from nVIDIA "See the enemy first" i am having a hard time understand.

To be entirely fair to that comparison screen, the fact that 240Hz looks like way less of a reduction over 144Hz vs 144Hz over 60Hz in that particular test scenario is a bit of an optical illusion; if we start from the fingertips and measure the difference in distance, the delivery speed increase of 240Hz over 144Hz is still 2/3rds of the delivery speed increase of 144Hz over 60Hz:



E.g. if they had captured that screen starting with the hand appearing from behind the wall first, nearly the entirety of the hand would have appeared at 240Hz before 144Hz even started to show the very tip of the fingers (if that).

Avean wrote: ↑

09 May 2020, 03:51

Like you say, research show reaction times below 100ms. But we are talking about less than 10ms of difference in this example between 144Hz and 240Hz. The idea that a player is gonna take advantage of this when its happening so extremely fast. Its just a few milliseconds ... Tons of other benefits of course, but that you have an advantage in seeing the enemy first

I'm not entirely certain where you're getting that 10ms number, but, again, let's say someone has a 200ms reaction time. You add 10ms to that, and you now effectively have a higher reaction time maximum. Then, add, say (pulling these numbers completely out of my hat; for example only), 8ms from your mouse, 5ms from your monitor, 15ms+ from the game engine buffering (etc), and you have now artificially increased your max average reaction time by a fair amount.

Look, I personally don't care about the difference between 144Hz and 240Hz either. It is of my opinion that such small decreases in ms only help players that are already good at the game. In other words, if you suck at the game, it probably isn't those extra few ms that are making you perform so poorly, and/or you have a really, really bad ISP.

That, and we're all free to think that 240Hz or higher "has [little to] no advantage," and to stick to our 144Hz display (I will also say here there can still be a slight justification to this, not because 240Hz isn't better, but because so few 240Hz panels perform quite as well with overdrive as 144Hz displays, but this is another subject, and something that is getting better). That said, again, it doesn't mean these advances are not "worth" it, just because some of us can't imagine any value to some aspects of the posed reductions.

I'll say it again; refresh rate is an artificial limitation, motion blur is an artificial limitation, input lag is an artificial limitation. The fact that we're event talking about this shows how far displays have to go before refresh rate discussions become obsolete. Ideally, refresh rates and achievable framerate should ultimately become so high that it's not even something the consumer thinks about any more.

But we're not there yet...

Yeah but considering the same setup except different monitors. Same mouse, same graphics card, playing in the same game with exact same ISP but one is on a 144Hz monitor and another on a 240Hz monitor. Then the difference is only like 3 or 4 milliseconds right between 144Hz and 240Hz? Its then i have a hard time believing a 240Hz player would see you first than someone on a 144Hz monitor because of human reaction times is much greater than 3-4ms. Sorry for long discussion btw, appreciate your answers Just have a habit of needing to know if my theories are correct hehe.

Imagine a mirror copy of you playing against you. 1 sits behind the monitor 240, 2 behind the monitor 144. That 1 will receive information about the second 2.7 ms earlier. The speed of your reaction does not matter. If it is 200ms, then 1 will press the button after 204.2 ms, and 2 will press 206.9 ms. If it is 1000ms, then 1 will press the button after 1004.2 ms, and 2 will press 1006.9 ms.

Avean wrote: ↑

09 May 2020, 09:15

Yeah but considering the same setup except different monitors. Same mouse, same graphics card, playing in the same game with exact same ISP but one is on a 144Hz monitor and another on a 240Hz monitor. Then the difference is only like 3 or 4 milliseconds right between 144Hz and 240Hz?

Depends on how you're counting it, and what context of input lag we're talking about. For G-SYNC, 143 FPS @144Hz vs 237 FPS @240Hz is between 3ms and 5ms where input lag reduction due to pure scanout duration decrease is concerned, yes, but there are other factors that make 240Hz more responsive; you've got the decreased scanout duration (regardless of framerate), and 96 more refreshes per second (with sustained 144+ FPS vs. 240+ FPS), which affects many other overall aspects, including motion clarity.

That, and as I said previously, no, I personally don't think that the difference will make you a better player if you're already struggling at 144Hz. But, again, that's not an excuse for display technology to stop dead in its tracks because the consumer personally feels satisfied with their 144Hz display.

Avean wrote: ↑

09 May 2020, 09:15

Its then i have a hard time believing a 240Hz player would see you first than someone on a 144Hz monitor because of human reaction times is much greater than 3-4ms.

The difference is counted after the human reaction time is accounted for. Additive.

But yes, if you're a 240Hz user with a middle-tier PC, mediocre ISP/ping, a 215ms reaction time, and an okay k/d, and you come across a 144Hz user with a beefy PC, an excellent ISP/ping, a 160ms reaction time, and an above average k/d, unless the netcode of the particular game is messed up, you're probably losing to said user 9 times out of 10.

Skill trumps ms, but ms reduction (even as slight as 3-5ms) can augment skill regardless, especially in situations where the players are otherwise matched.

1000WATT wrote: ↑

09 May 2020, 09:51

Imagine a mirror copy of you playing against you. 1 sits behind the monitor 240, 2 behind the monitor 144. That 1 will receive information about the second 2.7 ms earlier. The speed of your reaction does not matter. If it is 200ms, then 1 will press the button after 204.2 ms, and 2 will press 206.9 ms. If it is 1000ms, then 1 will press the button after 1004.2 ms, and 2 will press 1006.9 ms.

And yet, none of that matters, especially when playing via internet. And even on LAN, there is so much more to gaming than just "pressing the button". What will matter more is crosshair placement, where you are looking at vs where other player is looking, and the whole specific and situational reaction time. That photo by NVIDIA is misleading and was done as a marketing ploy, it doesn't really works that way in real life and it doesn't give you that kind of advantage of seeing other players earlier. Every player vs player interaction is unique and highly situational. The reason why 240Hz is better is because the game is smoother, feels more consistent and there is a slight reduction when it comes to motion blur. And as far as real world "advantage" of 2.7ms of 240Hz is concerned, it may and will only work in a very specific and limited scenarios. You also need to consider that everyone's eyes are different. Improving vision related motion and game harmonics while reducing the picture specific artifacts, distractions and imperfections is also person specific. Some players will benefit more than others from higher Hz, depending on their own personal susceptibility to these picture artifacts and imprefections.

mello

On the one hand, from such trifles as 2.7 ms on the monitor, 2.7 ms on the mouse, 2.7 ms on the processor, 2.7 ms, etc. .. A decent number is added up. And that all theoretical advantage can easily be destroyed by such an unpleasant thing as “high ping compensation”. I do not argue with the statement that nvidia is engaged in marketing. And I do not claim that the results will certainly improve. I wanted to say that the human reaction in these calculations is irrelevant. I wanted to say one of them will press the button 2.7 ms earlier. And what will be the results of clicking is a completely different topic.

I often do not clearly state my thoughts. google translate is far from perfect. And in addition to the translator, I myself am mistaken. Do not take me seriously. So that you understand how difficult it is for me, I will translate for you the name of this topic - "The benefits of high refreshment?"

1000WATT wrote: ↑

09 May 2020, 11:24

On the one hand, from such trifles as 2.7 ms on the monitor, 2.7 ms on the mouse, 2.7 ms on the processor, 2.7 ms, etc. .. A decent number is added up. And that all theoretical advantage can easily be destroyed by such an unpleasant thing as “high ping compensation”. I do not argue with the statement that nvidia is engaged in marketing. And I do not claim that the results will certainly improve. I wanted to say that the human reaction in these calculations is irrelevant. I wanted to say one of them will press the button 2.7 ms earlier. And what will be the results of clicking is a completely different topic.

I just wanted to add another angle or another layer to the discussion around that, that's all. There is a lot of complexity when it comes to benefits and advantages of high refreshrates, and it is just not what NVIDIA marketing department has put its focus on.

Avean wrote: ↑

09 May 2020, 03:51

"See the enemy first"

<SEMANTICS ALARM>
BEEP BEEP BEEP BEEP BEEP BEEP BEEP BEEP BEEP BEEP BEEP BEEPBEEP BEEP BEEP BEEP
</SEMANTICS ALARM>



This is exactly true.

It's the Olympics finish-line effect.

Both players are seeing the enemy simultaneously.

But because of 240Hz instead of 144Hz, the enemy's photons will (on average) hit the eyeballs of the player with the 240Hz monitor of identical best-case absolute latency (but lower scanout latency) as the 144Hz monitor.

So... it's true. Sure, it's an advertising spin. Sure, the humans can't tell they're seeing the enemy first. But neither does the Olympics sprinters that are only mere milliseconds apart -- there are situations of single-digit millisecond finishes. Those racers have to stare at the scoreboard to see if they won the 100 meter race.

Same situation for a simultaneous-draw situation where both players are reacting at the same time. Not all reaction-time-events trigger a race-to-finish event, since a sniper is often camping. But other situations, like two Fortnite players who suddenly see each other at the same time when a wall gets chopped down. Or two players going down halls/alleys and suddenly seeing the enemy around a corner or doorway. Anything that triggers a simultaneous starting pistol type event (sudden simultaneous visibility). They both thus suddenly have to shoot/attack at the same time (relative to their reaction times), and the less laggy system is often more likely to get that score.

The player technically saw the enemy first, even if they didn't know it (it just feels like a simultaneous-draw situation).

Think about this from a different angle. (Geometry versus physics!)

Imagine this. The competitive stratospheres. The reaction times of 10 good players may only be approximately 25 milliseconds apart. By being 2.5ms less laggy (the difference between 144Hz and 240Hz), you're 10% earlier in that statistical spread. That can be quite useful when winning enough money in esports for the career of an esports champion.

Now up there in the stratospheres, the reaction time differences between different humans will be much tighter statistically. Say, only 10ms spread instead of 25ms or 50ms spread between best/worst players in the match. Like the differences between Olympics sprinters at the starting line! At this tighter reaction-differences spread, the milliseconds amplify...

It's logical that these kinds of factors could cover that few-percent improvement noticed on 144Hz vs 240Hz.

It doesn't always matter to us mere mortals, but the humble millisecond commands thy obligatory automatic Blur Busters respect...

We've run into many surprises to the point where we never "assume" or "dismiss" thy millisecond. Not guilty unless proven guilty is the mandatory rule we apply to milliseconds at Blur Busters

NOTE: Yes, I know VSYNC OFF frameslices are a latency gradient, and absolute latency can be identical on 144Hz and 240Hz. But absolute latency is NOT identical to scanout latency. Even with VSYNC OFF, the accelerated scanout of 240Hz means the VSYNC OFF frameslice (at the same frame rate) is 240/144ths taller on the 240Hz than the 144Hz, delivering more pixels per VSYNC OFF frame slice at the same frame rate, if spewed out to a faster scanout (240Hz instead of 144Hz). So even if best-case absolute latency can sometimes be identical, the scanout latency is lower, with more refresh opportunities per pixel. See high speed videos of scanout latency at www.blurbusters.com/scanout and also heed attention to the VSYNC OFF diagram too. Scanout latency is not the same thing as absolute latency, and each pixel can have different latencies (TOP vs CENTER vs BOTTOM) depending on the sync technology and settings configured. So the lag numbers don't always tell the whole story, as lag is never a single number. Just like for GtG -- it's usually an average, there can be over 60,000 different GtG numbers on the same panel. These rabbit holes are massive...