His method is pretty standard, but it looks like he's switching away from average to just the middle bar. Shouldn't really change the numbers much though.
https://displaylag.com/testing-method/
https://displaylag.com/testing-method/
If "lag" is a metric describing latency from source to display, then based on displaylag.com (and whatever other goofballs), my CRT has 16.67ms of lag, or 8.3ms if you take the middle bar. It's quite idiotic. Lag in the realm of 60hz is [latency from source to display] - 16.67ms. 0ms is perfection. It's that simple.
Take any gamer who understands frames and tell them that a particular display has 9ms of lag and they will instantly think, "that's half a frame." But it's not. That's the goofy displaylag.com speak for what is in reality 1ms of lag. And to further my point, so far as I've seen, there are zero manufacturers that report latency in the displaylag.com fashion. BenQ touts some of their "1ms" monitors, and that's actually what they are. [/nerd rant]
-ud
Right, 8.3ms is correct though. CRTs aren't technically 'zero lag' because they still have to write from top to bottom, and that takes time, however quickly. We measure from the middle because that's generally where we're looking. Display manufacturers like to measure from the top because it makes the numbers better. Neither is really right or wrong, but middle feels reasonable to me for 'real world' use.
It's somewhat meaningless, we're not capably of detecting 8 milliseconds like that, but it's there in terms of the physics of it.
All this means is that 10ms LCD is practically as good as the CRT, and display lag as a concept just isn't really relevant anymore, provided you have the right hardware. Anyone who says they missed their combo because of 2ms of lag in the monitor can fuck off lol
But there are no lag free LCD's. The top zone of an LCD still has a slower response time than a CRT, so it shouldn't be 0ms. There is a delay. And with a camera test, how often are we testing input lag, looking at the top of the screen? It's usually in the middle zone or lower, as that is where the controlled object usually resides, in order to do the test.The lag values on that site are reported in a stupid way. The Leo Bodnar tester he uses shows 3 flashing bars which can be used to detect display latency at that point on the screen. Displaylag guy adds all those up and divides by 3 to get an average. Based on his numbers, my BenQ RL2460 has 10ms of lag. It surely does not (I've done high speed camera testing against a CRT). We already know that a 60 hz display with zero lag (i.e. a CRT) will take 16.67ms to update 1 frame, so a lag free LCD should have the following readings at each bar of the Leo Bodnar tester:This is a great resource for that kind of thing: https://displaylag.comAnyone own any 32” 1ms 1080p monitors that they recommend?
00.00ms
08.33ms
16.67ms
The average of those is ~8.3. Subtract that from displaylag's averages to get proper numbers.
-ud
Either way, camera, leo bodnar, etc, is all close approximations IMO. Close enough tho to where it's roughly negligable. I mean, 8.3ms vs 10ms tells me that 8.3 will mean 50% of the time I will be at 1f of input lag, and at 10ms, im at like 60% or something.
Or atleast that is how I understand it, and I misinterpreted what you said completely =)
Sorry, I may not have been clear. Those 3 values are the ideal numbers we would want for a lag free display. Best LCDs are off by 1ms, which ain't no slouch
This sort of highlights the confusion that comes with such goofy metrics. There are 2 quantities of time that are being (erroneously) treated as one:
1. Refresh rate - the number of times the screen is refreshed in 1 second (in our case 60Hz)
2. Display lag - the time between the source outputting a pixel and the display showing it on the screen
Any 60Hz display (doesn't matter CRT, LCD, whatever) takes 16.67ms to draw a frame, hands down. Any 60Hz display (doesn't matter CRT, LCD, whatever) takes 8.3ms to draw 1/2 a frame, hands down. Nothing about those timing observations speaks to lag from source.
A CRT is drawing each pixel real time, raw, no frame buffering. There are 0ms (at a physical level were dealing in nanoseconds) between the RGB signals changing at the source and that pixel showing up on the screen. The low latency display (I've done testing on BenQ and Asus 1ms monitors), pixels only trail by ~1ms, which is a few scanlines worth of time.
-ud
P.S. you know I love you, Aurich
Last edited:
I feel like we're arguing about something we agree on lol. So in the middle of that CRT is around 9ms of time for something to be drawn right? You can't see the frame change, the punch come out, whatever, until that time passes. Now obviously we're not capable of really "seeing" that 9ms in a meaningful way, which is why 60fps looks smooth to us, but it's there. It's built into the refresh rate, it's just the physics.
You're saying the LCD does the same thing in the middle of the screen, just maybe a ms behind. Agreed, with you. So that 10ms number, compared to the 9ms number (rounding here) is something we agree on, and also matches what DisplayLag is saying.
I'm not arguing they're the same thing exactly from a technical standpoint, they aren't, but from a practical standpoint those numbers are real, and match up to real world experience.
So the DisplayLag numbers basically say that these gaming monitors, or new TVs, that are sitting roughly in that 10ms range, are CRT-like in performance. Right?
I'm sending chocolates in a heart, watch the mail.
MoppelTheWhale
Grand Master
on other forums this would have ended in a shit throwing contest, here it ends in candy - i love this forum
We know each other IRL, there's a mutual respect. Well I think so, I know I respect him ...on other forums this would have ended in a shit throwing contest, here it ends in candy - i love this forum
Yeah, Aurich and I know each other outside of the interwebs so it's easier to assume he's not personally attacking me when he disagrees with something I said. Though, I hope I can learn to show respect to everyone the same. I know I've lost my cool here and there in forum battles.on other forums this would have ended in a shit throwing contest, here it ends in candy - i love this forum
-ud
If the argument is that the player is always focused on the middle of the screen, then the player will always have to wait a full frame (16.67ms) to see a change. I think the 9ms argument falls apart, but whatever, I think we understand both perspectives (even though I'm right! lol).
-ud
I mean, you're right, for sure, but the middle still makes sense because when you're measuring display lag that's bigger than a frame it's a good place to focus. So keeps things standardized.
But really I think the point we're both getting at is once you're sub-frame it's basically a non issue. A gaming LCD is as good as a CRT as far as response goes, there's no longer a reason to be snobby about it. I still prefer CRTs for 15khz games, always will, my cabs are awesome and I'd never want them on LCDs. But for modern gaming it's "just as good" provided the rest of your hardware and software is up to snuff.
Heh, just to save time gonna link you to this thread where we talked about this already: Twistedsymphony's "Night @ the Arcade" #10
Short version: not anymore
Gonna slap this on a mad catz tournament edition fight stick , gonna make it costume to play Neo geo/cps2 on the HAS 
((Artwork done by my brother))
((Artwork done by my brother))
Attachments
Can you post a link where you got the fan from please?
Gelid Solutions Silent 6 Computer Case Fan https://www.amazon.com/dp/B008S1HNPS/ref=cm_sw_r_cp_api_i_Q9YxCbTEPCMRS
