1 00:00:00,000 --> 00:00:07,400 Everyone knows that adding latency to your gaming experience is bad, but 2 00:00:04,584 --> 00:00:12,000 [music] exactly how much latency is too much latency? At what point can you 3 00:00:09,680 --> 00:00:17,920 legitimately blame the lag for being hard stuck at copper seven? To find out, 4 00:00:14,960 --> 00:00:22,960 we've assembled a cracked team of four certified gamers that range in skill 5 00:00:19,760 --> 00:00:24,600 level from absolute hack all the way to 6 00:00:22,960 --> 00:00:28,880 god like. And we're going to tell you exactly how 7 00:00:27,040 --> 00:00:33,560 much input latency it takes to measurably affect your KD ratio. 8 00:00:31,640 --> 00:00:37,640 And all of this is possible thanks to ASUS who sponsored this video. We'll be 9 00:00:35,840 --> 00:00:41,920 using their ROG gaming peripherals including their new [music] 27-in tandem 10 00:00:40,400 --> 00:00:49,960 OLED monitor that can switch between 720p 720 Hz and 1440p 540 Hz using just 11 00:00:48,120 --> 00:00:52,800 a hot key. And it can do all of that without relying on display stream 12 00:00:51,640 --> 00:00:56,760 compression. But enough ado, let's 13 00:00:55,080 --> 00:01:00,960 ado this. Every test subject will use one of the 14 00:00:58,680 --> 00:01:05,480 three identical setups behind me. They use the same high-performance ASUS PCs, 15 00:01:03,200 --> 00:01:09,680 the same low-latency gaming peripherals, and the same silky-smooth mousing 16 00:01:07,760 --> 00:01:15,800 surfaces courtesy of these desk pads from lttstore.com. The real star of the 17 00:01:12,200 --> 00:01:18,200 show today though is these black boxes. 18 00:01:15,800 --> 00:01:24,080 Inside each one of these is an Arduino Leonardo that connects to this USB hub. 19 00:01:21,240 --> 00:01:31,560 This setup allows us to add latency in 1 ms increments from 0 to 100 ms which I 20 00:01:28,440 --> 00:01:33,960 remind you is 1/10 of a second. 21 00:01:31,560 --> 00:01:37,320 That would be terrible. Or would it? 22 00:01:35,640 --> 00:01:42,520 Each participant will complete three sets of tests while David here secretly 23 00:01:40,200 --> 00:01:46,440 changes their latency in between and then benchmarks their gaming performance 24 00:01:44,320 --> 00:01:50,520 using Aim Labs. It's important to note that we tested our base system latency 25 00:01:48,480 --> 00:01:54,040 at about 12 and a quarter milliseconds which we will be adding to whatever we 26 00:01:52,320 --> 00:01:59,800 have on our box which will give us our theoretical click to photon latency. We 27 00:01:57,720 --> 00:02:03,320 did consider adding latency to different parts in the chain, but in our 28 00:02:01,480 --> 00:02:06,720 preliminary testing we found that it didn't have a major impact and so we 29 00:02:05,080 --> 00:02:10,560 decided to keep it in one spot to keep it simple and consistent. But enough 30 00:02:08,679 --> 00:02:15,680 with the boring facts, let's game. 31 00:02:11,880 --> 00:02:19,516 I'm a 30-year-old unk at this point. 32 00:02:15,680 --> 00:02:20,480 I I go by baby no money. 33 00:02:19,516 --> 00:02:27,280 >> [laughter] >> And I make music and I play video games 34 00:02:24,640 --> 00:02:32,040 occasionally joke. My gaming ability out of 10, I used to be pretty good, but now 35 00:02:30,200 --> 00:02:36,960 I'm garbage cuz all these these 12 12-year-old kids are just insane. I 36 00:02:34,640 --> 00:02:41,840 don't really play much like competitive gaming anymore cuz it's just unhealthy. 37 00:02:39,160 --> 00:02:46,120 Lock in, bro. Baby no accuracy. 38 00:02:43,120 --> 00:02:47,000 >> 45, come on now. Did I have any latency 39 00:02:46,120 --> 00:02:50,560 or no? No no no no no. WOAH. WOAH. 40 00:02:49,520 --> 00:02:55,600 THAT'S WHERE YOU'RE MESSING WITH THE VARIABLES. >> WOAH. OH, MAN. Oh my god, I'm sweating. 41 00:02:54,000 --> 00:03:00,400 Literally [music] dripping in sweat. Brother, he hit 50k. 42 00:02:58,320 --> 00:03:04,080 High score. Wait, hold on. Wait, you're actually talking. Wait wait yo, crazy? 43 00:03:03,000 --> 00:03:08,240 All right, this next one's going to be a doozy. Oh, this one this next one wraps 44 00:03:06,640 --> 00:03:11,680 wraps the whole thing up. This is straight torture. It looks so easy when 45 00:03:10,240 --> 00:03:15,560 you watch someone else do it and you're just like bro, just point at the thing. 46 00:03:14,240 --> 00:03:20,160 Oh, that's >> We've reached new depths. Wow, bro. 47 00:03:18,160 --> 00:03:23,080 >> New lows. It's got some tips. No, there's tips for you. There's tips. 48 00:03:21,760 --> 00:03:26,880 There's tips. Keep it up. >> [laughter] 49 00:03:25,360 --> 00:03:31,480 >> Yeah, and there's another tip on the bottom saying like yeah, we're getting weird latency issues. 50 00:03:29,602 --> 00:03:35,680 >> [laughter] >> That was better. Hey, look at that. It's 51 00:03:33,200 --> 00:03:40,880 suggesting you become a pro. Is that your arm hurting now? Oh, yeah. 52 00:03:38,760 --> 00:03:44,400 It's getting real now. All right, what if I what if I lock in 53 00:03:42,800 --> 00:03:49,240 like this? Okay. I'll share the screen. Less less input 54 00:03:46,360 --> 00:03:54,280 delay closer, bro. Yeah, yeah. Oh oh, locked in position. I really 55 00:03:51,680 --> 00:03:59,080 don't think this is going to help. All right, more. Definitely had latency 56 00:03:56,760 --> 00:04:06,720 there, too. Okay. Okay, pro perspective though. Is this actually useful? This is 57 00:04:02,200 --> 00:04:14,000 useful maybe 20% of the time in my 58 00:04:06,720 --> 00:04:15,120 opinion. 1 2 3 4 5 6 7 8 9 10 59 00:04:14,000 --> 00:04:19,640 11 Oh, wait. The counting is off. Wait wait 60 00:04:17,200 --> 00:04:23,520 wait wait, crazy crazy? Okay, over flicks under flicks perfect flicks. 61 00:04:21,840 --> 00:04:26,600 I just want to be an okay flick. Now, there's too many unders. Yeah, there's a 62 00:04:24,880 --> 00:04:32,120 lot of under. What does that mean? That means I'm shooting lower, right? >> tracking it, but you're shooting it 63 00:04:30,040 --> 00:04:34,800 basically way too behind it. Yeah, good job. Good job. 64 00:04:33,440 --> 00:04:38,120 Solid. Solid. Sorry sorry. 65 00:04:47,920 --> 00:04:53,800 My name is Leon Kim also known as Kanata. I'm a professional Fortnite 66 00:04:51,520 --> 00:04:59,280 player for Dignitas. I think I'm really good at gaming and like I can go like 67 00:04:57,400 --> 00:05:04,080 basically go to any game and be good at it. [music] I think I'm like 8.5 OUT OF 68 00:05:01,720 --> 00:05:06,598 10. >> [screaming] 69 00:05:07,760 --> 00:05:13,748 >> I THINK YOU'RE UNDERSELLING YOURSELF. >> [laughter] 70 00:05:10,800 --> 00:05:17,120 >> WHENEVER you are ready, go for it. [music] All right, bet. I'm starting. 71 00:05:15,080 --> 00:05:20,920 Oh, he's going for a base score. This one's lucky easy. 72 00:05:18,720 --> 00:05:23,880 His movements are so much more precise than mine. This guy's lucky messing with 73 00:05:22,400 --> 00:05:30,240 the input delay over here. You can't really tell on this one though cuz this one's just easy. Like I could be on like 74 00:05:26,960 --> 00:05:32,720 20 Hz type of thing and it'll be fine. 75 00:05:30,240 --> 00:05:36,400 It's not that it's not that hard. It's too easy over here. Okay, this one 76 00:05:34,720 --> 00:05:41,400 is crazy. This one like you're moving and you're not even moving. 77 00:05:39,680 --> 00:05:45,880 Like I want to move, but it's like it's like I'm dragging through mud [music] type of movement. 78 00:05:44,400 --> 00:05:51,320 It's like someone's holding my legs back in a race. Damn, dude. 79 00:05:48,640 --> 00:05:55,680 How my reaction time go up? All right, next task. Oh yeah, this one 80 00:05:53,360 --> 00:05:59,600 is bad. Yeah. Okay, that was terrible. Oh my god, that 81 00:05:57,440 --> 00:06:03,520 was terrible. Tenses up your forearms too much. Last time I did it. Oh my god. 82 00:06:02,440 --> 00:06:06,804 Oh, that's terrible. >> really bad. Like the pixels aren't even 83 00:06:05,680 --> 00:06:11,640 working. >> [laughter] >> No, it's over. It's over. No, that one 84 00:06:09,560 --> 00:06:14,840 that one was impossible. They were all the input delays actually it's actually 85 00:06:13,120 --> 00:06:17,800 just affecting the whole thing. Okay, this one feeling good. 86 00:06:18,320 --> 00:06:24,320 Okay. That one felt better. That one felt a lot better. 87 00:06:22,640 --> 00:06:27,800 All right, last one. Damn, your arm actually does like tense 88 00:06:25,920 --> 00:06:30,680 up, bro. Like it it it actually like messes you up. 89 00:06:31,760 --> 00:06:37,960 How long is this thing? Okay. Damn, you're so close to 67%. 90 00:06:35,880 --> 00:06:42,829 >> Oh my god, I hit 67. Oh no, you were. 67. 91 00:06:40,920 --> 00:06:46,680 You guys are too old for this, man. >> [laughter] >> Okay, this one was cocked this one's 92 00:06:44,960 --> 00:06:52,240 cocked. This one's cocked. This one's cocked. This one's cocked. Oh my god, 93 00:06:49,280 --> 00:06:52,240 it's really bad. 94 00:06:55,929 --> 00:06:57,949 >> [music] 95 00:07:01,640 --> 00:07:06,840 >> As many of you will know, I'm Linus from Linus Tech Tips. Love technology, love 96 00:07:05,600 --> 00:07:10,800 gaming. Not really as good at gaming as I am at 97 00:07:08,640 --> 00:07:14,680 tech. I'm confident that I will be able to try my best. Lock in, bro. 98 00:07:13,360 --> 00:07:18,199 Bro, I don't feel it. Oh, that one looks bad. >> to begin. Okay, you're not supposed to 99 00:07:16,720 --> 00:07:20,219 influence the test. >> [laughter] 100 00:07:21,400 --> 00:07:29,040 >> Okay. Pretty good. So theoretically, the more latency that 101 00:07:26,240 --> 00:07:33,800 you add, the reaction time should just naturally go up. Well, here's the thing. 102 00:07:30,880 --> 00:07:40,240 We're not adding so much latency that your human reaction time and movement is 103 00:07:37,520 --> 00:07:45,280 likely to be a major factor. The latency we're adding is more likely to make it 104 00:07:42,520 --> 00:07:49,320 so that you will overshoot things and you're not as It's like you're not 105 00:07:46,560 --> 00:07:52,760 as connected to the machine. Are the routes the exact same? No. No, 106 00:07:51,440 --> 00:07:56,000 it's they're randomly generated. >> Okay, yeah. How many other variables do 107 00:07:54,560 --> 00:08:01,720 you guys have? No, no. Don't worry about that. >> [laughter] 108 00:07:58,600 --> 00:08:01,720 >> Next task. Okay. 109 00:08:03,160 --> 00:08:07,680 Pretty good. >> [laughter] 110 00:08:06,520 --> 00:08:10,640 >> Hey, man. I'm trying to be positive reinforcement here. 111 00:08:14,530 --> 00:08:16,550 >> [music] 112 00:08:20,080 --> 00:08:25,120 >> So my name's Typical Gamer and I've been live streaming and making videos since 113 00:08:23,760 --> 00:08:29,760 2009. So I'd rate my gaming ability out of 10 114 00:08:27,600 --> 00:08:32,870 for Fortnite, I would put it at like a 8.5. I feel like there's got to be Yeah, 115 00:08:31,960 --> 00:08:36,880 there's latency on this one already. [music] They pre-fired it. Yeah. 116 00:08:35,440 --> 00:08:42,560 Don't even don't think about that though. Just do your best every time. >> Yeah, come on Typical, bro. 117 00:08:40,640 --> 00:08:44,919 Oh, I'm still good. I'm still good. I'm still good. 118 00:08:46,440 --> 00:08:53,880 Wait, hold on. Oh yeah, you're Oh, CHILL OUT. 119 00:08:51,040 --> 00:08:53,880 56 120 00:08:54,800 --> 00:09:00,600 UNK'S GOT IT. UNK'S GOT IT. UNK'S literally got it. 121 00:08:58,160 --> 00:09:05,480 Oh my Oh, what is this? Oh, this is maximum delay. 122 00:09:02,720 --> 00:09:08,960 That was you, though. Holy. That was so bad. All right, last one. 123 00:09:07,640 --> 00:09:12,760 I'm bricked up right now. What? 124 00:09:11,160 --> 00:09:15,200 What? That is a Fortnite term, though. 125 00:09:14,680 --> 00:09:19,480 Bricked up? >> Bricked up? Yeah, it's when you [music] have max brick. No. 126 00:09:18,320 --> 00:09:23,840 >> I don't know about that one, dude. >> Yo, can you can you give me your brick? I've never heard that one [laughter] 127 00:09:21,440 --> 00:09:23,840 before. 128 00:09:31,166 --> 00:09:37,360 >> [music] >> We have enough data at this point, but 129 00:09:35,800 --> 00:09:41,920 one thing we thought would be kind of interesting would be to go head-to-head 130 00:09:39,800 --> 00:09:46,040 and find out just how much latency it takes for 131 00:09:44,080 --> 00:09:50,161 one of our more typical gamers to beat one of well, typical gamer, but also 132 00:09:48,480 --> 00:09:54,440 like the better gamers. >> [clears throat] >> Are you interested to know? 133 00:09:53,360 --> 00:09:59,040 Let's do it. I'm like actually I'm talking I know you're talking about 134 00:09:57,560 --> 00:10:01,680 awareness. No, I mean you guys Leave a comment below if you want. 135 00:10:00,641 --> 00:10:06,160 >> [laughter] >> All right, ready? Let's go. Come on, match starting. I'm locking Linus. 136 00:10:05,226 --> 00:10:09,960 >> [laughter] >> All right, 3 2 1 peek. No. You saw him 137 00:10:08,280 --> 00:10:14,600 on the left. I'm not going to. I don't want it. Move this way then. I could 138 00:10:11,920 --> 00:10:18,560 stay alive. Go this way. Okay. Look down. I could be alive man. 139 00:10:16,440 --> 00:10:21,640 I could be alive man. All right, 3 2 1 peek. No, I don't want it. You should 140 00:10:20,120 --> 00:10:25,560 peek. I think it's a really good time to be Just peek. 141 00:10:28,152 --> 00:10:33,840 >> [groaning] >> Wow, that was disgusting. 142 00:10:32,082 --> 00:10:38,080 >> [laughter] >> Oh, okay. They they have nerfed me. I 143 00:10:36,040 --> 00:10:42,400 will say. I TOLD YOU BRO. 144 00:10:40,240 --> 00:10:45,837 I TOLD YOU. IT'S OBVIOUS. ALWAYS right side. It's 145 00:10:44,600 --> 00:10:49,360 obvious, man. >> [laughter] >> Obvious right side right there. Okay, 146 00:10:47,760 --> 00:10:53,680 this is Okay, now it feels a little wavy. 147 00:10:50,920 --> 00:10:57,560 Oh, okay Linus. He was locked in on that. It feels like 148 00:10:55,520 --> 00:11:02,680 I'm like it's like a pendulum when I'm swinging. 149 00:10:59,760 --> 00:11:07,720 Oh, eyes. Oh my. Awareness. It's a negative awareness, man. Now Did you 150 00:11:05,160 --> 00:11:11,480 just call me negative awareness? That was negative awareness right there. What 151 00:11:09,360 --> 00:11:14,657 kind of a thing to say is that? We are now at 75 ms. 152 00:11:13,720 --> 00:11:20,280 OH MY >> [laughter] >> WOW. THAT'S A TOUGH GO. 153 00:11:19,400 --> 00:11:26,480 All right, let's go, bro. >> Let's swap. >> Are we swapping? All right, we're swapping. 154 00:11:23,160 --> 00:11:28,360 GG GG good game. GG. 155 00:11:26,480 --> 00:11:31,360 Dude, this is probably the dopest lab I've ever been a part of. 156 00:11:31,440 --> 00:11:36,560 Oh my, what the 157 00:11:34,760 --> 00:11:39,960 Bro, why is Kenna actually nasty? Yeah, this is actually nasty, dude. You got to 158 00:11:38,560 --> 00:11:43,600 find awareness on it. You saw that, right? You saw that shot? Yeah, but 159 00:11:41,760 --> 00:11:46,960 that's that's just being better at it. That had nothing to do with awareness. 160 00:11:45,440 --> 00:11:50,080 He knew where you were just as much as you knew where he was. You just are way 161 00:11:48,839 --> 00:11:54,160 faster. All right, don't play with the same angle too much. You're doing great. 162 00:11:52,560 --> 00:11:57,839 Okay, you need awareness. You guys are the worst coaches ever. He's never going 163 00:11:56,560 --> 00:12:04,438 to peek it twice, all right? Unless right now. Then he's aware that I said that. 164 00:12:01,000 --> 00:12:05,640 Oh, I had that. Dude, I had that. 165 00:12:04,438 --> 00:12:09,023 >> [laughter] >> Like maybe like Okay, he's so sweaty, 166 00:12:08,280 --> 00:12:13,280 bro. >> [laughter] >> He's actually not. He's actually not. 167 00:12:11,839 --> 00:12:16,880 Now's maybe a good time to talk about the monitor. Tandem OLED that gives you 168 00:12:15,240 --> 00:12:20,440 boosted brightness as well as better longevity according to ASUS. They've got 169 00:12:18,520 --> 00:12:25,360 their true black glossy coating so you get that gorgeous OLED contrast and it 170 00:12:23,280 --> 00:12:30,880 looks almost as good from the back. Oh, you Got transparent accents, RGB, 171 00:12:28,440 --> 00:12:33,800 absolutely flipping gorgeous. It is gorgeous. 172 00:12:32,240 --> 00:12:37,800 Oh, and don't forget that it has ROG OLED Care Pro which includes the Neo 173 00:12:36,000 --> 00:12:42,240 Proximity Sensor which means the display will turn off when you step away and 174 00:12:39,600 --> 00:12:48,480 back on when you step back to games. Okay, we're at 50 ms. 50 ms. No way. 175 00:12:47,360 --> 00:12:56,120 Peek, bro. Peek that? Yeah, peek. Should I peek it? Peek it. You think so? Yeah, do it. 176 00:12:53,040 --> 00:12:56,120 All right, 3 2 1. 177 00:12:56,183 --> 00:12:58,203 >> [laughter] 178 00:12:59,560 --> 00:13:06,320 >> NO. >> [screaming] 179 00:13:03,360 --> 00:13:11,600 >> HOW'S THAT AWARENESS? AWARENESS, COACH. Maybe no chill. Yeah, you got clapped, 180 00:13:09,200 --> 00:13:11,600 bro. 181 00:13:16,800 --> 00:13:21,920 No, no, no, we're good. Yeah, give me some latency, bro. 182 00:13:22,720 --> 00:13:28,200 Put it in the hands of God there for a second. I saw that. 183 00:13:25,960 --> 00:13:31,960 Prayed it went the right way. Oh, I had that. I swear I swear you you hit him 184 00:13:30,600 --> 00:13:36,320 through the wall like twice. Yeah, I know. 185 00:13:34,360 --> 00:13:39,320 See that? See how I just did that? You think I would peek this thing? 186 00:13:40,000 --> 00:13:45,080 LET'S GO. SIT ON. I JUST [laughter] SAT THERE. I SHOULD 187 00:13:43,760 --> 00:13:48,920 HAVE JUST LET'S GO, DUDE. He's losing confidence. 188 00:13:47,000 --> 00:13:52,000 Losing confidence, losing confidence, no awareness? Yeah, ego challenge. Ego 189 00:13:50,680 --> 00:13:56,960 challenge. Every opportunity. 190 00:13:54,000 --> 00:14:00,560 Oh. How much has he got now, David? 75. Give me like 100, bro. Oh, it's over. 191 00:13:58,960 --> 00:14:03,600 Give me like 120 and that's it. >> It's over. You won. It's over. 192 00:14:02,320 --> 00:14:08,960 A lot more analysis is going to be needed for the Aim Lab stuff, but what is clear from the 1v1 is whether you 193 00:14:07,520 --> 00:14:12,819 call it awareness, whether you call it experience, 194 00:14:11,320 --> 00:14:16,959 if you're not good at the game, >> [music] >> no amount of latency advantage is going 195 00:14:14,839 --> 00:14:20,560 to make you good at the game compared to someone who's got it. It was a great 196 00:14:19,440 --> 00:14:25,440 time. Um yeah, I mean it was a What day is today? 197 00:14:23,520 --> 00:14:28,360 Wednesday? I think so. It's a good day to be a Wednesday. 198 00:14:26,880 --> 00:14:34,435 It was a lot of fun. Yeah, I had a great time. That was pretty fun. I actually had a good time playing that Aim Lab 199 00:14:32,720 --> 00:14:40,240 with the different latency. Like >> [music] >> I can see why it affects gaming heavily. 200 00:14:38,160 --> 00:14:44,839 That was rough. Like I know that I'm average, 201 00:14:42,640 --> 00:14:47,535 but standing next to someone who's not average just makes it so much more 202 00:14:46,560 --> 00:14:53,400 apparent. >> [laughter] >> What surprised you the most about today? 203 00:14:51,000 --> 00:14:57,320 That this is your guys' job. It's really sick. Yeah. [laughter] 204 00:14:55,360 --> 00:14:59,680 Like the ability to basically just do science with the bros. 205 00:14:59,325 --> 00:15:05,000 >> [laughter] >> It's sick. 206 00:15:02,480 --> 00:15:08,200 So, how'd it go? Initial testing, it was super clear right away that input 207 00:15:06,560 --> 00:15:14,200 latency has a big impact on gaming performance. At just 10 ms, we were 208 00:15:10,680 --> 00:15:17,720 seeing almost a 7% drop and by 50 ms, 209 00:15:14,200 --> 00:15:20,920 over 25% drop in scores on average. 210 00:15:17,720 --> 00:15:23,440 Really, that is surprising to me because 211 00:15:20,920 --> 00:15:26,839 as one of the test subjects, I would say that at just a small amount of input 212 00:15:25,560 --> 00:15:31,760 latency, I didn't feel it. It didn't bother me. 213 00:15:29,280 --> 00:15:37,480 For me, there was very much a threshold, but you're saying that across our four 214 00:15:34,680 --> 00:15:41,280 subjects, even a little bit clearly matters. Yes, even before you can feel 215 00:15:39,560 --> 00:15:45,839 it which everyone sort of made comments at about 35 to 50 ms about you can cuz 216 00:15:44,120 --> 00:15:49,320 it's so yeah, jiggly. [music] Even before then, it's a pretty clear line 217 00:15:47,760 --> 00:15:53,240 down. And then what I also found interesting was that each task had a 218 00:15:50,920 --> 00:15:57,280 different difference from the input latency. The biggest difference was the 219 00:15:55,000 --> 00:16:02,600 The second king one. Everyone had a very clear decline across all skill sets. The 220 00:16:00,320 --> 00:16:05,920 third one was a pretty similar linear. And the thing I found really interesting 221 00:16:04,200 --> 00:16:10,400 about the first test is that its secondary measurement was just reaction 222 00:16:07,520 --> 00:16:14,920 time and I assumed that the more input latency I added would linearly add to 223 00:16:12,720 --> 00:16:19,360 the reaction time score, but it seemed like it kind of compounds. When you 224 00:16:16,440 --> 00:16:23,240 added like 50, it would increase by 100 ms. So there's kind of like a 225 00:16:21,000 --> 00:16:27,720 compounding effect in the tracking aim click. But one thing that I realized was 226 00:16:25,560 --> 00:16:31,680 that 10 to 50 ms [music] was not granular enough. I chose those 227 00:16:29,920 --> 00:16:34,040 metrics because when I was testing, that's when I could start to feel it. So 228 00:16:32,920 --> 00:16:39,240 I was like, oh, that's going to be, you know, the the thing. So that's what you were doing the second time. 229 00:16:37,240 --> 00:16:43,440 >> We did a second round of testing with 3 ms increments. Surely you're not going 230 00:16:42,040 --> 00:16:48,079 to be able to tell a difference from like 3 to 6 ms. 231 00:16:46,200 --> 00:16:52,839 The first round of this ad hoc testing was really interesting where there is a 232 00:16:49,800 --> 00:16:56,920 clear pattern down to about 233 00:16:52,839 --> 00:16:59,520 7 8 9% at the 9 ms range, but there was 234 00:16:56,920 --> 00:17:04,320 a weird jump at 6 ms performance where at 6 ms of added input latency, 235 00:17:02,480 --> 00:17:09,640 the average performance was higher than zero which made zero sense. Not 236 00:17:07,360 --> 00:17:13,600 logically, not intuitively, but I wasn't really satisfied with that so I tested 237 00:17:11,120 --> 00:17:17,560 another like 14 15 people. Oh. >> Those results were 238 00:17:15,600 --> 00:17:22,560 weirder. And so it goes down at 3 ms, up a little 239 00:17:20,640 --> 00:17:26,640 bit again at 6 ms and then down to almost the same at 9. So 9 and 3 are 240 00:17:25,240 --> 00:17:32,000 almost identical. So I started to wonder, does [music] it not matter below 10 ms? 241 00:17:30,800 --> 00:17:35,920 But then I separated top performers from bottom 242 00:17:34,160 --> 00:17:41,640 performers. Oh. [music] And in the top performers, it's a super 243 00:17:38,320 --> 00:17:43,040 clear line. Like undeniably just 3% each 244 00:17:41,640 --> 00:17:47,920 added 3 ms. >> So if you're any good, 245 00:17:45,160 --> 00:17:50,440 then functionally speaking, there is no limit to how low you want your latency 246 00:17:50,040 --> 00:17:54,880 to be. [music] >> Yeah, we could try to get a system that would get us 247 00:17:53,120 --> 00:17:58,920 less than the 12 and a half that we were testing with, but it's not a realistic 248 00:17:57,600 --> 00:18:03,160 scenario in any game. >> gamers though, that [music] could potentially make a difference. See, I 249 00:18:01,640 --> 00:18:04,960 was not expecting this. Me neither. I was hoping that this video would 250 00:18:04,692 --> 00:18:09,960 >> [music] >> tell people that it's really your gaming skill that's the problem. Like your 251 00:18:08,040 --> 00:18:15,679 peripherals are fine, but what I found out is that >> is Wait, and that is true to a degree. I 252 00:18:13,560 --> 00:18:20,880 mean you saw how much latency we had to add to 253 00:18:17,560 --> 00:18:23,600 either Kenna or typical gamer, right? 254 00:18:20,880 --> 00:18:26,880 For me to have any shot of hitting them. But 255 00:18:24,840 --> 00:18:30,800 if you have the game sensor What did What [music] did Kenna call it? Uh 256 00:18:28,840 --> 00:18:34,640 awareness. You got to find awareness on it. You saw that, right? You saw that 257 00:18:32,320 --> 00:18:38,760 shot? Awareness. It's a negative awareness, man. Then there is no limit 258 00:18:37,360 --> 00:18:43,080 functionally that we found. >> No, we can confidently say that as 259 00:18:40,640 --> 00:18:47,880 little as 3 ms impacts gaming performance for high 260 00:18:44,960 --> 00:18:52,040 not even high, above average gamers. Huh. It's a It's the takeaway that I 261 00:18:50,400 --> 00:18:57,159 find interesting is that most peripherals now are pretty low latency. 262 00:18:54,280 --> 00:18:59,880 Gaming mice are sub 1 ms. Sub 1 ms if they're good, but I bet there's a lot of 263 00:18:59,159 --> 00:19:05,320 you out there that are playing competitive games on Bluetooth mice or on the wrong monitor mode. Monitors are 264 00:19:03,919 --> 00:19:09,320 often the biggest source of input latency. If you're gaming on a display 265 00:19:07,080 --> 00:19:13,760 that also has a film mode, make sure you are off that because that's how you get 266 00:19:10,840 --> 00:19:16,720 to like 100 ms of delay. And like that was enough to get you on on level with a 267 00:19:15,840 --> 00:19:22,679 pro gamer. >> Right. [music] And so When we say source of input latency, we mean because 268 00:19:21,000 --> 00:19:26,760 technically it's not the input, but it's the input to your eyes that will affect 269 00:19:24,600 --> 00:19:30,679 when you put your input in. It also makes sense that all the esports teams 270 00:19:28,640 --> 00:19:36,560 are all in the same cities because that 25 ms extra of LA to Dallas is enough to 271 00:19:34,720 --> 00:19:39,720 drastically affect your performance. Well, there it is. I 272 00:19:38,560 --> 00:19:46,320 mean ASUS will be happy about [music] that. PC gaming, is it pay to win? 273 00:19:43,960 --> 00:19:49,560 Well, if you have the awareness, then yes. 274 00:19:47,520 --> 00:19:53,320 If you don't, then there's not much you can do about it. No. I was kind of 275 00:19:51,360 --> 00:19:56,120 hoping there'd be some ambiguity here because 276 00:19:54,640 --> 00:20:01,240 you know, obviously ASUS sponsored the video and it's in their best interest for lower latency, high performance 277 00:19:59,600 --> 00:20:04,680 peripherals [music] to make a difference to gaming. Just like we would, sponsored 278 00:20:03,520 --> 00:20:10,360 or not. All we're doing is giving you guys the results that we have, semi-scientific as 279 00:20:08,360 --> 00:20:13,320 they are. Uh you can check out we'll have maybe some uh downloads of the 280 00:20:12,000 --> 00:20:17,640 graphs if you guys want to see them down below. >> sure. Yeah, we'll have a little package for you guys. We'll also have links to 281 00:20:16,360 --> 00:20:21,160 all the products that we used down there in the video description. Uh or you can 282 00:20:19,480 --> 00:20:24,440 check out the LTC Labs website for our growing database of in-depth testing of 283 00:20:23,000 --> 00:20:29,920 gaming peripherals including latency measurements of mice and keyboards. 284 00:20:27,440 --> 00:20:34,480 Huge thank you Typical Gamer, Baby No Money, and [music] Canada for joining us 285 00:20:31,800 --> 00:20:37,520 today. I had a ton of fun, at least as much fun as the last time we did a video 286 00:20:35,840 --> 00:20:41,360 like this, which maybe you should check out. It was more focused on how many 287 00:20:39,480 --> 00:20:42,760 frames per second you need before it affects gaming performance. 288 00:20:42,549 --> 00:20:46,920 >> [music] >> And the results may surprise you. 289 00:20:45,600 --> 00:20:51,760 If you completely weren't paying attention today. >> [laughter] 290 00:20:49,600 --> 00:20:54,000 >> Turns out more is better. Shh.