[Ethernet Onboard-vs-USB-vs-PCIe] I finally found the reason behind my input lag after 6 years

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
Post Reply
6yToFindTheAnswer
Posts: 5
Joined: 30 Oct 2022, 20:16

Re: I finally found the reason behind my input lag after 6 years

Post by 6yToFindTheAnswer » 07 Nov 2022, 16:50

Mims wrote:
02 Nov 2022, 17:41
I'm new to the forum, just wanted to reply to your post about the issue that you found with intel NICs.

I have the same scenario, moved out from one house to another, replaced all of my components including monitor, mouse, keyboard outside of my PC ... Recently redid my electrical box on my new appartment, felt some improvements but still had this input lag + desync with other players in multiplayer game, getting killed without even seeing nothing...

Tried with an USB Ethernet adapter with it own network card integrated, felt directly the improvement while playing valorant, fortnite ... The issue with taking invisible shots, and feeling like people i play against are godlike player was gone ! I was finally being able to take my time while gunfighting, seeing the bullets coming toward me and being able to adjust my movement while doing so. But still have some input lag (probably because the networking isn't processed directly from the motherboard).

I have a I225-V adapter, with a motherboard Z690-a prime gaming.
The 2 other motherboard that i replaced have exactly the same NIC, plus a third one having I219-V, with the same input lag / desync issue.
VERY interesting, so you also tested unplugging your ethernet cable from a motherboard with intel i219-V and directly felt an improvement in game and even on desktop, just like me.

Also :!: I have found a video of a russ.ian guy talking about that so we are definitely not the only ones.

Here is the video :

phpBB [video]


With the very bad auto-translation of youtube; what i understand is that : apparently he says that other players movements were "insane or weird" when he played with motherboards with intel network cards and when he switched to Realtek network cards their movements were a lot smoother and easier to kill. Also he noticed that with the intel network card when he ran into his teammates he was having weird "shakings movement or rollbacks" which is not the case with the realtek. If any of you guys understand Russlan could you please investigate this and what he is saying :?: cuz with youtube auto translate it is really hard to understand.

I think i am going to buy a few PCIe to ethernet cards with a Realtek chip to test this on my own and report back to you guys.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: I finally found the reason behind my input lag after 6 years

Post by Chief Blur Buster » 07 Nov 2022, 21:42

There are indeed crappy on-board Ethernet implementations, and I've sometimes found a PCIe Ethernet card to produce better results.

Other times, the Ethernet port on a motherboard is really good.

It's very motherboard dependant.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

joseph_from_pilsen
Posts: 166
Joined: 01 Apr 2022, 23:51

Re: I finally found the reason behind my input lag after 6 years

Post by joseph_from_pilsen » 09 Nov 2022, 19:12

Sorry to lurk here because im very interested in this topic but...
¨
what if the OPs troubles are not caused by chipset neither network card. But pure fact he plays games with unfair netcode preferring high pingers /csgo, valo, MW /.
Logically, he performs better at LTE than at his home fiber. If he adds fake latency to his ftth (around 60ms in csgo, 35ms in valorant to maximize lagcompensation advantage as these values are a sweetspot pings) he will stop being only prefired and feeling behind because he will start being lagcompensated too what he never will be at his low lateny ftth (thats why chinese esport players play with VPN to singapore - to get at least some latency to be able to do open kills just by lagcompensated peeks). So they no more can see you before you see them. Both games (csgo and valo but also MW) have unfair netcode, csgo has 200ms max lagcompensation (thats why backtrack works so fine in csgo), valorant reduced it to 35ms hard limit but the movement is slower here, so its in the end the same - the peeker wins always. I dont play MW, so i dont know the exact value there but its said its present and high too.
Try to make custom server, a server with fair netcode, disable lagcompensation like in cs 1.6 (you need set sv_unlag to 0 or let it at 1 and set sv_maxunlag 0 from default 0.2s so the lagcompensation is zero). Then connect your PC and play with others. Ideally, invite more players to your local LAN. If you will no more suffer from being prefired, problem was definitelly in the netcode.
OR
You can spoof your latency to be the same like at LTE, install a router outside your pc with linux, just any PC with 2 ethernets (1ETH + USB/ETH port), install netem packet, set at outgoing eth0/1 port netem delay 50ms (or more if you want Russlan prefire), and enjoy free kills. You may also use it only at T side so you have peekers advantage when rushing (i do it with my LTE, when switching sides from CT to T, i vote a timeout and switch from cable to LTE so i have easy opens).

Again, if you start getting better and not feeling behind, its not hardware issue but netcode issue. Also mark server quality, for a valid test you need a good server with server variance max. 3ms +-0.x ms and 0.00x var. Otherwise its not a valid test, at trash server (nearly all official ones at csgo, all DM servers except few ones of warmupserver.net and a half of faceit servers) the aim is a coinflip and aim stutters due to bad server performance. Even 5ms sv is bad, i can notice very easy the difference between 5ms sv and 2.5ms sv server. OF course, its easy to notice bad server var too (the last row and last column in net_graph), any var higher than zero point zero zero x or any var unstability result in noticable microlags and aim and movement stuttering. There is also noticable inertia at higher sv or server var, and crosshair tends to keep moving after sidestep or jumps a little off after you stop your mouse and press counterstep and also the sprays are absolutelly wrong with added false horizontal recoil and which shouldnt be there, also the shooting speed is not exactly correct (almost if you have fast reactions, the server quality is crucially limiting, the faster you act the worse results).
Yes and important thing, throttle your connection! MW seems to run best at 3/0.5mbit limit so the opponents cant read your movement so easy, csgo runs fine at 80000 rates and valorant i dont know exactly now but this game seems to be least prone to throttling since they implemented a server sided player movement computing so players can no more trick their outgoing rates to trash their movement for being tough to hit by others. But in csgo and MW the throttling works perfectly. It also removes stuttering by high bandwith usage and reduces jitter and packet skipping.
Its very possible that at LTE/work the game is set to lower bandwith or simpler rendering automatically as it cant handle your home settings.

Then after you do all of this you can make a final verdict.
Im experiencing the same, but i cant tell definitelly a verdict as there are too many variables and this feeling you described i have only occasionally and yes, i feel better or at least not worse K:D and TTK ratio performance at 2mbps hard limited LTE than at my home ftth, what baffles me because the LTE jitters terribly with random rollback lag every few seconds as the ping spikes for 100+ms for a while. Now im installing kali linux and netem at a secondary PC to start to play with fake latency and making some more valid and exact results (LTE at my secondary address with random lags is not very good for that).
One day i will post my verdict of my research and investigation. Using Intel chipset :o

triplese
Posts: 130
Joined: 13 Dec 2021, 12:20

Re: I finally found the reason behind my input lag after 6 years

Post by triplese » 13 Nov 2022, 11:07

joseph_from_pilsen wrote:
09 Nov 2022, 19:12
But pure fact he plays games with unfair netcode preferring high pingers /csgo, valo, MW /.
Logically, he performs better at LTE than at his home fiber. If he adds fake latency to his ftth (around 60ms in csgo, 35ms in valorant to maximize lagcompensation advantage as these values are a sweetspot pings) he will stop being only prefired and feeling behind because he will start being lagcompensated too what he never will be at his low lateny ftth (thats why chinese esport players play with VPN to singapore - to get at least some latency to be able to do open kills just by lagcompensated peeks).
So choose warmupserver with 60ms and you will find that this is not lagcompensation related, because with such ping you will performing even worse. Dont have pleasant experience to be always prefired by Russlan kid with 150ms ping?
Problem is that internet is free (read as chaotic) network that not controlled by anybody, and no real debug instruments or strict standards is implemented. Disable ICMP anywhere? Yes, please. Fuck off gamers that cant even traceroute to the server. Its total mess and games developers cant do anything with that.

Here is great investigation from Riots how "reported" ping and real ping differs. Best illustration of desync.
https://lolesports.com/article/riot-gam ... 33b5d5a616

espresso
Posts: 63
Joined: 11 Feb 2022, 04:56

Re: I finally found the reason behind my input lag after 6 years

Post by espresso » 13 Nov 2022, 17:00

triplese wrote:
13 Nov 2022, 11:07


Here is great investigation from Riots how "reported" ping and real ping differs. Best illustration of desync.
https://lolesports.com/article/riot-gam ... 33b5d5a616
This is 1000% what Valve is doing as well!

User avatar
dervu
Posts: 249
Joined: 17 Apr 2020, 18:09

Re: I finally found the reason behind my input lag after 6 years

Post by dervu » 14 Nov 2022, 05:12

Lol. Are you serious or trolling?
This has nothing to do with their game being played online, but with solution made specifically for some specific conditions and you are trying to extrapolate that to other game. :lol:
Ryzen 7950X3D / MSI GeForce RTX 4090 Gaming X Trio / ASUS TUF GAMING X670E-PLUS / 2x16GB DDR5@6000 G.Skill Trident Z5 RGB / ASUS ROG SWIFT PG279QM / Logitech G PRO X SUPERLIGHT / SkyPAD Glass 3.0 / Wooting 60HE / DT 700 PRO X || EMI Input lag issue survivor

Anonymous768119

Re: I finally found the reason behind my input lag after 6 years

Post by Anonymous768119 » 14 Nov 2022, 07:07

How do you want to fix anything if you can't read with understanding?

espresso
Posts: 63
Joined: 11 Feb 2022, 04:56

Re: I finally found the reason behind my input lag after 6 years

Post by espresso » 14 Nov 2022, 13:22

dervu wrote:
14 Nov 2022, 05:12
Lol. Are you serious or trolling?
This has nothing to do with their game being played online, but with solution made specifically for some specific conditions and you are trying to extrapolate that to other game. :lol:
How do you think all these bots with 120ms Ping on Faceit or MM are even able to get out of silver.
In a Raw real world scenario without any assistance, players with 70+ ms have ZERO Chance. ZERO.

I am to lazy to rewrite:
viewtopic.php?f=10&t=9722&p=78169#p78213

Worstylolzs
Posts: 38
Joined: 23 Jan 2020, 10:28

Re: I finally found the reason behind my input lag after 6 years

Post by Worstylolzs » 14 Nov 2022, 18:52

Hello,
i have the same intel chip on my motherboard MSI 270 tomahawk. After read this post i try play apex legends from my phone like USB tethereing and i must say it looks like i see some little differencies. Is there any test method how i can confirm your findings? To be clear i have my phone connected to my home WIFI and then tethering with USB. I dont have any LTE connection on my phone, just WIFI. I thinking about buy this https://ventiontech.com/collections/hub ... et-adapter USB hub because he have realtek chip, so if i buy i will tell you differences.

But again, i dont have any real test method, so its only feeling testing...

Worstylolzs
Posts: 38
Joined: 23 Jan 2020, 10:28

Re: I finally found the reason behind my input lag after 6 years

Post by Worstylolzs » 15 Nov 2022, 14:28

6yToFindTheAnswer wrote:
07 Nov 2022, 16:50
Mims wrote:
02 Nov 2022, 17:41
I'm new to the forum, just wanted to reply to your post about the issue that you found with intel NICs.

I have the same scenario, moved out from one house to another, replaced all of my components including monitor, mouse, keyboard outside of my PC ... Recently redid my electrical box on my new appartment, felt some improvements but still had this input lag + desync with other players in multiplayer game, getting killed without even seeing nothing...

Tried with an USB Ethernet adapter with it own network card integrated, felt directly the improvement while playing valorant, fortnite ... The issue with taking invisible shots, and feeling like people i play against are godlike player was gone ! I was finally being able to take my time while gunfighting, seeing the bullets coming toward me and being able to adjust my movement while doing so. But still have some input lag (probably because the networking isn't processed directly from the motherboard).

I have a I225-V adapter, with a motherboard Z690-a prime gaming.
The 2 other motherboard that i replaced have exactly the same NIC, plus a third one having I219-V, with the same input lag / desync issue.
VERY interesting, so you also tested unplugging your ethernet cable from a motherboard with intel i219-V and directly felt an improvement in game and even on desktop, just like me.

Also :!: I have found a video of a russ.ian guy talking about that so we are definitely not the only ones.

Here is the video :

phpBB [video]


With the very bad auto-translation of youtube; what i understand is that : apparently he says that other players movements were "insane or weird" when he played with motherboards with intel network cards and when he switched to Realtek network cards their movements were a lot smoother and easier to kill. Also he noticed that with the intel network card when he ran into his teammates he was having weird "shakings movement or rollbacks" which is not the case with the realtek. If any of you guys understand Russlan could you please investigate this and what he is saying :?: cuz with youtube auto translate it is really hard to understand.

I think i am going to buy a few PCIe to ethernet cards with a Realtek chip to test this on my own and report back to you guys.
Can you please try this two optimazing video if it help? I try both and in my feeling testing it's better.
https://www.youtube.com/watch?v=kzkxbWXUXc8 trying first. Was better than default.
https://www.youtube.com/watch?v=MsTGC0IHH8E Only take settings from start of video. In my opinion is better than first.

Post Reply