Re: sv_clockcorrection_msecs

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

Comberbach .
Can someone explain what sv_clockcorrection_msecs does?

Cheers,

Phantom
 
> Why even use "100" 64 is enough or if you think you need more why not go straight to 128. 64 tick with sv_clockcorrection_msecs 15 "feels" like 128. Personally I use sv_clockcorrection_msecs 0 which also works fine if every player has decent jitter
>


_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

Wareya
IIRC it restricts how far out of sync the server allows clients to go from the time which the server thinks the client is supposed to be at.
On 2014/12/19 8:59, Comberbach . wrote:
Can someone explain what sv_clockcorrection_msecs does?

Cheers,

Phantom
 
> Why even use "100" 64 is enough or if you think you need more why not go straight to 128. 64 tick with sv_clockcorrection_msecs 15 "feels" like 128. Personally I use sv_clockcorrection_msecs 0 which also works fine if every player has decent jitter
>



_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers


_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

The H4x0r
In reply to this post by Comberbach .
It's "time travel" used for jitter (ping variance) correction and is what causes among other things the "getting killed behind corners". 128 tick halves it's effects so manually setting it to 15 (instead of the default 30) has mostly the same visible effects as using 128 tick though keep in mind that many 128 tick servers suchs as CEVO's or ESEA's use clock correction 15 and 128 tick which would equal to clock correction 7.5 on 64 tick but the diffrence between 7.5 and 15 isn't really noticeable anyway. Overall it's quite complicated but 15 is considered to be "optimal" reducing the negative effects clock correction has on everyone while still providing adequate jitter correction for people with poor connections.


From: [hidden email]
To: [hidden email]
Date: Fri, 19 Dec 2014 13:59:34 +0000
Subject: Re: [Csgo_servers] sv_clockcorrection_msecs

Can someone explain what sv_clockcorrection_msecs does?

Cheers,

Phantom
 
> Why even use "100" 64 is enough or if you think you need more why not go straight to 128. 64 tick with sv_clockcorrection_msecs 15 "feels" like 128. Personally I use sv_clockcorrection_msecs 0 which also works fine if every player has decent jitter
>


_______________________________________________ Csgo_servers mailing list [hidden email] https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers

_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

Wareya
 >and is what causes among other things the "getting killed behind corners".

You w0t? That's caused by input prediction on your end and lag
compensation on the other person's. clockcorrection will just restrict
how much their lag compensation varies. If they have 250 ping you're
still going to get killed from behind corners after you get into cover.

 >128 tick halves it's effects

I don't understand. It's counted in milliseconds, and I can't see how
valve would mess up a time to tick conversion.

_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

The H4x0r
The reason 128 tick halves clock corrections negative effects is because there's twice as many frames. Ofcourse clock correction alone doesn't cause the getting killed behind corners but it's what allows it to happen. Less clock correction = less forgiving lag compensation for clients with crappy connections. Also with getting killed behind corners I specifically meant clients with good connections and low ping who should suffer from it, clients with bad connections are going to suffer either way but lowering clock correction lessens the negative effects on clients with proper connections (at the expense of high ping/high jitter players)

> Date: Fri, 19 Dec 2014 20:53:51 -0500

> From: [hidden email]
> To: [hidden email]
> Subject: Re: [Csgo_servers] sv_clockcorrection_msecs
>
> >and is what causes among other things the "getting killed behind corners".
>
> You w0t? That's caused by input prediction on your end and lag
> compensation on the other person's. clockcorrection will just restrict
> how much their lag compensation varies. If they have 250 ping you're
> still going to get killed from behind corners after you get into cover.
>
> >128 tick halves it's effects
>
> I don't understand. It's counted in milliseconds, and I can't see how
> valve would mess up a time to tick conversion.
>
> _______________________________________________
> Csgo_servers mailing list
> [hidden email]
> https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers

_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

The H4x0r
low ping who shouldn't suffer from it*


From: [hidden email]
To: [hidden email]
Date: Mon, 22 Dec 2014 11:15:23 +0200
Subject: Re: [Csgo_servers] sv_clockcorrection_msecs

The reason 128 tick halves clock corrections negative effects is because there's twice as many frames. Ofcourse clock correction alone doesn't cause the getting killed behind corners but it's what allows it to happen. Less clock correction = less forgiving lag compensation for clients with crappy connections. Also with getting killed behind corners I specifically meant clients with good connections and low ping who should suffer from it, clients with bad connections are going to suffer either way but lowering clock correction lessens the negative effects on clients with proper connections (at the expense of high ping/high jitter players)

> Date: Fri, 19 Dec 2014 20:53:51 -0500

> From: [hidden email]
> To: [hidden email]
> Subject: Re: [Csgo_servers] sv_clockcorrection_msecs
>
> >and is what causes among other things the "getting killed behind corners".
>
> You w0t? That's caused by input prediction on your end and lag
> compensation on the other person's. clockcorrection will just restrict
> how much their lag compensation varies. If they have 250 ping you're
> still going to get killed from behind corners after you get into cover.
>
> >128 tick halves it's effects
>
> I don't understand. It's counted in milliseconds, and I can't see how
> valve would mess up a time to tick conversion.
>
> _______________________________________________
> Csgo_servers mailing list
> [hidden email]
> https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers

_______________________________________________ Csgo_servers mailing list [hidden email] https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers

_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers
Reply | Threaded
Open this post in threaded view
|

Re: sv_clockcorrection_msecs

Wareya
In reply to this post by The H4x0r
 >The reason 128 tick halves clock corrections negative effects is
because there's twice as many frames.
That doesn't make any sense, at all. Clock correction is measured in
milliseconds. It will apply to twice as many frames on 128tick, giving
the same amount of time shift allowance.

 >Ofcourse clock correction alone doesn't cause the getting killed
behind corners but it's what allows it to happen
No, this is just wrong. All clock correction does is let someone's ping
go higher for a moment. It doesn't "restrict" how much lag compensation
is theoretically possible. Someone with a perfectly stable 200 ping will
still be able to kill you around corners you just went behind, and if
you have a stable 200 ping, you will also be killed around corners you
just went behind. It lets it happen in slightly more situations, but
that's a good thing. Getting killed around corners is literally just an
edge case of having good hit registration.

 >Less clock correction = less forgiving lag compensation for clients
with crappy connections.
Only in terms of jitter, but again, they can still kill you around
corners. They just have to shoot *incorrectly* instead. It makes it even
more RNG.

 >Also with getting killed behind corners I specifically meant clients
with good connections and low ping who should suffer from it
Has literally nothing to do with ping and everything to do with jitter.

 > clients with bad connections are going to suffer either way but
lowering clock correction lessens the negative effects on clients with
proper connections (at the expense of high ping/high jitter players)
People with bad connections are *always* suffering more than people with
good connections. You're just making it worse.

If you want a perfect game, play on LAN. Further, please stop spreading
misinformation.

On 2014/12/22 4:15, The H4x0r wrote:
> The reason 128 tick halves clock corrections negative effects is
> because there's twice as many frames.


_______________________________________________
Csgo_servers mailing list
[hidden email]
https://list.valvesoftware.com/cgi-bin/mailman/listinfo/csgo_servers