[Infntnub]

Just for the hell of it I asked a few people last night on xfire what fps they use while playing urt and 4 out the 6 people I asked said they never adjusted it from the default (com_maxfps 85 = which rounds off to 90 if your vid card can handle it).

I have always been under the assumption 125 was ideal since it was the highest allowable setting - but just for argument's sake, what setting do you use/prefer?

[mitsubishi]

the apparent reason it's 85 is that maxpackets is max 42 (currently) and if you drop below 42 or its double or its triple you won't get its highest latency advantage available. So if some are going to drop below 125 and above 85 they're gonna lose the packets.

i expect it to go to 125 in 4.2 since packets goes to 125.

[no, it's actually optimized for 125 and com_maxfps 85 caps it to really 90.9 (which loses that advantage by dropping client->server packets to 30), it should be com_maxfps 83 to get that advantage]

--

if all that is very confusing (it usually is) see http://www.www0.org/...packets_and_FPS

This post has been edited by mitsubishi: 17 June 2010 - 04:02 PM

[Infntnub]

So in some cases 85 is a better choice than 125.. that's surprising.

[mitsubishi]

not strictly even there; e.g. you may gain in networking latency but you lose on visual latency if it would be able to go above 85.

--

Actually I got that wrong here (unless 85 is replaced with 83). 42 (maxpackets) seems to be optimized for 125 (FPS) but when com_maxfps is 85 it effectively caps FPS to 90.9. I've no idea why it's set to 85 (FPS) (it caps client->server packets to 30/sec in that case).

--

Maybe it's a mistake, someone really wanting to do com_maxfps 83. With 85 it caps FPS to 90.9 and it loses the advantage by capping client->server packets to 30 (assuming it's steady).

--

To not re-write it, from that guide:

An unintuitive internal rounding up

Notice that the ceiling of the var is being rounded up internally in an unintuitive fashion: com_maxfps 84 goes to 90.090.. while 83 stays to 83.333...

i.e. com_maxfps 84 gives effectively 30.30.. packets per second since it puts effective max fps to 90.91.. while the correct value would be 83 to get 41.666.. since that is the com_maxfps value that will give an effective max fps of 83.333 which we'd want for lowering networking latency with the cl_maxpackets ceiling available.

[All that is confirmed by a 'draw client->server packets /second' meter I've cooked for a client]

This post has been edited by mitsubishi: 21 June 2010 - 09:27 AM

[rYuuJiN]

125

[BRAVO]

I use 90... because my lo-end computer always works on 60-80 depending on the map and how many players on server.

a very interesting data..., if you set max fps to 60.. the ci advice disappears, very useful if your ISP doesn't work well (like mine).

[mitsubishi]

you might want to do 83 per the reasons above.

[BRAVO]

Really good mitsu (like always) but my english is not really good, and what you are explaining seems to be very technical. i didn't understand.

i have to set my graphics the way having allways 83+ ? or i should set com_maxfps 83 to win those extra fps ?.

[Mod ObScUrE]

BRAVO, on 21 June 2010 - 02:38 AM, said:

i should set com_maxfps 83 to win those extra fps packets.

Yes when you can't get stable 125 fps the next lower step will be 83.333... to get the best network latency.

Quote

What cl_maxpackets does
cl_maxpackets governs the max amount of packets of information the client is willing to send to the server (client->server) a second. It is locked between 30 and 42 in UrT.
This is highly related to the FPS of the client.

90.9 : 3 = 30.3 packets
90 : 2 = 45 packets (This is not possible in 4.1 because maxpackets are locked down to 42).


83.3 : 2 = 41.666...
125 : 3 = 41.666...

[BRAVO]

lots of thanks mitsubishi and ObScUrE.

changing mi .cfg right now... :)