fps and etpro
Moderators: Forum moderators, developers
fps and etpro
Hey guys, i'm just posting this to ask whether anything significant in terms of graphics or whatever has been added to etpro, i am having big problems with fps and stability.
I only get it on etpro servers more so 3.2.0 + , tonight i did a fresh install of ET, ran the desktop shortcut made a standard server on goldrush and got 125 fps stable in 1 position, i closed down ET and loaded with the latest 3.2.2 loaded via +set fs_game etpro , and this time my fps is halved and i can't manage 100 in the same places. I realize ET is cpu intensive so a few months ago i would sit down and say it is but playing BF2 at 100fps and CS source at 60+ makes me think whether something is wrong with ET. Could multiple ETpro installations conflict with others?
Cheers for any help
System spec-
AMD 3000+ stock speeds
GF 6800XT XFX 128MB AGP (81.94)
Asus NForce2 A7N8X Deluxe
1GB x2 512mb Geil 5-3-3-2.5T
120GB SATA HDD
Windows XP Home SP2
I only get it on etpro servers more so 3.2.0 + , tonight i did a fresh install of ET, ran the desktop shortcut made a standard server on goldrush and got 125 fps stable in 1 position, i closed down ET and loaded with the latest 3.2.2 loaded via +set fs_game etpro , and this time my fps is halved and i can't manage 100 in the same places. I realize ET is cpu intensive so a few months ago i would sit down and say it is but playing BF2 at 100fps and CS source at 60+ makes me think whether something is wrong with ET. Could multiple ETpro installations conflict with others?
Cheers for any help
System spec-
AMD 3000+ stock speeds
GF 6800XT XFX 128MB AGP (81.94)
Asus NForce2 A7N8X Deluxe
1GB x2 512mb Geil 5-3-3-2.5T
120GB SATA HDD
Windows XP Home SP2
If you are hosting locally, you should only have 1 version of etpro. However, this shouldn't cause FPS problems.
A number of other people have reported FPS problems after 3.2.0
http://forums-center.com/viewtopic.php?t=7389
As far as I can tell, you are the first person to actually do a worthwhile test. You might also want to try timedemoing the two versions, although this only gives you a part of the picture. If you post a screenshot of the location you used for testing, others can try do comparable tests.
Note that the PB issue I mention in the above thread would only cause very brief drops. It also shouldn't be present if you don't use a server config that has a bunch of cvar, md5 checks and autoscreenshots.
A number of other people have reported FPS problems after 3.2.0
http://forums-center.com/viewtopic.php?t=7389
As far as I can tell, you are the first person to actually do a worthwhile test. You might also want to try timedemoing the two versions, although this only gives you a part of the picture. If you post a screenshot of the location you used for testing, others can try do comparable tests.
Note that the PB issue I mention in the above thread would only cause very brief drops. It also shouldn't be present if you don't use a server config that has a bunch of cvar, md5 checks and autoscreenshots.
send lawyers, guns and money
Hi just done all that and here is the result, all however on a local server, the demo was taken with a standard config no cvar editing
etmain
etpro 3.2.2
Doesn't seem much difference but on a server with airstrikes going off etc maybe it could be more.. I'm not sure what to think of it lol, try it yourselves. My problems are more FPS on a whole map, i don't get lags or drops just a solid drop to like 40 fps from a stable 125. Maps i once could run around on 125 i can barely manage 40 on now. Oh and btw i had PB enabled during the recording of the demo... Would it be worth testing with PB disabled?
Cheers
Demo is http://www.serialkilla.pwp.blueyonder.c ... demo.dm_84
etmain
etpro 3.2.2
Doesn't seem much difference but on a server with airstrikes going off etc maybe it could be more.. I'm not sure what to think of it lol, try it yourselves. My problems are more FPS on a whole map, i don't get lags or drops just a solid drop to like 40 fps from a stable 125. Maps i once could run around on 125 i can barely manage 40 on now. Oh and btw i had PB enabled during the recording of the demo... Would it be worth testing with PB disabled?
Cheers
Demo is http://www.serialkilla.pwp.blueyonder.c ... demo.dm_84
k. first i did /cvar_restart and /vid_restart so i was testing with a default config. here's what i got with your demo:Tragedy wrote:try it yourselves
etmain - 1204 frames, 6.4 seconds: 187.3 fps
etpro - 1204 frames, 6.6 seconds: 183.1 fps
pretty insignificant difference on my machine, which is:
athlon 64 3000+ @ 1935
6800 ultra @ 425/1190
I would assume the fps drops aren't a result of clientside code (since most cputime is used for rendering) but rather a sync with the server.
Etpro dev has cleaned a lot of the code, a lot useless data was saved from being sent between client-server and even numerous calculations were being improved. (you gotta play with the code to understand how serious those things are, imo the power of etpro is with it's code improvement and bug fixes, rather than the various cvars which are simple for anyone to code), on one hand this should improve network capabilities, but remember some stuff was removed serverside (like hitbox checking etc), thus the player relays more on the server for calculations and in turn recieving the packets and syncing the client.
Though i gotta admit i doubt serverside calculations could be avoided.
Etpro dev has cleaned a lot of the code, a lot useless data was saved from being sent between client-server and even numerous calculations were being improved. (you gotta play with the code to understand how serious those things are, imo the power of etpro is with it's code improvement and bug fixes, rather than the various cvars which are simple for anyone to code), on one hand this should improve network capabilities, but remember some stuff was removed serverside (like hitbox checking etc), thus the player relays more on the server for calculations and in turn recieving the packets and syncing the client.
Though i gotta admit i doubt serverside calculations could be avoided.
this reminds me of a question i had. everybody remembers the bug in etmain 2.55 where vsay could boot everybody from the server. well, there's a somewhat notorious clan right now who finds it funny to spam ungodly amounts of text and vsays. and although it no longer boots people, it can cause some lesser servers to get a bit laggy. not to mention it's annoying - and league configs remove all limits on those things so it's free reign for asshats. also, having keys bound to each team, like everyone does, and rapidly tapping them to jump back and forth at blistering speeds (between rounds, etc), can actually cause problems on some servers. i've seen it myself. almost as if it was too much information to handle.
now the question: i assume that even if i have cg_novoicechats 1, the commands are still sent to me, but not played, correct? if that's the case, what about making it so my preference tells the server i don't even want that data to be sent? same for text chats - although they can't be disabled entirely, perhaps cg_teamchatheight 0 (which would effectively be disabling them) could tell the server to buzz off with that info, i don't want it.
might save some bandwidth too
now the question: i assume that even if i have cg_novoicechats 1, the commands are still sent to me, but not played, correct? if that's the case, what about making it so my preference tells the server i don't even want that data to be sent? same for text chats - although they can't be disabled entirely, perhaps cg_teamchatheight 0 (which would effectively be disabling them) could tell the server to buzz off with that info, i don't want it.
might save some bandwidth too
afaik, there's a spamprotection filter available serverside (too lazy to double check it), but i remember quite a few matches where someone tried to spam ascii art, and the server blocked it after a few lines resulting in only a part of the ascii logo shown, you might look it up if you have your own server.
But that's kinda offtopic, since the original poster was reffering to fps drop on usual gameplay
But that's kinda offtopic, since the original poster was reffering to fps drop on usual gameplay
Ok after a little more testing i've figured it isn't an ETpro problem as i have been getting this on some ETmain servers now. It seems any server with over 14 people is when the fps begins to drop. It has to be player models or something server-side since if i join a server on ETmain on battery with 12 players + my fps is between 45-70 outside were the allies spawn, but on a locally hosted server i get 125 stable in the exact same places.
Trying other things i did a Windows XP repair to SP1 and the fps was similar so it isn't exactly a Windows issue. Maybe its just wear and tear on my CPU over time.. i don't know but hopefully with a new PC coming for christmas i'll have to bare with it..
Trying other things i did a Windows XP repair to SP1 and the fps was similar so it isn't exactly a Windows issue. Maybe its just wear and tear on my CPU over time.. i don't know but hopefully with a new PC coming for christmas i'll have to bare with it..
Me and a friend crashed a baserace server by calling votes to kick each other which kept failing, took afew mins then it evenutally crashed.ouroboro wrote:this reminds me of a question i had. everybody remembers the bug in etmain 2.55 where vsay could boot everybody from the server. well, there's a somewhat notorious clan right now who finds it funny to spam ungodly amounts of text and vsays. and although it no longer boots people, it can cause some lesser servers to get a bit laggy. not to mention it's annoying - and league configs remove all limits on those things so it's free reign for asshats. also, having keys bound to each team, like everyone does, and rapidly tapping them to jump back and forth at blistering speeds (between rounds, etc), can actually cause problems on some servers. i've seen it myself. almost as if it was too much information to handle.
now the question: i assume that even if i have cg_novoicechats 1, the commands are still sent to me, but not played, correct? if that's the case, what about making it so my preference tells the server i don't even want that data to be sent? same for text chats - although they can't be disabled entirely, perhaps cg_teamchatheight 0 (which would effectively be disabling them) could tell the server to buzz off with that info, i don't want it.
might save some bandwidth too
Notice something here, on crowded servers try going to an area far away from everyone, if you'll notice your fps is still low - this is because the lag is caused from the network sync.Tragedy wrote:Ok after a little more testing i've figured it isn't an ETpro problem as i have been getting this on some ETmain servers now. It seems any server with over 14 people is when the fps begins to drop. It has to be player models or something server-side since if i join a server on ETmain on battery with 12 players + my fps is between 45-70 outside were the allies spawn, but on a locally hosted server i get 125 stable in the exact same places.
Trying other things i did a Windows XP repair to SP1 and the fps was similar so it isn't exactly a Windows issue. Maybe its just wear and tear on my CPU over time.. i don't know but hopefully with a new PC coming for christmas i'll have to bare with it..
however if your fps will get back to normal, this is clientside fps hit caused from more player models.
The problem seems to be in certain places on certain maps, for example on Goldrush the 2nd spawn for axis around the bridge and the courtyard is killer for me, and anywhere around Venice is horrible on FPS, imo the problem really started at 2.60 thats the furthest i can remember having the problem. It can't be Opengl or that kinda thing related since i can manage 100fps stable on a CS1.6 server, even CS source runs around 50-60 ok.. Just ET doesn't like something..Svartberg wrote: Notice something here, on crowded servers try going to an area far away from everyone, if you'll notice your fps is still low - this is because the lag is caused from the network sync.
however if your fps will get back to normal, this is clientside fps hit caused from more player models.
I can remember pre ETpro-3.1.9 before all those compatiability issues with ETpatch 2.60 when i could play stable 76-125 on a 5700FX, now around 60fps seems to be my fps now on the same maps with a graphics card twice as powerful. Others have reported issues.. I've read it could possibly be PB, but can PB do so much to lower fps? I thought PB at most could cause lag on your ping since its sending screenshot data or whatever but not on your FPS unless PB is requesting a lot of screenshots...
http://www.etnation.co.uk/index.php?nam ... pic&t=1881