trashso until it's figured out what breaks comanglia's config, default is actually the way to go, I guess
when ive been experimenting for a couple of hours with dozens of cvars benchmarking that dustbowl demo i found out that r_rootlod 0 and r_lod -1
gave me 2-4 more fps than r_rootlod 2 and r_lod 2 dont know if its placebo or not, but ive been using r_rootlod 0 and r_lod -1 since then, have a shot
e: forgot to mention that both r_rootlod 0 and r_lod -1 are game default btw
[quote=trash]so until it's figured out what breaks comanglia's config, default is actually the way to go, I guess[/quote]
when ive been experimenting for a couple of hours with dozens of cvars benchmarking that dustbowl demo i found out that r_rootlod 0 and r_lod -1
gave me 2-4 more fps than r_rootlod 2 and r_lod 2 dont know if its placebo or not, but ive been using r_rootlod 0 and r_lod -1 since then, have a shot
e: forgot to mention that both r_rootlod 0 and r_lod -1 are game default btw
done w/timedemo benchmark1
dx8
comanglia's highfps config (w/some edits)
2639 frames 25.222 seconds 104.63 fps ( 9.56 ms/f) 8.886 fps variability
-default with fps commands cut out of autoexec
2639 frames 27.514 seconds 95.92 fps (10.43 ms/f) 8.299 fps variability
dx9
comanglia
2639 frames 25.308 seconds 104.28 fps ( 9.59 ms/f) 10.732 fps variability
-default
2639 frames 31.602 seconds 83.51 fps (11.97 ms/f) 9.139 fps variability
i5-3570k
radeon hd 6770
6gb ram
done w/timedemo benchmark1
dx8
comanglia's highfps config (w/some edits)
2639 frames 25.222 seconds 104.63 fps ( 9.56 ms/f) 8.886 fps variability
-default with fps commands cut out of autoexec
2639 frames 27.514 seconds 95.92 fps (10.43 ms/f) 8.299 fps variability
dx9
comanglia
2639 frames 25.308 seconds 104.28 fps ( 9.59 ms/f) 10.732 fps variability
-default
2639 frames 31.602 seconds 83.51 fps (11.97 ms/f) 9.139 fps variability
i5-3570k
radeon hd 6770
6gb ram
i would also suggest doing timedemos with a demo of people rocking unusuals and random cosmetics
edit.
unusuals and that cart/player glow EAT fps, you probably should consider them when testing. I know timedemo stats are good, but that benchmark dem is dated
i would also suggest doing timedemos with a demo of people rocking unusuals and random cosmetics
edit.
unusuals and that cart/player glow EAT fps, you probably should consider them when testing. I know timedemo stats are good, but that benchmark dem is dated
Disabling steam beta shit alone made my fps go from 20-60 to 50-100 in pubs.
Disabling steam beta shit alone made my fps go from 20-60 to 50-100 in pubs.
after updating windows (lol) and restarting
dx8 comanglia
2639 frames 20.511 seconds 128.66 fps ( 7.77 ms/f) 8.969 fps variability
which is more or less what i had before tough break
so if you're still ahving problems after opting out of beta and trying -default...update windows?
i still get shit fps on the new maps tho
after updating windows (lol) and restarting
dx8 comanglia
2639 frames 20.511 seconds 128.66 fps ( 7.77 ms/f) 8.969 fps variability
which is more or less what i had before tough break
so if you're still ahving problems after opting out of beta and trying -default...update windows?
i still get shit fps on the new maps tho
Increasing my graphics settings to the max ingame without having -default on has no effect on my framerate in the benchmark, but turning -default on made the benchmark run in like slow motion at lower fps? Is anyone else experiencing this? I have a gtx 970 and an amd fx-8370.
Increasing my graphics settings to the max ingame without having -default on has no effect on my framerate in the benchmark, but turning -default on made the benchmark run in like slow motion at lower fps? Is anyone else experiencing this? I have a gtx 970 and an amd fx-8370.
Its pretty good, turned off ragdolls and gibs though
Its pretty good, turned off ragdolls and gibs though
I launched the game with -default on dx90 and these are some of the settings that are in config.cfg after using host_writeconfig full
Show Content
mat_queue_mode -1
cl_threaded_bone_setup 0
mat_forcehardwaresync 0
mat_forcemanagedtextureintohardware 0
r_queued_decals 0
r_queued_ropes 1
r_queued_post_processing 0
r_threaded_client_shadow_manager 0
r_threaded_particles 1
r_threaded_renderables 0
mod_load_anims_async 0
mod_load_mesh_async 0
mod_load_vcollide_async 0
I have a GTX 760 + i5 4670k @4.2ghz. Shouldnt the queue/threaded settings all be set to 1 for a multicore optimized setup? (Except mat_queue_mode).
I also looked a bit at forcehardwaresync, forcemanagedtextureintohardware and the three mod_load_x_async but didnt get any wiser from that. Could anyone explain those commands and what values would work best?
Thank you.
edit: typo
I launched the game with -default on dx90 and these are [b]some of the[/b] settings that are in config.cfg after using host_writeconfig full
[spoiler]mat_queue_mode -1
cl_threaded_bone_setup 0
mat_forcehardwaresync 0
mat_forcemanagedtextureintohardware 0
r_queued_decals 0
r_queued_ropes 1
r_queued_post_processing 0
r_threaded_client_shadow_manager 0
r_threaded_particles 1
r_threaded_renderables 0
mod_load_anims_async 0
mod_load_mesh_async 0
mod_load_vcollide_async 0[/spoiler]
I have a GTX 760 + i5 4670k @4.2ghz. Shouldnt the queue/threaded settings all be set to 1 for a multicore optimized setup? (Except mat_queue_mode).
I also looked a bit at forcehardwaresync, forcemanagedtextureintohardware and the three mod_load_x_async but didnt get any wiser from that. Could anyone explain those commands and what values would work best?
Thank you.
edit: typo
martenmy net_graph says im on around 50 on a pub, but it felt so smooth. How is that possible?
There's a difference between frame times and frame rate.
Frame rate is generally the average number of frames rendered per second. However, each frame might be slightly offset from the others. In an extreme case, let's say the first 50 frames take 10ms each. That's half of a second, and you've rendered 50 frames. However, then suddenly your CPU is heavily taxed, and the next 10 frames each take 50ms to render. You've rendered 60 frames in a full second, and the frame rate counter will display 60fps, but because of the huge gap in frame times, you're going to notice stutter. It won't look smooth.
My guess is that all of these old broken cvars caused horrible stuttering and unnecessary CPU stress, which in turn would increase frame time variance, decreasing smoothness.
[quote=marten]my net_graph says im on around 50 on a pub, but it felt so smooth. How is that possible?[/quote]
There's a difference between frame times and frame rate.
Frame rate is generally the average number of frames rendered per second. However, each frame might be slightly offset from the others. In an extreme case, let's say the first 50 frames take 10ms each. That's half of a second, and you've rendered 50 frames. However, then suddenly your CPU is heavily taxed, and the next 10 frames each take 50ms to render. You've rendered 60 frames in a full second, and the frame rate counter will display 60fps, but because of the huge gap in frame times, you're going to notice stutter. It won't look smooth.
My guess is that all of these old broken cvars caused horrible stuttering and unnecessary CPU stress, which in turn would increase frame time variance, decreasing smoothness.
i5 760, gtx470 1920x917
dx 8.1 comanglia's config
2639 frames 21.553 seconds 122.44 fps ( 8.17 ms/f) 12.265 fps variability
2639 frames 20.536 seconds 128.51 fps ( 7.78 ms/f) 9.999 fps variability
2639 frames 20.364 seconds 129.59 fps ( 7.72 ms/f) 9.974 fps variability
2639 frames 20.656 seconds 127.76 fps ( 7.83 ms/f) 10.051 fps variability
2639 frames 20.416 seconds 129.26 fps ( 7.74 ms/f) 9.681 fps variability
-default dx9.0 valve
2639 frames 37.172 seconds 70.99 fps (14.09 ms/f) 5.633 fps variability
2639 frames 35.754 seconds 73.81 fps (13.55 ms/f) 4.711 fps variability
2639 frames 35.566 seconds 74.20 fps (13.48 ms/f) 4.668 fps variability
2639 frames 35.836 seconds 73.64 fps (13.58 ms/f) 4.547 fps variability
2639 frames 35.899 seconds 73.51 fps (13.60 ms/f) 4.751 fps variability
altho ragdolls and other stuff is obviously having an impact on the default
i5 760, gtx470 1920x917
dx 8.1 comanglia's config
2639 frames 21.553 seconds 122.44 fps ( 8.17 ms/f) 12.265 fps variability
2639 frames 20.536 seconds 128.51 fps ( 7.78 ms/f) 9.999 fps variability
2639 frames 20.364 seconds 129.59 fps ( 7.72 ms/f) 9.974 fps variability
2639 frames 20.656 seconds 127.76 fps ( 7.83 ms/f) 10.051 fps variability
2639 frames 20.416 seconds 129.26 fps ( 7.74 ms/f) 9.681 fps variability
-default dx9.0 valve
2639 frames 37.172 seconds 70.99 fps (14.09 ms/f) 5.633 fps variability
2639 frames 35.754 seconds 73.81 fps (13.55 ms/f) 4.711 fps variability
2639 frames 35.566 seconds 74.20 fps (13.48 ms/f) 4.668 fps variability
2639 frames 35.836 seconds 73.64 fps (13.58 ms/f) 4.547 fps variability
2639 frames 35.899 seconds 73.51 fps (13.60 ms/f) 4.751 fps variability
altho ragdolls and other stuff is obviously having an impact on the default
flatlineyea i ran -default and none of my settings changed so rip
Got the same thing too. All my graphics settings stayed at what they were at before, even after turning off Steam Cloud sync, and deleting both the custom cfg folder, and the regular one.
[quote=flatline]yea i ran -default and none of my settings changed so rip[/quote]
Got the same thing too. All my graphics settings stayed at what they were at before, even after turning off Steam Cloud sync, and deleting both the custom cfg folder, and the regular one.
Here's what it booted up with:
http://images.akamai.steamusercontent.com/ugc/392175615064974667/E4E1B0C2F1C3A74B65FE00C51D8D9B421A760AC5/
And here's the actual recommended settings that the game tells me to use (notice the *s)
http://images.akamai.steamusercontent.com/ugc/392175615064977077/38645D7BD55DB0D7BBEA9B5CF2AB2088BB532981/
Should I just use the stuff that it booted me up with? Also I use anti aliasing and anisotropic filtering through nvidia control panel is it better to use it within TF2 or nvidia control panel?
Here's what it booted up with:
[img]http://images.akamai.steamusercontent.com/ugc/392175615064974667/E4E1B0C2F1C3A74B65FE00C51D8D9B421A760AC5/[/img]
And here's the actual recommended settings that the game tells me to use (notice the *s)
[img]http://images.akamai.steamusercontent.com/ugc/392175615064977077/38645D7BD55DB0D7BBEA9B5CF2AB2088BB532981/[/img]
Should I just use the stuff that it booted me up with? Also I use anti aliasing and anisotropic filtering through nvidia control panel is it better to use it within TF2 or nvidia control panel?
thx flame my game looks fat af now mmmm
[url=https://www.youtube.com/watch?v=Vwr6OuQ8dJo]https://www.youtube.com/watch?v=Vwr6OuQ8dJo[/url]
Sadly didn't help :/ i5-4460/R9 280X oc
Also still no idea why I lost ~30-40 fps with this update
Sadly didn't help :/ i5-4460/R9 280X oc
Also still no idea why I lost ~30-40 fps with this update
-default changes my game and music volumes. how to fix?
-default changes my game and music volumes. how to fix?
lol_goat-default changes my game and music volumes. how to fix?
change your game and music volume
[quote=lol_goat]-default changes my game and music volumes. how to fix?[/quote]
change your game and music volume
Finally someone who gets it :D
Increasing quality for the fat graphics, shiny particles and silly outlines Valve is forcing on you will increase your fps by stressing that beast of a GPU you have been keeping on a leash by redirecting every bit of load to your CPU by running potato settings because you came from a toaster.
FlameConfig-default when?! :D
Finally someone who gets it :D
Increasing quality for the fat graphics, shiny particles and silly outlines Valve is forcing on you will increase your fps by stressing that beast of a GPU you have been keeping on a leash by redirecting every bit of load to your CPU by running potato settings because you came from a toaster.
FlameConfig-default when?! :D
try setting:
r_lod -1
r_rootlod 2
mat_picmip 2
in comanglias cfg, for more fps
try setting:
r_lod -1
r_rootlod 2
mat_picmip 2
in comanglias cfg, for more fps
My second reason for using fps configs was that it generally reduced visual clutter/made stuff easier to distinguish.
Stuff like pyro flames and explosions. Do you have any screenshots or vids?
This sounds p cool
My second reason for using fps configs was that it generally reduced visual clutter/made stuff easier to distinguish.
Stuff like pyro flames and explosions. Do you have any screenshots or vids?
This sounds p cool
I recall people mentioning that -dxlevel 98 is a bit more graphics-balancing a while back. I got about the same framerate as 91, but I have a pretty mid-range graphics card, so it might work out better for others.
it would help to find out what commands specifically offload to the CPU, if this is the case as people are assuming
I recall people mentioning that -dxlevel 98 is a bit more graphics-balancing a while back. I got [url=http://www.teamfortress.tv/post/525138/tf2-benchmarks]about the same framerate as 91[/url], but I have a pretty mid-range graphics card, so it might work out better for others.
it would help to find out what commands specifically offload to the CPU, if this is the case as people are assuming
the steam beta was really effecting my game, I got a LOT of frames back after getting out of it
the steam beta was really effecting my game, I got a LOT of frames back after getting out of it
It looks like most people (including me) get more frames from using comanglia's config rather than -default
It looks like most people (including me) get more frames from using comanglia's config rather than -default
EmmaWatsonIt looks like most people (including me) get more frames from using comanglia's config rather than -default
flame ...BUT I have a CRAZY idea that works for me and will probably be worth trying for people with higher end cards and processors (im on a 970 and an oc'ed i7 4770k @4ghz)...
...If your fps goes up, great, if it doesn't oh well. I went from 60-70 in pubs to I kid you not 200+ without dips just by maxing everything...
...Idk, my game runs like cinematic butter. I'm in no way telling you this is going to work, or that your fps config is bad, I'm simply suggesting you try it since it 100% worked for me.
[quote=EmmaWatson]It looks like most people (including me) get more frames from using comanglia's config rather than -default[/quote]
[quote=flame] ...BUT I have a CRAZY idea that [b]works for me[/b] and will probably be worth trying for people with [b]higher end cards and processors[/b] (im on a 970 and an oc'ed i7 4770k @4ghz)...
...[b]If your fps goes up, great, if it doesn't oh well[/b]. I went from 60-70 in pubs to I kid you not 200+ without dips just by maxing everything...
...Idk, my game runs like cinematic butter. [b]I'm in no way telling you this is going to work[/b], or that your fps config is bad, [b]I'm simply suggesting[/b] you try it since it 100% worked for [b]me[/b].[/quote]
I used to get 180 fps in timedemo benchmark1, after the update my frames were less and it felt really laggy. So i tried this, on dx90 i got 126 fps and on dx98 i got 130fps, but i tried it in a lobby and it was way smoother. So gonna try use this for now :) (on dx98)
Also tf2 looks so good with this :o
I used to get 180 fps in timedemo benchmark1, after the update my frames were less and it felt really laggy. So i tried this, on dx90 i got 126 fps and on dx98 i got 130fps, but i tried it in a lobby and it was way smoother. So gonna try use this for now :) (on dx98)
Also tf2 looks so good with this :o
For some reason, my class cfgs dont want to execute automatically. Help?
EDIT: They do, but my interp commands don't change
For some reason, my class cfgs dont want to execute automatically. Help?
EDIT: They do, but my interp commands don't change
what needs to happen now is a mix of -default and comanglia's cfg...stressing gpu, combined with whatever you can get to squeeze those frames...i'm sure some cvars -default changed could get even more fps when used a different value...
i'm getting 131fps capped thou, fps_max at 132....it's pretty constant, no dips, which is cool because source is influenced so much by frames and frame fluctuation really fucks up with aim, but comanglia says that a fps constantly reaching it's cap gives input lag....so far i have to notice it, but is it really true? should i just use fps_max 0 and have fluctuation rather than getting 131 constant and dealing with this input lag?
what needs to happen now is a mix of -default and comanglia's cfg...stressing gpu, combined with whatever you can get to squeeze those frames...i'm sure some cvars -default changed could get even more fps when used a different value...
i'm getting 131fps capped thou, fps_max at 132....it's pretty constant, no dips, which is cool because source is influenced so much by frames and frame fluctuation really fucks up with aim, but comanglia says that a fps constantly reaching it's cap gives input lag....so far i have to notice it, but is it really true? should i just use fps_max 0 and have fluctuation rather than getting 131 constant and dealing with this input lag?
Quertwhat needs to happen now is a mix of -default and comanglia's cfg...stressing gpu, combined with whatever you can get to squeeze those frames...i'm sure some cvars -default changed could get even more fps when used a different value...
i'm getting 131fps capped thou, fps_max at 132....it's pretty constant, no dips, which is cool because source is influenced so much by frames and frame fluctuation really fucks up with aim, but comanglia says that a fps constantly reaching it's cap gives input lag....so far i have to notice it, but is it really true? should i just use fps_max 0 and have fluctuation rather than getting 131 constant and dealing with this input lag?
I think you should put it atleast higher to like 300, its still not confirmed but your game feels smoother with more fps even if you have 144hz
[quote=Quert]what needs to happen now is a mix of -default and comanglia's cfg...stressing gpu, combined with whatever you can get to squeeze those frames...i'm sure some cvars -default changed could get even more fps when used a different value...
i'm getting 131fps capped thou, fps_max at 132....it's pretty constant, no dips, which is cool because source is influenced so much by frames and frame fluctuation really fucks up with aim, but comanglia says that a fps constantly reaching it's cap gives input lag....so far i have to notice it, but is it really true? should i just use fps_max 0 and have fluctuation rather than getting 131 constant and dealing with this input lag?[/quote]
I think you should put it atleast higher to like 300, its still not confirmed but your game feels smoother with more fps even if you have 144hz
Excuse my ignorance, but how do you exec your autoexec?
Excuse my ignorance, but how do you exec your autoexec?
MaclinExcuse my ignorance, but how do you exec your autoexec?
some ancient technique called "launch the game"
although if youve changed a few values in autoexec.cfg while in the game you can exec it manually in the console simply by pasting: exec autoexec
typo
[quote=Maclin]Excuse my ignorance, but how do you exec your autoexec?[/quote]
some ancient technique called "launch the game"
although if youve changed a few values in autoexec.cfg while in the game you can exec it manually in the console simply by pasting: exec autoexec
typo