Upvote Upvoted 245 Downvote Downvoted
1 ⋅⋅ 25 26 27 28 29 30 31 ⋅⋅ 58
mastercomfig - fps/customization config
posted in Customization
811
#811
1 Frags +
pancake_stacksSo what's with the net_splitrate commands? I've read they're only for server-side and useless for clients. And is it tied to the rate command, as they're both the same in your cfgs.

I believe it works on clients, as clients still split packets if they're too big. Unfortunately, I don't have access to any source code that uses these split packet commands, so I can't be too sure about them.

The two rates for normal packets and splitpackets are the same for simplicity because rates are dependent on your Internet connection, and it is easier to adjust the rates as one value according to your upload speed, rather than mess around with trying to optimize the two rates individually for a lot of different upload speeds.

hi_How can I turn on the disguise animation for Spy? I looked through the config and don't see any commands related to enabling or disabling it.

You mean the white outline that appears on your screen? I believe that's controlled by your HUD.

DashieGasaiAwesome! I am using default 5.1.3 config on Nvidia 384.90 drivers.

Is there any console output related to medibeam/mediguns?

[quote=pancake_stacks]So what's with the net_splitrate commands? I've read they're only for server-side and useless for clients. And is it tied to the rate command, as they're both the same in your cfgs.[/quote]
I believe it works on clients, as clients still split packets if they're too big. Unfortunately, I don't have access to any source code that uses these split packet commands, so I can't be too sure about them.

The two rates for normal packets and splitpackets are the same for simplicity because rates are dependent on your Internet connection, and it is easier to adjust the rates as one value according to your upload speed, rather than mess around with trying to optimize the two rates individually for a lot of different upload speeds.

[quote=hi_]How can I turn on the disguise animation for Spy? I looked through the config and don't see any commands related to enabling or disabling it.[/quote]
You mean the white outline that appears on your screen? I believe that's controlled by your HUD.

[quote=DashieGasai]
Awesome! I am using default 5.1.3 config on Nvidia 384.90 drivers.[/quote]
Is there any console output related to medibeam/mediguns?
812
#812
0 Frags +

In-game bug looks like this

https://i.imgur.com/AfbW70y.jpg

I'll try to see if I can spot anything in console.

Nothing medigun-related in console as much as I can see.

In-game bug looks like this [img]https://i.imgur.com/AfbW70y.jpg[/img]

I'll try to see if I can spot anything in console.

Nothing medigun-related in console as much as I can see.
813
#813
SizzlingStats
0 Frags +
mastercomsI believe it works on clients, as clients still split packets if they're too big. Unfortunately, I don't have access to any source code that uses these split packet commands, so I can't be too sure about them.

Have you tried IDA Pro or something similar on the linux/osx binaries to find references to the splitpacket convars?

[quote=mastercoms]I believe it works on clients, as clients still split packets if they're too big. Unfortunately, I don't have access to any source code that uses these split packet commands, so I can't be too sure about them.[/quote]

Have you tried IDA Pro or something similar on the linux/osx binaries to find references to the splitpacket convars?
814
#814
1 Frags +
DashieGasaiIn-game bug looks like this https://i.imgur.com/AfbW70y.jpg

I'll try to see if I can spot anything in console.

Nothing medigun-related in console as much as I can see.

Hm, that's strange, though are you sure it has to be something with my config? It could be a TF2 bug.

SizzlingCalamariHave you tried IDA Pro or something similar on the linux/osx binaries to find references to the splitpacket convars?

I have done some debugging and a bit of decompiling but it honestly takes a long time to get anything out of it so I haven't looked at much stuff. Hopefully I'll get to figuring out what the splitpacket cvars do eventually.

[quote=DashieGasai]In-game bug looks like this [img]https://i.imgur.com/AfbW70y.jpg[/img]

I'll try to see if I can spot anything in console.

Nothing medigun-related in console as much as I can see.[/quote]
Hm, that's strange, though are you sure it has to be something with my config? It could be a TF2 bug.

[quote=SizzlingCalamari]Have you tried IDA Pro or something similar on the linux/osx binaries to find references to the splitpacket convars?[/quote]
I have done some debugging and a bit of decompiling but it honestly takes a long time to get anything out of it so I haven't looked at much stuff. Hopefully I'll get to figuring out what the splitpacket cvars do eventually.
815
#815
0 Frags +
mastercomspancake_stacksSo what's with the net_splitrate commands? I've read they're only for server-side and useless for clients. And is it tied to the rate command, as they're both the same in your cfgs.I believe it works on clients, as clients still split packets if they're too big. Unfortunately, I don't have access to any source code that uses these split packet commands, so I can't be too sure about them.

The two rates for normal packets and splitpackets are the same for simplicity because rates are dependent on your Internet connection, and it is easier to adjust the rates as one value according to your upload speed, rather than mess around with trying to optimize the two rates individually for a lot of different upload speeds.

Have you looked into testing with the net_showsplits command? I believe it dumps to console how it handles splits.

[quote=mastercoms][quote=pancake_stacks]So what's with the net_splitrate commands? I've read they're only for server-side and useless for clients. And is it tied to the rate command, as they're both the same in your cfgs.[/quote]
I believe it works on clients, as clients still split packets if they're too big. Unfortunately, I don't have access to any source code that uses these split packet commands, so I can't be too sure about them.

The two rates for normal packets and splitpackets are the same for simplicity because rates are dependent on your Internet connection, and it is easier to adjust the rates as one value according to your upload speed, rather than mess around with trying to optimize the two rates individually for a lot of different upload speeds. [/quote]

Have you looked into testing with the net_showsplits command? I believe it dumps to console how it handles splits.
816
#816
0 Frags +
pancake_stacksHave you looked into testing with the net_showsplits command? I believe it dumps to console how it handles splits.

IIRC I did, but it was too spammy to give any useful information. Could be wrong though, it was more than a year ago since I used network debugging to test things.

[quote=pancake_stacks]Have you looked into testing with the net_showsplits command? I believe it dumps to console how it handles splits.[/quote]
IIRC I did, but it was too spammy to give any useful information. Could be wrong though, it was more than a year ago since I used network debugging to test things.
817
#817
0 Frags +

These are my final results:
Comp quality
1024x768
i3-6006U
Intel HD 520
dxlevel 9.1

2639 frames 34.502 seconds 76.49 fps (13.07 ms/f) 6.272 fps variability

Overall the comp quality pretty much matches Comanglia's stability, I'm gonna stick with this config.

These are my final results:
Comp quality
1024x768
i3-6006U
Intel HD 520
dxlevel 9.1

2639 frames 34.502 seconds 76.49 fps (13.07 ms/f) 6.272 fps variability

Overall the comp quality pretty much matches Comanglia's stability, I'm gonna stick with this config.
818
#818
0 Frags +

is the comp.cfg supposed to be used with another cfg as it seems to have a significantly smaller amount of console commands compared to the others

is the comp.cfg supposed to be used with another cfg as it seems to have a significantly smaller amount of console commands compared to the others
819
#819
1 Frags +
Viviis the comp.cfg supposed to be used with another cfg as it seems to have a significantly smaller amount of console commands compared to the others

Comp.cfg is an addon to your main Comfig that changes some settings due to specific situation (e.g comp.cfg is more for competitive players, obviuosly). It should be executed right after the main part of the config

[quote=Vivi]is the comp.cfg supposed to be used with another cfg as it seems to have a significantly smaller amount of console commands compared to the others[/quote]

Comp.cfg is an addon to your main Comfig that changes some settings due to specific situation (e.g comp.cfg is more for competitive players, obviuosly). It should be executed right after the main part of the config
820
#820
0 Frags +

I installed your config but when I opened the game the fog seems thicker.

Launch Options:

-novid -nojoy -high -nosteamcontroller -nohltv -reuse -primarysound -threads 2 -nogammaramp -disable_d3d9_hacks -usetcp -softparticlesdefaultoff -snoforceformat

Can you help? The only way to remove this fog is to use comangliacoms config (toaster preset) back or set sv_cheats to 1 and set fog_enable to 0 and fog_override to 1.

I tried to set r_pixelfog to 0 but it still doesn't work.

I'm using my custom.cfg preset btw.

Specs:
1280x720
i3-4170 3.7 ghz
AMD Radeon (TM) R5 M330
dxlevel 90
I installed your config but when I opened the game the fog seems thicker.

Launch Options:
[code]-novid -nojoy -high -nosteamcontroller -nohltv -reuse -primarysound -threads 2 -nogammaramp -disable_d3d9_hacks -usetcp -softparticlesdefaultoff -snoforceformat[/code]

Can you help? The only way to remove this fog is to use comangliacoms config (toaster preset) back or set sv_cheats to 1 and set fog_enable to 0 and fog_override to 1.

I tried to set r_pixelfog to 0 but it still doesn't work.

I'm using my [url=https://pastebin.com/YcwKTV9L]custom.cfg[/url] preset btw.

[code]Specs:
1280x720
i3-4170 3.7 ghz
AMD Radeon (TM) R5 M330
dxlevel 90[/code]
821
#821
1 Frags +
solly6

Remove -disable_d3d9_hacks
Also you shouldn't put -threads in any value lower than 4 so its probably better that you remove it

[quote=solly6][/quote]
Remove [b]-disable_d3d9_hacks[/b]
Also you shouldn't put [b]-threads[/b] in any value lower than 4 so its probably better that you remove it
822
#822
0 Frags +
sagesolly6Remove -disable_d3d9_hacks
Also you shouldn't put -threads in any value lower than 4 so its probably better that you remove it

Thanks, I'm gonna try it.

EDIT: Thanks,man. Glad you helped. The fog is gone now.

[quote=sage][quote=solly6][/quote]
Remove [b]-disable_d3d9_hacks[/b]
Also you shouldn't put [b]-threads[/b] in any value lower than 4 so its probably better that you remove it[/quote]

Thanks, I'm gonna try it.

EDIT: Thanks,man. Glad you helped. The fog is gone now.
823
#823
1 Frags +

Benchmarkss:

https://i.imgur.com/14o1MkY.png

With AMDGPU drivers.

launch options: -novid -gl_disable_arb_buffer_storage -nojoy -nosteamcontroller -noff -nohltv -softparticlesdefaultoff -reuse -primarysound -snoforceformat -swapcores -high -console

mastercomfig with presets/maxperformance:
2639 frames 18.516 seconds 142.53 fps ( 7.02 ms/f) 14.062 fps variability
2639 frames 18.429 seconds 143.20 fps ( 6.98 ms/f) 11.368 fps variability
2639 frames 18.421 seconds 143.26 fps ( 6.98 ms/f) 11.758 fps variability

Comanglia edited by mastercoms with mat_queue_mode 2/r_occlusion 0:
2639 frames 18.343 seconds 143.87 fps ( 6.95 ms/f) 12.022 fps variability
2639 frames 18.293 seconds 144.26 fps ( 6.93 ms/f) 10.572 fps variability
2639 frames 18.336 seconds 143.92 fps ( 6.95 ms/f) 10.565 fps variability

Comanglia edited by mastercoms with mat_queue_mode -1/r_occlusion 0:
2639 frames 18.387 seconds 143.52 fps ( 6.97 ms/f) 10.553 fps variability
2639 frames 18.499 seconds 142.66 fps ( 7.01 ms/f) 10.360 fps variability
2639 frames 18.488 seconds 142.74 fps ( 7.01 ms/f) 10.248 fps variability

Since this weekend my desktop is dual boot Windows 10/Linux. Once i'll be on win, i'll benchmark there too.
Having tf2 on SSD on win and on HDD on linux shouldn't invalidate the benchmarks, right? SSD is just about loading times afaik

I've HUGE fps drops on linux while playing MGE on mge_training_v8_beta4b:

7257 frames 60.088 seconds 120.77 fps ( 8.28 ms/f) 50.627 fps variability

i know @mastercoms plays on linux too... any idea of what it could be?
If i play even one match of mge, i have to restart tf2 or i'll have those fps drops in all other maps too

Benchmarkss:
[img]https://i.imgur.com/14o1MkY.png[/img]
With AMDGPU drivers.

[code]launch options: -novid -gl_disable_arb_buffer_storage -nojoy -nosteamcontroller -noff -nohltv -softparticlesdefaultoff -reuse -primarysound -snoforceformat -swapcores -high -console


mastercomfig with presets/maxperformance:
2639 frames 18.516 seconds 142.53 fps ( 7.02 ms/f) 14.062 fps variability
2639 frames 18.429 seconds 143.20 fps ( 6.98 ms/f) 11.368 fps variability
2639 frames 18.421 seconds 143.26 fps ( 6.98 ms/f) 11.758 fps variability


Comanglia edited by mastercoms with mat_queue_mode 2/r_occlusion 0:
2639 frames 18.343 seconds 143.87 fps ( 6.95 ms/f) 12.022 fps variability
2639 frames 18.293 seconds 144.26 fps ( 6.93 ms/f) 10.572 fps variability
2639 frames 18.336 seconds 143.92 fps ( 6.95 ms/f) 10.565 fps variability


Comanglia edited by mastercoms with mat_queue_mode -1/r_occlusion 0:
2639 frames 18.387 seconds 143.52 fps ( 6.97 ms/f) 10.553 fps variability
2639 frames 18.499 seconds 142.66 fps ( 7.01 ms/f) 10.360 fps variability
2639 frames 18.488 seconds 142.74 fps ( 7.01 ms/f) 10.248 fps variability[/code]

Since this weekend my desktop is dual boot Windows 10/Linux. Once i'll be on win, i'll benchmark there too.
Having tf2 on SSD on win and on HDD on linux shouldn't invalidate the benchmarks, right? SSD is just about loading times afaik



I've HUGE fps drops on linux while playing MGE on mge_training_v8_beta4b:
[code]7257 frames 60.088 seconds 120.77 fps ( 8.28 ms/f) 50.627 fps variability[/code]
i know @mastercoms plays on linux too... any idea of what it could be?
If i play even one match of mge, i have to restart tf2 or i'll have those fps drops in all other maps too
824
#824
0 Frags +

On Windows 10 (same setup) as above:
CPU: i5 -6400 2.7GHz (dunno why linux says it's 3.3)
GPU: AMD R9 380
Driver GPU: Radeon Software Version 17.7
dxlevel: 95
Resolution: 1920x1080 full screen

Launch options:

-novid -nojoy -nosteamcontroller -noff -nohltv -softparticlesdefaultoff -reuse -primarysound -snoforceformat -swapcores -high

mastercomfig with presets/maxperformance:

2639 frames 18.161 seconds 145.31 fps ( 6.88 ms/f) 12.153 fps variability
2639 frames 18.214 seconds 144.89 fps ( 6.90 ms/f) 10.549 fps variability
2639 frames 18.260 seconds 144.53 fps ( 6.92 ms/f) 10.797 fps variability

Comanglia edited by mastercoms with mat_queue_mode 2/r_occlusion 0:

2639 frames 18.251 seconds 144.59 fps ( 6.92 ms/f) 11.262 fps variability
2639 frames 18.186 seconds 145.11 fps ( 6.89 ms/f) 10.375 fps variability
2639 frames 18.239 seconds 144.69 fps ( 6.91 ms/f) 10.212 fps variability

Comanglia edited by mastercoms with mat_queue_mode -1/r_occlusion 0:

2639 frames 18.240 seconds 144.68 fps ( 6.91 ms/f) 10.331 fps variability
2639 frames 18.414 seconds 143.32 fps ( 6.98 ms/f) 10.568 fps variability
2639 frames 18.293 seconds 144.26 fps ( 6.93 ms/f) 10.095 fps variability

Same demo i recorded on mge_training_v8_beta4b:

7257 frames 29.011 seconds 250.14 fps ( 4.00 ms/f) 24.543 fps variability

Oh the strange thing is that if i go on a mge_ohiguv_sucks_a12 server, i've no issues on linux!

On Windows 10 (same setup) as above:
CPU: i5 -6400 2.7GHz (dunno why linux says it's 3.3)
GPU: AMD R9 380
Driver GPU: Radeon Software Version 17.7
dxlevel: 95
Resolution: 1920x1080 full screen

Launch options:
[code]-novid -nojoy -nosteamcontroller -noff -nohltv -softparticlesdefaultoff -reuse -primarysound -snoforceformat -swapcores -high[/code]
mastercomfig with presets/maxperformance:
[code]2639 frames 18.161 seconds 145.31 fps ( 6.88 ms/f) 12.153 fps variability
2639 frames 18.214 seconds 144.89 fps ( 6.90 ms/f) 10.549 fps variability
2639 frames 18.260 seconds 144.53 fps ( 6.92 ms/f) 10.797 fps variability[/code]

Comanglia edited by mastercoms with mat_queue_mode 2/r_occlusion 0:
[code]2639 frames 18.251 seconds 144.59 fps ( 6.92 ms/f) 11.262 fps variability
2639 frames 18.186 seconds 145.11 fps ( 6.89 ms/f) 10.375 fps variability
2639 frames 18.239 seconds 144.69 fps ( 6.91 ms/f) 10.212 fps variability[/code]

Comanglia edited by mastercoms with mat_queue_mode -1/r_occlusion 0:
[code]2639 frames 18.240 seconds 144.68 fps ( 6.91 ms/f) 10.331 fps variability
2639 frames 18.414 seconds 143.32 fps ( 6.98 ms/f) 10.568 fps variability
2639 frames 18.293 seconds 144.26 fps ( 6.93 ms/f) 10.095 fps variability[/code]

Same demo i recorded on mge_training_v8_beta4b:
[code]7257 frames 29.011 seconds 250.14 fps ( 4.00 ms/f) 24.543 fps variability[/code]

Oh the strange thing is that if i go on a mge_ohiguv_sucks_a12 server, i've no issues on linux!
825
#825
0 Frags +

Ran some benchmarks on Mint 17.3. Yikes. Anyone got any advice to improve performance?

2639 frames 36.642 seconds 72.02 fps (13.88 ms/f) 3.436 fps variability
2639 frames 36.493 seconds 72.32 fps (13.83 ms/f) 3.159 fps variability
2639 frames 36.790 seconds 71.73 fps (13.94 ms/f) 3.107 fps variability

CPU: Core i7 870
GPU: GeForce GTX 970
Driver: 387.34
OS: Linux Mint 18.3 Cinnamon 64-bit.

Ran some benchmarks on Mint 17.3. Yikes. Anyone got any advice to improve performance?

2639 frames 36.642 seconds 72.02 fps (13.88 ms/f) 3.436 fps variability
2639 frames 36.493 seconds 72.32 fps (13.83 ms/f) 3.159 fps variability
2639 frames 36.790 seconds 71.73 fps (13.94 ms/f) 3.107 fps variability

CPU: Core i7 870
GPU: GeForce GTX 970
Driver: 387.34
OS: Linux Mint 18.3 Cinnamon 64-bit.
826
#826
3 Frags +

Linux users: maybe try gl_batch_tex_creates 0;gl_batch_tex_destroys 0.

If that doesn't work, set them back to 1 and then try gl_nullqueries 0.

If that doesn't work, set that back to 1 and then try gl_minimize_rt_tex 0.

DashieGasai

What are your launch options and preset?

Linux users: maybe try gl_batch_tex_creates 0;gl_batch_tex_destroys 0.

If that doesn't work, set them back to 1 and then try gl_nullqueries 0.

If that doesn't work, set that back to 1 and then try gl_minimize_rt_tex 0.

[quote=DashieGasai][/quote]
What are your launch options and preset?
827
#827
0 Frags +
mastercomsWhat are your launch options and preset?

Just mastercomfig with
-novid -nojoy -nosteamcontroller -noff -nohltv -softparticlesdefaultoff -reuse -primarysound -snoforceformat -swapcores

[quote=mastercoms]
What are your launch options and preset?[/quote]

Just mastercomfig with
-novid -nojoy -nosteamcontroller -noff -nohltv -softparticlesdefaultoff -reuse -primarysound -snoforceformat -swapcores
828
#828
2 Frags +
mastercomsLinux users: maybe try gl_batch_tex_creates 0;gl_batch_tex_destroys 0.

Awesome! That' fixed the deal for me on mge_training_v8_beta4b:

7257 frames 29.217 seconds 248.38 fps ( 4.03 ms/f) 22.865 fps variability
[quote=mastercoms]Linux users: maybe try gl_batch_tex_creates 0;gl_batch_tex_destroys 0.[/quote]

Awesome! That' fixed the deal for me on mge_training_v8_beta4b:
[code]7257 frames 29.217 seconds 248.38 fps ( 4.03 ms/f) 22.865 fps variability
[/code]
829
#829
5 Frags +

Cleaned configs up, added gl_batch_tex_creates 0;gl_batch_tex_destroys 0 and bam!

2639 frames 22.137 seconds 119.21 fps ( 8.39 ms/f) 8.710 fps variability
2639 frames 22.469 seconds 117.45 fps ( 8.51 ms/f) 7.959 fps variability
2639 frames 22.400 seconds 117.81 fps ( 8.49 ms/f) 8.030 fps variability

Thanks a lot! This is so much better.

Cleaned configs up, added gl_batch_tex_creates 0;gl_batch_tex_destroys 0 and bam!

[code]
2639 frames 22.137 seconds 119.21 fps ( 8.39 ms/f) 8.710 fps variability
2639 frames 22.469 seconds 117.45 fps ( 8.51 ms/f) 7.959 fps variability
2639 frames 22.400 seconds 117.81 fps ( 8.49 ms/f) 8.030 fps variability
[/code]

Thanks a lot! This is so much better.
830
#830
0 Frags +

I'd appreciate it if people could test net_compresspackets 0;net_maxroutable 576 to see its effects on FPS and networking (hitreg/projectiles)

I'd appreciate it if people could test net_compresspackets 0;net_maxroutable 576 to see its effects on FPS and networking (hitreg/projectiles)
831
#831
0 Frags +

mastercoms, where should I put dxsupport_override.cfg and texture_preload_list.txt? Does it improve my framerates?

Is it ok to use gl commands on Windows?

mastercoms, where should I put dxsupport_override.cfg and texture_preload_list.txt? Does it improve my framerates?

Is it ok to use gl commands on Windows?
832
#832
0 Frags +
solly6mastercoms, where should I put dxsupport_override.cfg and texture_preload_list.txt? Does it improve my framerates?

Is it ok to use gl commands on Windows?

You should be putting the mastercomfig and mastercomfig_exec folders into your tf/custom folder. They do improve framerates and load times.

gl commands won't work on Windows.

[quote=solly6]mastercoms, where should I put dxsupport_override.cfg and texture_preload_list.txt? Does it improve my framerates?

Is it ok to use gl commands on Windows?[/quote]
You should be putting the mastercomfig and mastercomfig_exec folders into your tf/custom folder. They do improve framerates and load times.

gl commands won't work on Windows.
833
#833
1 Frags +

i am currently using Comanglia's stability config on my shitty computer.
i want to switch to this config but i have a a question.

does any of my personal settings conflict with this config?
:
//My Stuff =D
sensitivity 3.33
volume .15
fov_desired 90
viewmodel_fov 70
viewmodel_fov_demo 70
cl_hud_minmode 0
cl_disablehtmlmotd 0
cl_autorezoom 0
tf_hud_no_crosshair_on_scope_zoom 0
cl_autoreload 1
hud_fastswitch 1
hud_classautokill 0
hud_combattext 1
tf_dingalingaling 1
tf_dingaling_volume .6
tf_dingaling_pitchmindmg 140
tf_dingaling_pitchmaxdmg 5
tf_medigun_autoheal 1
cam_ideallag 0
cl_spec_carrieditems 1
cl_hud_playerclass_playermodel_showed_confirm_dialog 1
tf_hud_target_id_disable_floating_health 1
con_enable 1
cl_interp 0

thanks.

i am currently using Comanglia's stability config on my shitty computer.
i want to switch to this config but i have a a question.

does any of my personal settings conflict with this config?
:
//My Stuff =D
sensitivity 3.33
volume .15
fov_desired 90
viewmodel_fov 70
viewmodel_fov_demo 70
cl_hud_minmode 0
cl_disablehtmlmotd 0
cl_autorezoom 0
tf_hud_no_crosshair_on_scope_zoom 0
cl_autoreload 1
hud_fastswitch 1
hud_classautokill 0
hud_combattext 1
tf_dingalingaling 1
tf_dingaling_volume .6
tf_dingaling_pitchmindmg 140
tf_dingaling_pitchmaxdmg 5
tf_medigun_autoheal 1
cam_ideallag 0
cl_spec_carrieditems 1
cl_hud_playerclass_playermodel_showed_confirm_dialog 1
tf_hud_target_id_disable_floating_health 1
con_enable 1
cl_interp 0

thanks.
834
#834
0 Frags +
mastercomsI'd appreciate it if people could test net_compresspackets 0;net_maxroutable 576 to see its effects on FPS and networking (hitreg/projectiles)

Not really, atleast for me.

net_compresspackets 0; net_maxroutable 576

Benchmarks: 
2639 frames 15.886 seconds 116.2 fps ( 6.02 ms/f) 7.576 fps variability

net_compresspackets 1; net_maxroutable 1200

Benchmarks:
2639 frames 15.886 seconds 116.07 fps ( 6.02 ms/f) 7.563 fps variability
Specs:
1280x720
i3-4170 3.7 ghz
AMD Radeon (TM) R5 M330
dxlevel 90
[quote=mastercoms]I'd appreciate it if people could test net_compresspackets 0;net_maxroutable 576 to see its effects on FPS and networking (hitreg/projectiles)[/quote]

Not really, atleast for me.

net_compresspackets 0; net_maxroutable 576
[code]Benchmarks:
2639 frames 15.886 seconds 116.2 fps ( 6.02 ms/f) 7.576 fps variability[/code]

net_compresspackets 1; net_maxroutable 1200
[code]Benchmarks:
2639 frames 15.886 seconds 116.07 fps ( 6.02 ms/f) 7.563 fps variability[/code]

[code]Specs:
1280x720
i3-4170 3.7 ghz
AMD Radeon (TM) R5 M330
dxlevel 90[/code]
835
#835
0 Frags +

I might actually try again if there are any effects.

EDIT: Nothing at all.
EDIT2: One question btw, is there any commands that will make my game look blocky? I can use the toaster preset from comangliacoms and get blocky textures.

I might actually try again if there are any effects.

EDIT: Nothing at all.
EDIT2: One question btw, is there any commands that will make my game look blocky? I can use the toaster preset from comangliacoms and get blocky textures.
836
#836
0 Frags +
solly6

It won't affect benchmarks since those are local. You'll have to test in an actual multiplayer game.

The maxperformance preset will make it look like that. The commands that cause it are mat_filtertextures 0;mat_filterlightmaps 0.

Zharga_

Just remove cl_interp 0.

[quote=solly6][/quote]

It won't affect benchmarks since those are local. You'll have to test in an actual multiplayer game.

The maxperformance preset will make it look like that. The commands that cause it are mat_filtertextures 0;mat_filterlightmaps 0.

[quote=Zharga_][/quote]
Just remove cl_interp 0.
837
#837
0 Frags +
mastercomsFreudAlso, is the tickrate in TF2 66 or 66.66...? (ie, is the minimum interp I should use 0.01515... or 0.015?) Again, I recall reading differing advice on this.
TF2's tickrate is 66. So, if you have cl_interp_ratio set to 1, and cl_updaterate set to the max of 66, you'll get a cl_interp of 0.01515151515.

From console when creating local server:

SV_ActivateServer: setting tickrate to 66.7

Are you sure the tickrate is 66? This seems strange if that is the case.

comfig.cfgnet_compresspackets 0 // Remove CPU load from network packet compression
//net_compresspackets 1 // Compress packets on weaker CPUs or on <2.4Mbps stable upload

Two things:

  • The comments here seem unclear. net_compresspackets 0 removes CPU load from packet compression, yet you recommend net_compresspackets 1 for weaker CPUs?
  • You seem to be using settings that reduce CPU load at the cost of increasing network usage? (Again, the comments are a bit unclear.) Are there settings that could be used to do the opposite - reduce network usage possibly at the cost of increased CPU load?
[quote=mastercoms][quote=Freud]Also, is the tickrate in TF2 66 or 66.66...? (ie, is the minimum interp I should use 0.01515... or 0.015?) Again, I recall reading differing advice on this.[/quote]

TF2's tickrate is 66. So, if you have cl_interp_ratio set to 1, and cl_updaterate set to the max of 66, you'll get a cl_interp of 0.01515151515.[/quote]

From console when creating local server:

[code]SV_ActivateServer: setting tickrate to 66.7[/code]

Are you sure the tickrate is 66? This seems strange if that is the case.

[quote=comfig.cfg]net_compresspackets 0 // Remove CPU load from network packet compression
//net_compresspackets 1 // Compress packets on weaker CPUs or on <2.4Mbps stable upload[/quote]

Two things:
[list]
[*]The comments here seem unclear. net_compresspackets 0 removes CPU load from packet compression, yet you recommend net_compresspackets 1 for weaker CPUs?
[*]You seem to be using settings that reduce CPU load at the cost of increasing network usage? (Again, the comments are a bit unclear.) Are there settings that could be used to do the opposite - reduce network usage possibly at the cost of increased CPU load?
[/list]
838
#838
0 Frags +
FreudFrom console when creating local server:
SV_ActivateServer: setting tickrate to 66.7

Are you sure the tickrate is 66? This seems strange if that is the case.

Seeing as you were talking about networking,

Freudie, is the minimum interp I should use 0.01515... or 0.015

my answer was correct. You can only send a max of 66 snapshots every second, since it's impossible to send 2/3 of a snapshot. The fact that the server ticks slightly faster is irrelevant.

FreudThe comments here seem unclear. net_compresspackets 0 removes CPU load from packet compression, yet you recommend net_compresspackets 1 for weaker CPUs?

Yes, because disabling compression increases the number of split packets you'll have to send, which by default are sent on another thread. If you have a weaker CPU, it would be better to use compression to avoid splitpackets whenever possible.

FreudYou seem to be using settings that reduce CPU load at the cost of increasing network usage? (Again, the comments are a bit unclear.) Are there settings that could be used to do the opposite - reduce network usage possibly at the cost of increased CPU load?

My settings don't do one or the other. It balances CPU load and network usage where it is useful for performance and network performance.

Besides, network usage and CPU load aren't inversely related, except for this one example with packet compression vs splitpackets, and even then the relation is not very clear most of the time.

[quote=Freud]From console when creating local server:

[code]SV_ActivateServer: setting tickrate to 66.7[/code]

Are you sure the tickrate is 66? This seems strange if that is the case.
[/quote]
Seeing as you were talking about networking,
[quote=Freud]ie, is the minimum interp I should use 0.01515... or 0.015[/quote]
my answer was correct. You can only send a max of 66 snapshots every second, since it's impossible to send 2/3 of a snapshot. The fact that the server ticks slightly faster is irrelevant.

[quote=Freud]The comments here seem unclear. net_compresspackets 0 removes CPU load from packet compression, yet you recommend net_compresspackets 1 for weaker CPUs?
[/quote]
Yes, because disabling compression increases the number of split packets you'll have to send, which by default are sent on another thread. If you have a weaker CPU, it would be better to use compression to avoid splitpackets whenever possible.

[quote=Freud]You seem to be using settings that reduce CPU load at the cost of increasing network usage? (Again, the comments are a bit unclear.) Are there settings that could be used to do the opposite - reduce network usage possibly at the cost of increased CPU load?
[/quote]
My settings don't do one or the other. It balances CPU load and network usage where it is useful for performance and network performance.

Besides, network usage and CPU load aren't inversely related, except for this one example with packet compression vs splitpackets, and even then the relation is not very clear most of the time.
839
#839
0 Frags +

One question btw, is there any commands that will make my game look blocky? I can use the toaster preset from comangliacoms and get blocky textures.

One question btw, is there any commands that will make my game look blocky? I can use the toaster preset from comangliacoms and get blocky textures.
840
#840
0 Frags +
solly6One question btw, is there any commands that will make my game look blocky? I can use the toaster preset from comangliacoms and get blocky textures.

http://www.teamfortress.tv/42867/mastercomfig-fps-customization-config/?page=28#836

[quote=solly6]One question btw, is there any commands that will make my game look blocky? I can use the toaster preset from comangliacoms and get blocky textures.[/quote]
http://www.teamfortress.tv/42867/mastercomfig-fps-customization-config/?page=28#836
1 ⋅⋅ 25 26 27 28 29 30 31 ⋅⋅ 58
Please sign in through STEAM to post a comment.