Skip to content

Instantly share code, notes, and snippets.

@mastercoms
Created May 13, 2017 14:12
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save mastercoms/1da43d6303ec11012fca7a8b4db705b5 to your computer and use it in GitHub Desktop.
Save mastercoms/1da43d6303ec11012fca7a8b4db705b5 to your computer and use it in GitHub Desktop.
tf2cfg3 beta3
// ~~~~ Mastercoms's TF2 Config ~~~~
// --- Network ---
// send to server at this many times per second
cl_cmdrate 66
// get from server at this many times per second
cl_updaterate 66
// reduce delay between entity interpolation to minimum,
// set to 2 for some room for error
cl_interp_ratio 1
// adjust delay between interpolating entities, set to 0 to use interp ratio
cl_interp 0
// smooth out entities to their new position after a prediction error
// prediction errors are often large and making entities jump instantly
// to their correct location is disorienting and makes it harder for you
// the server will still accept hits on an entity that is being smoothed,
// so there isn't much benefit to turning this off
cl_smooth 1
// time over which to smooth entities
cl_smoothtime 0.15
// server times out players after 65 seconds
// so why not wait until the server kicks you out?
cl_timeout 65
// compress packets save bytes, at the cost of
// extra cpu usage. if you have a fast enough network,
net_compresspackets 0
// how many packets we can split per frame
net_splitrate 3
// scaled rate for how many packets we are sending
// each packet can be a max size of 1260 bytes
// we can send up to 132 packets per second
rate 167580
// packet level ping control
net_maxcleartime 1.0
net_maxpacketdrop 660
// allow for split packets at the higher rate
net_splitpacket_maxrate 167581
// max out file upload size
net_maxfilesize 64
// always thread packets on a separate thread.
// very demanding of your computer, and the network system
//net_queued_packet_thread 581304
// --- Render ---
// -- Threading
// These settings will boost your performance on a 4 core CPU
// loop through the leaf asynchronously
cl_threaded_client_leaf_system 1
// matsys queue in multithreaded mode
mat_queue_mode 2
// use a threaded game logic loop
host_thread_mode 1
// do some rendering on a separate thread. rarely buggy.
r_threaded_renderables 1
// queue decals to matsys
// queuing decals also allows for batched decals
r_queued_decals 1
// usually doesn't render a texture for post processing effects, due to timing
// issues. better to disable it by default. it might work for your driver.
//r_queued_post_processing 1
// -- Optimization
// These settings don't affect quality that much and improve performance
// don't do smoothing on ropes
rope_smooth 0
// don't do memory flushes
host_flush_threshold 0
// don't try pixelvis, which is used for halos and pretty lights
r_dopixelvisibility 0
// don't use pixelvis in rendering
r_drawpixelvisibility 0
// don't use a costly precise pixelvis algorithm
r_pixelvisibility_partial 0
// don't store a useless frame time variable
r_norefresh 1
// force using occluders
r_occludermincount 2
// autodetect is broken for fast z rejects, so make it available.
// if it's not supported, it just won't be used
r_fastzreject 1
// renders textures dynamically
// but it's very buggy and flashes textures most of the time
//disp_dynamic 1
// use optimized fog rendering
fast_fogvolume 1
// -- Anti Aliasing
// don't do software based AA
mat_software_aa_strength 0
// don't do software based AA on the HUD
mat_software_aa_strength_vgui 0
// don't do software based AA
mat_software_aa_blur_one_pixel_lines 0
// don't limit performance when the engine is not focused.
// the focus detection can be buggy on some systems and may
// throttle your game, without you knowing
engine_no_focus_sleep 0
// see more of the battlefield. more to render. decrease for ABSOLUTE POTATOS
fov_desired 90
// prevent a memory leak with fps_max by setting to 240
fps_max 240
// don't do the costly "xray" outline
//glow_outline_effect_enable 0
lod_TransitionDist -5000
mp_usehwmodels -1
mp_usehwvcds -1
muzzleflash_light 0
tracer_extra 0 // don't add extra fluff to bullet lines
// -- Lighting --
r_ambientmin 0.35 // tweak lighting a bit to be realistic, no performance impact
r_rimlight 0 // disable that weird light around model edges, slightly better perf and better looks
r_lightaverage 0 // don't do this inefficient high cost lighting job that is barely noticeable
r_maxdlights 16 // reduce maximum dynamic lights that can be active in a frame
// TF2 won't hit this limit often
r_worldlightmin 0.0008 // don't render insignificant lighting
r_worldlights 2 // max lights applied to a vertex. 2 is a boost from the default 4
// and isn't a noticeable drop in quality
// -- Particles --
// do particles on a separate thread
r_threaded_particles 1
// if the thread pool is not available, use this many cores for particle jobs
particle_sim_alt_cores 3
// batch particles with best mode
cl_particle_batch_mode 2
// reduce particles, but it only helps a little bit because
// the real performance impact comes from creating particle system
mat_reduceparticles 1
// use the new particle system for creating new particles
cl_new_impact_effects 1
// don't create particle systems when things hit surfaces, perf and distracting
r_drawflecks 0
// --- Animations ---
// threshold for expression optimization (0.152 = 1/66)
ai_expression_frametime 0.0152
// enable expression optimization
ai_expression_optimization 1
// disable 3-way animation blending
anim_3wayblend 0
// duration of an eye blink (do not set to 0, or else TF2 will divide by 0)
blink_duration 0.0304
// don't follow flex rules
flex_rules 0
// don't smooth flexes
flex_smooth 0
// --- Shadows ---
// don't do a second shadow render pass for sharper shadows
cl_blobbyshadows 1
// threaded shadow manager, not used by default in source 2013
r_threaded_client_shadow_manager 1
// don't waste memory and rendering for shadows
r_flashlightdepthres 32
// disable shading/lighting useless performance hit
r_flashlightmodels 0
r_flashlightrender 0
// don't render shadows far away (nextbot system)
nb_shadow_dist 300
// don't render more shadows than needed
r_shadowmaxrendered 17
// --- Gibs --
// disable burning gibs
cl_burninggibs 0
// -- Props
// draw detail props up to this distance
cl_detaildist 1000
// fade in detail props across this distance, 0 to disable fading
cl_detailfade 0
// disable clientside physics props
cl_phys_props_enable 0
// limit the count of clientside physics props to a sane limit for TF2
cl_phys_props_max 20
// do not spawn props until we can see them
cl_phys_props_respawndist 1000
// delay in seconds before the client will respawn applicable props
cl_phys_props_respawnrate 120
func_break_max_pieces 0
// --- Misc ---
// don't draw monitors
cl_drawmonitors 0
// disable jigglebones if we are under the optimal framerate
cl_jiggle_bone_framerate_cutoff 66
// --- Effects ---
// disable shell ejection from pistols, shotgun and minigun
cl_ejectbrass 0
// every 15 frames, check for a temporary entity collision
// syringes, some effects are temp ents
cl_fasttempentcollision 15
// disable muzzle flashes in first person
cl_muzzleflash_dlight_1st 0
// use the impact effects system
cl_new_impact_effects 1
// disable water splashes
cl_show_splashes 0
cl_mvm_wave_status_visible_during_wave 1 // show useful information when you're fighting robots
cl_notifications_max_num_visible 1 // don't crowd up my game with notifications!
cl_notifications_move_time 0.1 // make notifications leave faster
hud_achievement_glowtime 1 // glow achievements for a little while shorter
hud_saytext_time 8 // fade chat a bit faster
cl_idealpitchscale 1.3 // may recover faster to center point from fliching
// --- Decals ---
mp_decals 9
r_decals 9
r_drawbatchdecals 1 // don't draw decals immediately
r_spray_lifetime 1 // freshen up the walls by clearing sprays every round
r_decal_cover_count 2 // if decals are almost fully covered by this many decals, remove them
r_decal_cullsize 20 // if decals are smaller than this on screen in pixels, don't render them
r_decal_overlap_area 0.5 // if decals are covered by other decals this much or more, remove them
r_decal_overlap_count 0 // all the other decal rules are fine. don't do this heavy impact check.
r_decalstaticprops 0 // don't decal and thus use ambient lighting for props,
// it's heavy and does pratically nothing
r_drawmodeldecals 0 // blood decals on bodies, hurts performance
r_maxmodeldecal 10 // optimized model decal count if you happen to enable it
r_overlayfadeenable 0 // fading overlays causes performance issues
// maps can already have their own fading rules
r_overlayfademax 1000 // optimize overlay fades if you happen to enable it
r_overlayfademin 745 // optimize overlay fades if you happen to enable it
r_renderoverlayfragment 1 // render overlays, pretty minor impact
// --- Crosshairs ---
// disable crosshair translucency
cl_crosshairalpha 255
// --- Console ---
con_nprint_bgalpha 255
con_nprint_bgborder 1
// --- Ragdolls ---
// disable ragdoll collisions, which are disabled anyway in the code
cl_ragdoll_collide 0
// duration of the fade out effect of ragdolls, 0 instantly removes
cl_ragdoll_fade_time 0
// start fading ragdolls without delay, even if the player is looking at them
cl_ragdoll_forcefade 1
// disable ragdoll physics
cl_ragdoll_physics_enable 0
// how fast a ragdoll fades out
g_ragdoll_fadespeed 10000
// how fast a ragdoll fades out in low violence mode
g_ragdoll_lvfadespeed 10000
g_ragdoll_maxcount 4 // ragdolls are expensive, so don't have a lot
ragdoll_sleepaftertime 1 // ragdolls are irrelevant once they've settled
// --- HUD ---
// adjust the translucency of the killstreak banner, 255 for solid
cl_hud_killstreak_display_alpha 255
// set how long the killstreak banner stays, 0 to disable banner
cl_hud_killstreak_display_time 0
// minimize clutter in HUD
cl_hud_minmode 1
// disable live player model in HUD
cl_hud_playerclass_use_playermodel 0
hud_achievement_tracker 0
hud_deathnotice_time 4
hud_reticle_minalpha 255
// --- Bug Reporter ---
// upload bug reporter attachments without blocking
bugreporter_uploadasync 1
// --- Memory ---
datacachesize 256
lzma_persistent_buffer 1
mem_min_heapsize 378
mem_max_heapsize 512
mem_max_heapsize_dedicated 96
mem_compact
// --- Filesystem ---
filesystem_buffer_size 32768 // keep files up to 32KB away in buffer
filesystem_max_stdio_read 64
filesystem_native 1
filesystem_unbuffered_io 0
mod_load_anims_async 1
mod_load_mesh_async 1
mod_load_vcollide_async 1
mod_offline_hdr_switch 1
// --- Sound ---
// disable spatial
dsp_enhance_stereo 0
// disable room FX mixers
dsp_slow_cpu 1
// disable auto dsp
dsp_room 0
// disable dsp effects
dsp_player 0
dsp_facingaway 0
dsp_speaker 0
dsp_spatial 0
dsp_water 0
// disable dsp
dsp_db_mixdrop 1 // don't scale volume
dsp_db_min 0 // don't scale volume
dsp_mix_min 0 // disable dsp
dsp_mix_max 0 // disable dsp
snd_async_fullyasync 1 // play sounds independently of main engine work
snd_async_minsize 256000 // wait until 256K of audio file has been loaded
snd_disable_mixer_duck 1 // disable ducking, which ruins positional audio
snd_mix_async 1 // have the sound mixer run asynchronously
snd_mixahead 0.08 // keep sounds for 80 millis and mix them
snd_noextraupdate 1 // don't update sounds twice
snd_spatialize_roundrobin 3 // spatialize sound every 8 frames
phonemedelay 0.1 // sounds are delayed for 0.08 + 0.02 seconds
phonemesnap 1 // don't render extra phonemes on even higher LODs
voice_buffer_ms 150 // buffer voice stream for better quality, at a slight delay
voice_overdrive 1.25 // decreases other sounds by 20% when voice comms are played
voice_overdrivefadetime 0.25 // fades sounds quicker in and out when voice is played
voice_steal 2 // reuse unimportant sound channels, increases perf and quality
// --- Input ---
in_usekeyboardsampletime 0
m_mousespeed 0
m_rawinput 1
m_filter 0
// --- Backpack ---
tf_item_selection_panel_sort_type 0 // see backpack-tweaks.cfg
tf_backpack_page_button_delay 0.25 // decrease button delay for moving items
tf_time_loading_item_panels 0.0002 // time spent per frame for loading item panels
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment