DDraceNetwork - questions

DDraceNetwork

DDraceNetwork / questions

For newcomers to ask questions and get help. Also check https://wiki.ddnet.org/wiki/FAQ

Between 2024-06-15 00:00:00Z and 2024-06-16 00:00:00Z

Jupstar ✪

if i can also spot the difference between 700 fps and 1000 fps is probably only possible if i play with both for a week or smth but under 700 fps i see it directly

You can tell if a 240hz monitor is getting less than 700fps? It would be very interesting to see you in a double blind test

Learath2

You can tell if a 240hz monitor is getting less than 700fps? It would be very interesting to see you in a double blind test

sure

11:55

why should i not be able to tell?

11:56

for you i can explain it a bit simpler: imgine a framebuffer is about to be used by the graphics card to be send to the monitor. if u have more frames, u more likely get a more fresh frame. since frame times are not perfect u will always sometimes miss to hit the perfect timing for the buffer to be sent

11:56

so more frames generally make the timing less "random"

11:57

and the second important thing i wrote is, for a desktop GPU 500 FPS or 1000 FPS is almost the same power wise

11:57

so why should i not play with 1000 fps if it gives me a smoother image

11:58

ah and ofc assuming playing in entities which are light for the GPU to compute

The variation in average frametime between 740 and 660 fps on a 240Hz monitor would be tiiiiny. That's why I'm very skeptical that you can tell those apart

well it progressively gets worse

12:00

i didnt do frame by frame decrease

12:00

i tested 1000 fps

12:00

then 900

12:00

800

12:01

and so on

12:01

maybe i can even spot the difference between 700 and 1000 if i use both for a longer time

12:03

if i start ddnet right now and test 1000 fps vs 500 fps i can just by moving left and right tell the 1000 fps is smoother

Jupstar ✪

if i start ddnet right now and test 1000 fps vs 500 fps i can just by moving left and right tell the 1000 fps is smoother

Well you know which is which. Your brain can kill you if you believe you are about to die

12:06

For a proper test there should be no indication of performance at all, you just get a monitor and keyboard/mouse. The case should be in another room where you can't even hear the fans

Learath2

Well you know which is which. Your brain can kill you if you believe you are about to die

ok, but i think the difference here is big enough so it is not placebo

I'm eating rn, but I'll math it out after. I'm curious how much fresher of a frame you can get

12:07

Anything above 240 and you should never get a duplicated frame, that is easy enough to show

note i dont use vsync or freesync the tearing is what makes it unsmooth

Yeah, I'll calculate with no sync

on 240hz tearing is already very hard to observe, but 500 fps feels a bit less smooth while moving around

12:08

notice too, that i only see it less smooth, i dont claim my skill drops or i react worse

12:08

the eye can see high refresh rates, even if it has no advantage

12:09

it's simply not as fluent, dunno how to describe it

Jupstar ✪

notice too, that i only see it less smooth, i dont claim my skill drops or i react worse

Yeah I'm just curious if the difference in frametime is actually within what we documented humans to be capable of

buy my setup and u can notice it too i guess xD

Learath2

Yeah I'm just curious if the difference in frametime is actually within what we documented humans to be capable of

well youtubers can also see 480hz vs 240hz, so why not

12:10

or do they all lie for the ads xd

Jupstar ✪

or do they all lie for the ads xd

That or they might accidentally unconsciously bias themselves

i assume that if u play a game very often, u notice the differences but if u play a random game where u dont really have focus to anything specific

12:11

it might not give u any clear indicator which is better

12:11

in 3D is harder anyway i'd say, u'd need to flick around or smth

12:12

come to germany i show u my gores setup

12:12

Also a lot of devices have gsync/freesync in them nowadays. So that also helps

Learath2

Also a lot of devices have gsync/freesync in them nowadays. So that also helps

i dont use freesync because i feel the tiny delay

12:13

sync is always a bit of delay

12:13

for minecraft... totally fine

12:13

for gores, annoying xd

12:13

i even notice it on wayland, where every frame is perfect

12:13

well before they added tearing protocol xD

12:14

for the true gamers

Jupstar ✪

i dont use freesync because i feel the tiny delay

Huh, that is really teetering on the absurd. You are insanely sensitive. Freesync/gsync introduces what, nanoseconds? Not like they work similar to vsync

Learath2

Huh, that is really teetering on the absurd. You are insanely sensitive. Freesync/gsync introduces what, nanoseconds? Not like they work similar to vsync

they are basically like vsync but dont block the CPU

12:15

they still do the same effect, just on the monitor

I would really love to have you and a couple other people in a room with a proper double blind

lets say it like that

12:16

vsync is not a good indicator

12:16

let's talk about mailbox

12:16

that is tearing with many FPS

12:16

sry

12:16

no tearing ofc

12:16

https://registry.khronos.org/vulkan/specs/1.3-extensions/man/html/VkPresentModeKHR.html

12:17

"VK_PRESENT_MODE_MAILBOX_KHR specifies that the presentation engine waits for the next vertical blanking period to update the current image. Tearing cannot be observed. An internal single-entry queue is used to hold pending presentation requests. If the queue is full when a new presentation request is received, the new request replaces the existing entry, and any images associated with the prior entry become available for reuse by the application. One request is removed from the queue and processed during each vertical blanking period in which the queue is non-empty."

learath about to destroy jupstar using MATH and SCIENCE

12:20

don't trust your eyes, studies say that's dangerous

Learath2

I would really love to have you and a couple other people in a room with a proper double blind

but pls add gores pros

No destruction. Only curious how he is so much more sensitive than anyone else I have seen

Learath2

No destruction. Only curious how he is so much more sensitive than anyone else I have seen

i've talked about this with other tw players

12:21

sadly not everyone has 240hz monitors

You can talk all you want. It's sadly not even close to a valid trial when you all have fps counters up and know your settings

@Learath2 do u even have a high refresh rate monitor?

12:22

there are also these weird ppl that say 60hz vs 144hz is not noticable

12:22

so yes maybe others are not as sensitive lmao

12:22

i'd even say between 144hz and 240hz there is a huge difference

Jupstar ✪

@Learath2 do u even have a high refresh rate monitor?

I used to have a 144Hz one back in turkey. Went down to 60hz here

12:23

There is obviously a difference. But not one my brain can't adapt to within just a day or two

sure, but that is not our discussion i guess

12:23

as i said i can also play on 60hz

12:23

not like my skill is suddenly 0

Yeah, saying that one is imperceptible is insane. It just gets imperceptible as your brain adapts

12:24

240 -> 480 I'm starting to think there is a hidden sponsorship

training > higher refresh rates

12:25

i'd say hardware wise the most important in order ist:

good mouse
good keyboard
stable fps & network
good monitor

Also I seriously don't see how you can perceive freesync adding input lag. Maybe something is wrong with the implementation in your monitor/driver/gpu?

so monitor is least important

I would put stable network above all others. It gets exponentially worse as you get jitter and packet loss

Learath2

Also I seriously don't see how you can perceive freesync adding input lag. Maybe something is wrong with the implementation in your monitor/driver/gpu?

it forces the full image to wait 1000/240 ms

12:26

so u see the same image longer than with tearing

12:26

whihc already shows the updated image on some parts of the monitor

Learath2

I would put stable network above all others. It gets exponentially worse as you get jitter and packet loss

packet loss is a problem indeed, jitter can at least for solo play be contered with cl_prediction_margin 300

12:27

but ofc also only to a certain extend

12:27

but tbf u could say, u can play offline

12:27

if it's just about playing

12:28

but anyway

12:28

buy rtx 4090, i am not sponsered by nvidia

12:28

just buy it

When I win the lottery I will

Jupstar ✪

it forces the full image to wait 1000/240 ms

I think this is wrong though. That 4ms is already in your monitors display cycle

Learath2

I think this is wrong though. That 4ms is already in your monitors display cycle

it's partially true i think. if u are unlucky u look at the bottom of your screen where u see an old image

12:30

while with tearing it would have been a more recent one

12:31

up to 4ms

Iirc variable refresh rate will wait more for a new frame if it's not ready at 4ms, it won't re-display

Learath2

Iirc variable refresh rate will wait more for a new frame if it's not ready at 4ms, it won't re-display

but it always displays the whole buffer at once

12:31

it never partially updates it

12:32

and it takes 4ms to display the whole monitor

12:32

it cannot do that in an instant

But you already have insane fps, far beyond what it can sample all. I think average worst case it can add half a frametime (edited)

else we would not need any "sync"

Learath2

But you already have insane fps, far beyond what it can sample all. I think average worst case it can add half a frametime (edited)

yep

12:33

and i can ensure u i can still play well on 240hz even with VSync

12:33

it's absolutely not comparable to 60hz sync

Vsync is a completely different story. It adds very perceptible constant overhead

yeah vsync is harder indeed

12:34

mailbox is good

Anyway, we are just babbling. We are both just rationalizing our beliefs. There is only one way to settle this, a nicely set up double blind

12:35

With controls for gaming experience and domain knowledge

i could ask my brother, lmao

Has to be double blind. Your knowledge of the state can spoil the trial

if i dont see fps, how should i guess it?

12:37

if i get it 10/10

You know the config

well why should i cheat lmao

12:37

buy your 480hz oled benq today

I guess you could write a script that flips a coin to pick between 500 and 1000

at jup.star.shop

Even stuff like how much fan noise the system is making is not good. I can tell from the coil whine that I get in my headphones about how much my gpu is being utilized

yep

12:40

i can just tell you rx 6900 xt in entities, 500 fps vs 1000 fps basically makes no difference

12:40

2 Watt

12:40

41Watt vs 43Watt for a 300 Watt TDP card

That's good enough. I guess give it a try

12:44

That's 2ms per frame vs 1ms per frame. Which I guess on average will be 1ms vs .5ms of old frames

insane that our eyes are so good

At specific things. We are good at detecting change in the middle of our fov. But some trials showed frames shown for shorter than 13ms are very incompletely ingested by the brain

ok then i am glad i have a super brain

I said incomplete, not imperceptible

no matter what the cause is, but i am 99.9999999999% sure i can spot a difference when switching between 1000 fps and 500 fps ingame

12:49

xD

If I win the lottery I'll set up an experiment one day

nice, i will defs follow it

12:52

also do input lag, windows vs linux tests

Jupstar ✪

no matter what the cause is, but i am 99.9999999999% sure i can spot a difference when switching between 1000 fps and 500 fps ingame

yeah i could never, i would oinly ever notice a difference when i cap the FPS below my monitor's HZ

14:54

i didnt even feel a difference playing on 20K fps :D

meloƞ

yeah i could never, i would oinly ever notice a difference when i cap the FPS below my monitor's HZ

but are you certain you have no *sync active?

14:56

else you don't have the vegan power

14:56

im not sure what you mean with *sync, i hope my monitor is async! - justatest

justatest

like freesync

oh no i dont have those active

thats good

15:00

then u just blind

they for some reason stop text from beeing rendered completly

300 fps is unplayable

meloƞ

they for some reason stop text from beeing rendered completly

wait what?

15:00

r u on AMD?

yes.

15:00

idk how to properly describe it

Embedded image

15:01

like this?

nah - wait let me find a picture

15:01

language barrier fr

write german

cant, murpi is watching

he speaks german too (edited)

15:01

he aint noticing

cant find a picture rn, lets just say the text will look blurred and torn apart

15:03

imagine it like you're using a non-valid font in your terminal and you put the resolution of the text to 50x10

meloƞ

imagine it like you're using a non-valid font in your terminal and you put the resolution of the text to 50x10

mh weird tho

15:06

freesync should not alter text quality

15:06

u have 240hz?

ye

and u dont see diff between... lets say 300 fps and 1000 fps?

not noticable to the point that i care no

15:08

the second i go below 240 i see and feel it

15:08

but lets say 1000 -> 10000 i dont really see a difference

meloƞ

not noticable to the point that i care no

but what does care mean

15:08

do you see it or not

nah

lmao

15:08

geh augenarzt bruder

i passed the sehtest for the führerschein bruder, i dont need no augenarzt (edited)

15:09

even tho that was 6-7 years ago, but thats not the point here! (edited)

@meloƞ if u set ui_smooth_scroll_time 0 and scroll in server browser while hovering one element

15:11

do u sometimes for a very smal frame see an extra bar?

15:12

if u have a fast scroll wheel u can simply go all out

oh yeah i see that

would u say it looks consistent or would u say it looks like it's broken

15:13

the extra bar

not really consistent, i mean its not intended so i wouldnt expect it to be lol

now test with 1500 fps for difference (edited)

its the same KEKW

KEKW

shut up

it disappears slighly faster

15:15

but its still inconsistent and flickers

1500 fps it looks completely broken xD

KEKW

yeah

15:15

it flickers like crazy

15:15

compared to 300 fps

15:15

often it's only visible partially etc

its completly consistent on 200 fps tho, no flickering, just pure white greenthing

greenthing

but anyway.. that u can spot the difference in flickering is good

15:17

that kinda proofs my point

15:17

haters gonna hate

dude made a sehtest with teeworlds

15:17

gg

where can i fand game skins ?

19:49

the skins.tw site is suuuper slow

nicozicoyt

the skins.tw site is suuuper slow

so you know where to find them)

19:51

if this website is slow for you then probably others will be too

brother

justatest

1

nicozicoyt

where can i fand game skins ?

prob this https://ddnet.org/skins/

lolkildal

prob this https://ddnet.org/skins/

gameskins are texturepacks (edited)

21:36

not tee skins

skins.tw or teedata.net

Teeskins: Teeworlds Skins Database

Welcome to Teeskins — the largest collection of community created Teeworlds Skins. Download, Upload and Share Skins, Mapres, Gameskins, Emoticons, Cursors, Particles, Entities, Fonts, Grids and useful tools.

Welcome - Teedata

Teedata is a web-based database for Teeworlds. Browse through Teeworlds graphics, Tools and more.

Exported 210 message(s)