forum

At what point does fps increases have diminishing returns?

posted
Total Posts
87
show more
peppy
240 is a good number. if you can keep stable 500fps, then i'd go higher, else 240 is good.
I Give Up
420 fps is also a good number. Diminishing returns happen at 1337 fps unless you're mlg.
Multtari

Philosofikal wrote:

Motion blur is entirely in the eye because you don't see in frames per second - your sight is a continuous, real time chemical reaction to the exposure to light. The less time an image is exposed to your eye, the less it persists in your vision. This is why CRTs have basically no "motion blur" relative to their FPS - they refresh in strobes, the actual image is only exposed to your eye for only a small amount of time (which is why they flicker and cause eye strain). It's like a prehistoric version of Lightboost/ULMB (Ultra Low Motion Blur).

When you have inadequate pixel response times, you get ghosting, an entirely different problem. You get the "ghost" of previous frames in the current frame because the pixels can't keep up with the color transition, and are continuously in the middle of transitioning by the time the next frame appears, causing trails of improperly transitioned colors to appear behind moving objects.

http://www.overclock.net/t/1430257/what ... r-monitors
TLDR; Inputs can be calculated multiple times between still frames and are not dependant on your monitor refresh rate. What you want to look at is your tablet / mouse polling rate where adding more frames becomes useless.

When talking about tablet or mouse input the comfortablity is not entirely dependant on perceivable refresh rate of your screen or amount of motion blur but input lag which is decided by your in-game FPS. There is difference between 120fps and 240fps input lag. I don't think input lag is highly noticeable after 240.

I haven't tried playing with 120hz monitor but difference between 60hz and 80hz is huge when looking at notes and such. If you have enough rhythm skill and you know exactly where your cursor is going to land refresh rate is not important factor as framerate.

At some point i tried playing with game capped at 30hz as an experiment. Though being very uncomfortable i couldn't see huge hit in my playing ability as i would with capped framerate of 60 (V-Sync) which was suprising.


Correct me if i'm wrong.
Kert
We need special osu!tablets with 1000hz polling rate
bamboos will go to trash
E m i
>hi wacom-san
>can i has custom 1000hz tablet pls
ZenithPhantasm
ULMB/Lightboost strobing are both crappy imitations of CRT's no motion blur :^)
Saphirshroom

ZenithPhantasm wrote:

1000 fps because game logic runs at the same rate as fps and 1000hz is the maximum polling rate supported by USB 2.0
You're right about 1000 fps being the limit but game logic does not have to be limited by the fps rate at all.
Except, of course, when the PC begins to be limited by the game's graphical processing because it's not actually able to reach the fps threshold you gave it.
Full Tablet

Saphirshroom wrote:

ZenithPhantasm wrote:

1000 fps because game logic runs at the same rate as fps and 1000hz is the maximum polling rate supported by USB 2.0
You're right about 1000 fps being the limit but game logic does not have to be limited by the fps rate at all.
Except, of course, when the PC begins to be limited by the game's graphical processing because it's not actually able to reach the fps threshold you gave it.
The game logic of osu! is limited by the fps, fixing that would require rewriting a lot of the coding of the game.

Since the game takes the input once per frame, accuracy can be affected by low fps.

An estimation of how much accuracy is affected by fps (the table takes a theoretical input with a hit errors that follow a Normal Distribution with certain UR so the average accuracy at infinite fps is 95%, and estimates how accuracy and measured UR changes by changes of fps and the rounding done by the game of the hit errors)
(OD11 is added as a case where fps has an extreme effect on accuracy, it's not exactly OD10+DT, but is close)

Edit: made a new table with the timing windows shown in the game client instead of the ones in the wiki.
Saphirshroom

Full Tablet wrote:

Saphirshroom wrote:

You're right about 1000 fps being the limit but game logic does not have to be limited by the fps rate at all.
Except, of course, when the PC begins to be limited by the game's graphical processing because it's not actually able to reach the fps threshold you gave it.
The game logic of osu! is limited by the fps, fixing that would require rewriting a lot of the coding of the game.

Since the game takes the input once per frame, accuracy can be affected by low fps.
That's a... strange design decision. Thanks for pointing that out though, I didn't know that.
E m i
103,92 UR = 94.45%?

what
is it because the 100s were major lapses?
Full Tablet

[ Momiji ] wrote:

103,92 UR = 94.45%?
what
is it because the 100s were major lapses?
There are several possible reasons for that variation.

- The previous table only accounted for accuracy for circles (not sliders nor spinners). Assuming all sliders were a 300, 98.29% accuracy is equivalent to 97.40% accuracy on circles for that specific map. The game takes the error on sliders as well as circles for the calculation, if you tend to hit sliders less accurately than circles, then it is likely that in that play your UR on circles was less than the value shown.

- The previous table assumed a distribution of hit errors close to a Normal Distribution with mean error equal to 0 (Normal Distribution + an Uniform Distribution disturbance caused by the fps error; and rounded to the nearest millisecond). If your hit error distribution is not close to that distribution, then the estimation wouldn't be accurate.

- The table took the mean accuracy on circles. Because of the asymmetry of the Binomial Distribution, the median is different to the mean (unless the amount of circles tends to infinity, in that case they are equal); even though the mean is a certain value, there is a probability higher than 50% of getting an accuracy higher than the mean.

- The table took the timing windows shown in the Wiki, which are slightly wrong (18ms instead of 19.5ms for 300s).


According to a perfect Normal Distribution of hits (and fixed timing values), the probability of getting an accuracy of at least 98.29% (assuming all sliders are 300s) in that map with an UR on circles of 103.84 is 31.69% (so it's a very slight fluke according to the estimation). The median accuracy with that UR is 98.00% (the mean is 97.87%).

Edit: Also I noticed that that play was OD9.8, not OD10, fixed the values.
E m i
you're my favourite poster
autoteleology

ZenithPhantasm wrote:

ULMB/Lightboost strobing are both crappy imitations of CRT's no motion blur :^)
See signature, you have no idea what you're talking about
ZenithPhantasm
K person who posted a CoD video to show "omg i gud with mouse"
E m i
pls no bully
bigfeh

[ Momiji ] wrote:

pls no bully
autoteleology

ZenithPhantasm wrote:

K person who posted a CoD video to show "omg i gud with mouse"
K person who thinks liking a game completely invalidates the entire person as a whole

u shure been ridin that horse fer a while there cowboy, think it's time to give 'er a rest?
ZenithPhantasm

Philosofikal wrote:

Is there some problem with that? I see myself going up 40,000 rank in one month of playing, and hitting about 75,000 before that with two weeks of playing before selling my PC.

I have tons of experience using a mouse in a competitive setting.



Regardless of whether you like Call of Duty or not, you can't tell me with a straight face I don't know how to handle a mouse. You should give me some advice on how to be a rude moron so I can fit in better around here with people like you. :roll:
http://osu.ppy.sh/forum/t/303080/start=60
K m8
Yuudachi-kun
Zenny is our resident retard; pay him no attention. We use the RSI to keep him in check.
autoteleology

Kheldragar wrote:

Zenny is our resident retard; pay him no attention. We use the RSI to keep him in check.
The thing is, I don't even see what's wrong with that post. I absolutely understand how to handle a mouse well - I win almost every single pub match in CS:GO I play and I only have 86 hours in the game, and barely understand any of the strategy or metagame. That gameplay clip is the shit and totally proves my point.

I think the implication is that I was saying that made me immediately good at this particular game, but I didn't say that anywhere.
cheezstik

Philosofikal wrote:

Kheldragar wrote:

Zenny is our resident retard; pay him no attention. We use the RSI to keep him in check.
The thing is, I don't even see what's wrong with that post. I absolutely understand how to handle a mouse well - I win almost every single pub match in CS:GO I play and I only have 86 hours in the game, and barely understand any of the strategy or metagame. That gameplay clip is the shit and totally proves my point.

I think the implication is that I was saying that made me immediately good at this particular game, but I didn't say that anywhere.
Well if you admit that it doesn't make you good at this game, then what's the point of bringing it up, in an on-topic (or at least meant to be) section of an osu! forums? Cool story, you're good at cod or csgo, bring it to off-topic or whatever relevant section it's meant to be in.

Even in the original thread that you posted that cod gameplay in, look at the context, if you really didn't imply that knowing how to use a mouse in another game made you good at osu!, then why bother bringing it up?
autoteleology
Even in the original thread that you posted that cod gameplay in, look at the context, if you really didn't imply that knowing how to use a mouse in another game made you good at osu!, then why bother bringing it up?
I'm not the person who brought it up. Zenith follows me around Gameplay and Rankings and posts it in reply to half the posts I make, as if it's the only thing that defines me as a person. It's completely retarded and I don't get the point.

I brought it up originally because I basically gave the same tips that made me really good at mouse aim in everything else I've done. They've worked in many different games and there's no reason why they won't work here after a couple of years of practice. This game is not special when it comes to mouse control.

People questioned the legitimacy of my advice so I'm like, hey, I've gotten really good at mouse usage elsewhere, I have a system that works. I just happened to have visual evidence in this one particular game because I upload stuff to my YouTube, I never predicted there would be this huge nonsense backlash to it.
bigfeh

Philosofikal wrote:

Even in the original thread that you posted that cod gameplay in, look at the context, if you really didn't imply that knowing how to use a mouse in another game made you good at osu!, then why bother bringing it up?
I'm not the person who brought it up. Zenith follows me around Gameplay and Rankings and posts it in reply to half the posts I make, as if it's the only thing that defines me as a person. It's completely retarded and I don't get the point.

I brought it up originally because I basically gave the same tips that made me really good at mouse aim in everything else I've done. They've worked in many different games and there's no reason why they won't work here after a couple of years of practice. This game is not special when it comes to mouse control.

People questioned the legitimacy of my advice so I'm like, hey, I've gotten really good at mouse usage elsewhere, I have a system that works. I just happened to have visual evidence in this one particular game because I upload stuff to my YouTube, I never predicted there would be this huge nonsense backlash to it.
I gotta say I love how g&r has the ability to derail each and every single thread that's posted here. Isn't it lovely?

So as for the actual answer. First off, are you new to the internet? Because cod. Internet. Cod. Internet. No like. You were obviously going to get bashed for that video pretty much anywhere. Bringing it up originally was pretty obviously a mistake and just oh so happened to backfire so miserably that some people will always remember you as "that guy who posted a cod video on g&r". I'm among the most forgiving here but I'm sorry to tell you it is what it is.

The problem with your posts, you will recall, was not the content itself, but the way you posted said content. You posted as if your advice was the shit truth, while despite your best efforts to contribute (which, admittedly, could've been better), you just came off as incredibly cocky, because it was really all personal opinion (as I noted in one of my replies and as you said yourself, "here's something that works for me, why don't you try it?"). That's the whole reason you got bashed (that, and cod. cod too)

And on another note, mouse control in cod is barely as hard as csgo (in which pub matches are abso-fucking-lutely not proof of skill, or lack thereof) or this. Just figured I'd throw this in even though I'm pretty sure you'll disagree and try to take a dump on my head.


Here's a tip, though: don't.
drum drum

bigfeh wrote:

I gotta say I love how g&r has the ability to derail each and every single thread that's posted here. Isn't it lovely?
i forgot to ask if we could get back on topic
lets start
like
now
ZenithPhantasm
Hi drum drum! o/
bigfeh

ZenithPhantasm wrote:

Hi drum drum! o/
hi drum drumerino
pls no lockerino

this is so going to get locked, isn't it?

Anyways, not even sure if game logic runs at the same rate as your framerate, because that reeks an awful lot like bullshit. Rendering != processing unless peppy has some seriously weird code going on there

I'd say the question hasn't been answered yet, so I'm serious with the "pls no lockerino" thing
Full Tablet

bigfeh wrote:

ZenithPhantasm wrote:

Hi drum drum! o/
hi drum drumerino
pls no lockerino

this is so going to get locked, isn't it?

Anyways, not even sure if game logic runs at the same rate as your framerate, because that reeks an awful lot like bullshit. Rendering != processing unless peppy has some seriously weird code going on there

I'd say the question hasn't been answered yet, so I'm serious with the "pls no lockerino" thing
http://ask.fm/Tom94/answer/126791631310
http://ask.fm/Tom94/answer/126803371470

You can also test yourself by playing an OD10 map with a circle on the middle of the screen every 1/2 beat (preferably with a map with exactly 2 or 3 notes per second), and comparing accuracy and hit error bars between 60fps and 1000+fps.
autoteleology

Full Tablet wrote:

You can also test yourself by playing an OD10 map with a circle on the middle of the screen every 1/2 beat (preferably with a map with exactly 2 or 3 notes per second), and comparing accuracy and hit error bars between 60fps and 1000+fps.
Literally what the fuck. This needs to be stickied somewhere, I feel bad for the poor sap that has been playing this game on 60fps Vsync and wondering why his accuracy is so bad.
bigfeh

Philosofikal wrote:

Full Tablet wrote:

You can also test yourself by playing an OD10 map with a circle on the middle of the screen every 1/2 beat (preferably with a map with exactly 2 or 3 notes per second), and comparing accuracy and hit error bars between 60fps and 1000+fps.
Literally what the fuck. This needs to be stickied somewhere, I feel bad for the poor sap that has been playing this game on 60fps Vsync and wondering why his accuracy is so bad.
Literally what the fuck too, but not for the poor guy whose accuracy is shit because he's playing at 60 FPS.

I mean, if game logic and rendering are together, then it means they're both probably on a single thread, which is kind of retarded. The game logic should be running separately from other stuff (go check this link out for what I'm talking about), and knowing peppy, it probably is (because he's good). Could there be some legacy code going on there though

Anyways, separate things should run on separate threads, and if they are, great. It'll update consistently and will work well. Tying rendering to pretty much anything else is a bad decision because FPS fluctuates a lot, particularly if you don't have a frame buffer (which in some cases, you really shouldn't since buffers increase input lag). If the game really runs like you guys are saying it does, then this needs some serious rework done.


EDIT: wow, holy shit
just read tom on askfm and yeah, turns out the code is retarded. Can't really blame them though, because the game's old. This should still, however, be priority number fucking zero because it's literally game breaking
Saphirshroom

bigfeh wrote:

Can't really blame them though, because the game's old. This should still, however, be priority number fucking zero because it's literally game breaking
You can blame them because, as you've said, it's priority number fucking zero.
Good code is still good in 5 years from now. Which means osu is coded badly.
B1rd
How is it game breaking when there is zero negative impact from this 'bad code' besides low fps affecting accuracy?
Yuudachi-kun

B1rd wrote:

How is it game breaking when there is zero negative impact from this 'bad code' besides low fps affecting accuracy?

Because accuracy is like, important or something
Saphirshroom

B1rd wrote:

How is it game breaking when there is zero negative impact from this 'bad code' besides low fps affecting accuracy?
I heard accuracy is quite important. It's also a very bad kind of game breaking because it is so subtle that you blame it on yourself instead of bad programming.
-sev
Maybe this wouldn't be an issue if people stopped playing on toasters.
B1rd
I said with low fps. Anything over 240 is negligible. Only people with toasters are affected.
bigfeh

Saphirshroom wrote:

bigfeh wrote:

Can't really blame them though, because the game's old. This should still, however, be priority number fucking zero because it's literally game breaking
You can blame them because, as you've said, it's priority number fucking zero.
Good code is still good in 5 years for now. Which means osu is coded badly.
Yeah, I actually rewrote that sentence 5 times because for one, I understand legacy code. But on the other hand, yeah, it's game breaking, retarded and should be fixed asap.

edit:
for you people who aren't programmers (aka normal people), let me put it this way: imagine if you had a car that would only turn if you were flooring it, because for some reason, one day someone decided the steering wheel should depend on the engine to work. This is literally how bad this is

edit 2:

B1rd wrote:

I said with low fps. Anything over 240 is negligible. Only people with toasters are affected.
not true at all. read thread for reason
Saphirshroom

B1rd wrote:

I said with low fps. Anything over 240 is negligible. Only people with toasters are affected.
Oh yeah, let's make a low-end game and completely turn it on its head because game logic is dependant on fps. Let's also not tell the player about it so they can enjoy playing that low-end game on their low-end computers without ever knowing what on earth is going on.
Way to find excuses for badly designed code.
autoteleology
I think I should make a thread about this. This is critical. Should I make it a feature request or something here for everyone to complain in?
otoed1
The answer to this thread doesn't really exist, what defines diminishing returns in this case? Your FPS will continue to have an effect all the way up to the refresh rate of your input devices. What defines diminishing is difficult. Frankly, I don't notice the difference between when I played at 600fps and playing at 1400 fps now. There is an obvious different sub 400 however, and playing at 240 feels like a slideshow.
bigfeh

otoed1 wrote:

The answer to this thread doesn't really exist, what defines diminishing returns in this case? Your FPS will continue to have an effect all the way up to the refresh rate of your input devices. What defines diminishing is difficult. Frankly, I don't notice the difference between when I played at 600fps and playing at 1400 fps now. There is an obvious different sub 400 however, and playing at 240 feels like a slideshow.
Fun fact, you don't actually see the difference between 400 and 240 unless you have a 400Hz monitor, which doesn't exist
show more
Please sign in to reply.

New reply