Do you REALLY need over 9000 FPS to play the game?

I admit, I'm kinda FPS-obsessed, but that's only because I emulate some games (Pokemon B/W, GC Twilight Princess, etc.) - and that's where I can see significant improvement even over the platform it was designed for - like in Pokemon games, the gameplay is faster, and the text in Ace Attorney games is scrolling with proper speed. But for Xonotic? Most commercial games target 60 FPS, maybe 100 FPS tops. They do, however, motion-blur between the frames, so the outcome looks extremely well. I've seen it done in Source games - and they look awesome. So, instead of trying to get 500 FPS, why not settle for 100 FPS and try to optimize everything for that target framerate? It wouldn't improve the gameplay, but the quality would be significantly better.

What do you think?

You can't optimize a PC game for 100fps because different PCs' performance can differ by a few orders of magnitude.

i still agree on Minkovsky. How is it even possible to see a difference between 75 and anything above 75 FPS from a game when your TFT monitor updates at a frequency of 75 Hertz (75 updates per second)?

I've heard that the difference between 60fps and 120fps is the double mouse polling rate. Information is updated twice as fast(better sniper shots ). At least, I believe this was true with Nexuiz.
For realz though, part of the issue comes from inconsistencies in the framerate. If the target is 60fps, then the lowest rate under worst conditions will probably be around 15fps. On the other hand, if you shoot for 120fps, the lowest will be 30fps, far more manageable. Also, at the current moment, most mappers are limited by the number of realtime lights, models, and reflections they can put in a map. That's not a good thing, optimization(especially LOD which I one of my priorities) frees them to create better looking maps. There was an old article on this. Having your framerate at 45 and having it drop to 20 breaks flow, but having your framerate at 100 and having it drop to 60 does not. In a game as fast as Xonotic, that's important. Keep in mind, the faster we get the framerate, the more effects that can be added.

maybe because if you perform 75 fps on average, it's likely you will perform less face to a complex scene.
And you then might notice it with your 75 Hz TFT.

Above 100 fps, fps obsession has somewhat similar to penis obsession.

I would agree with naryl. Targeting a certain value might work for console games, but with Xonotic you would first have to define on what kind of hardware/software setup you want to base the target FPS.

If you have too many FPS (what the...), you can always add more effects (antialiasing, realtime light, offsetmapping, motion blur).
I don't think you'll still get 500 FPS after that.

No, I don't need 9000fps but I do need 100+ fps. In a fast paced first-person shooter where you play with a mouse, it's definitely not the same as playing on a xbox with a gamepad.

FPS is not the same thing as your monitor's refresh rate. FPS is how many times per second the engine updates, receives input, etc. Basically, the more fps, the better response time the game will receive input when you press a key or move your mouse. Or course, there's many other factors (hardware, mouse, etc., also the fact the server is capped at fixed fps so what you see if not the same as was server sees), but that's basically why people want lots of fps, because it "feels" better. However if you don't use a gaming mouse with fast refresh rate, I don't know the maths behind it but I would assume you're not gaining much by having 300+fps.

In any case I doubt anyone could notice a difference between 300fps and 600fps, I would say it's mostly happening in their head. Anyway, personally anything over 200fps feels the same. But the difference between 75fps and 200fps, you can bet your ass I can notice. Anything under 90fps feels pretty bad to me. I definitely wouldn't play with less than 70fps (because then it not only "feels" bad, you can actually see it visually if it randomly drops below your monitor refresh rate), unless it were only to explore a map filled with a lot of eyecandy or test high quality settings.

Mind you, I know there's many players that are very good even with low fps and a cheap mouse, so having better fps doesn't instantly make you better. You get used to whatever you can have at hand.

YESS YOU DO!!! lol seriously 46fps and higher should be playable for most, I personally cap it at 63 and I play on a CRT.

It may be playable, but not optimal, I never cap my framerate, let my GPU's do their job

Yea, I definitely like to have a way high framerate. Visuals can never be too smooth

This thread makes me sad ;_;
I need at least over 30 fps to play the game, which I'm hardly getting on a lower-than-omg setup. And besides, I feel zero difference between 60 and 200 (I can compare on another machine which I rarely get to use).

https://secure.wikimedia.org/wikipedia/e...frame_rate

As Akari said, the benefit bought by each additional frame/second just about levels off at 60fps. And framerate is just the speed at which the scene can be rendered, you still need to take into account ping, packet loss, mouse and keyboard responsiveness, monitor resolution, monitor refresh rate, monitor response time, sound quality, sound response, etc, etc, etc.

Personally I want at least 60 fps for dp to feel good. As nifrek pointed out above, far more then screen refresh depends on FPS in darkplaces (since its single threaded). Whats more important, for me, tough is as stable fps (and thus stable updates) as possible. When i play "seriously" i setup the game to run at a steady 120 fps.

Another common reason to want over 60 fps is "framedrops" eg periods when your fps dips far more then usual. If your rendering at 200 fps and you have a 50% drop over 5 seconds, your still rendering above your screen refresh. If the same thing happens with 60 fps, your stuck with half screen update speed for this duration. Of course a momentary drop in render speed is less likely to cut 50% in the lower ranges then it is in the higher.

What tZork said.
If I had consistent 60fps, no problems. It's my monitors refresh rate, so no frame intepolation and visually nice and smooth. Problem....it's not stable, and courtesy of the way DP works, it most likely never will be. So It's better to overspec and then have the framerate drop be 60fps instead of 15. That massive drop to 15 can seriously break flow. I also cannot stand 30fps. It seems far too jerky without any good motion blur. It's why I hate playing FPS on a console. It does not feel smooth.

For a game like Xonotic, I'd say 60 FPS at least. I tried capping both Nex and Xonotic at 30 FPS and the player movement looked like a slideshow. Everyone's eyes are different though... I showed a friend of mine the difference between 60 and 30 FPS in Nex and he could not tell the difference. When I dropped it down to 15 FPS I got a "whoa!". Personally, I can look at a game running and tell you what range it is in between 0-90 FPS, give or take. AssaultCube was quite playable at 30 FPS.

Funny story, if I cap the rendering frame rate below the value of the refresh rate (accounting for frame buffers) while vsync is active, I get little to no input lag.

Lastly, I think motion blur in Xonotic needs to be reworked to look better at <200 FPS. Much ghosting. I played the console version of Resident Evil 5 (don't hate me) and it was chock full of motion blur effects, which were obviously optimized for the 30 FPS margin, and they looked amazing. I'm sure Xonotic could have even better effects.

I cap my fps at 125 in both nexuiz and xonotic for consistency. I've tried different framerates, from 300 to even 30, and I have to say that in my experience, 125 is the best if you are looking for smooth gameplay while still keeping consistency with your mouse. A lot of people, including one of my clanmates, like a lower frame rate, from 60 to even 40 for it's jagged, rough feel. I think it's a matter of taste, but to be safe I would use at least a 60 cap and no lower to avoid loss of mouse input that otherwise might make a difference. Also, although I'm not entirely sure if this is true or not, but I have heard that keeping your net fps (cl_netfps) synchronized with your framerate helps also.

I'm going to screw with my settings and see if I can get fps to stay capped at 1337. lol

i run 40-70 fps , works great

then again my fov is at 1 too....

yea im hawt , ya know it :p


:^

Most screens can't even display anything over 60FPS, so why bother?

Because the Framerate is never consistent. It may be a smooth 60 now, but when it drops to 45, it's noticeable. When you shoot for a target framerate of 100-120, the it may only drop down to 60, in which case, no change is noticed.

No, I don't need 9000fps but I do need 100+ fps. In a fast paced first-person shooter where you play with a mouse, it's definitely not the same as playing on a xbox with a gamepad.

FPS is not the same thing as your monitor's refresh rate. FPS is how many times per second the engine updates, receives input, etc. Basically, the more fps, the better response time the game will receive input when you press a key or move your mouse. Or course, there's many other factors (hardware, mouse, etc., also the fact the server is capped at fixed fps so what you see if not the same as was server sees), but that's basically why people want lots of fps, because it "feels" better. However if you don't use a gaming mouse with fast refresh rate, I don't know the maths behind it but I would assume you're not gaining much by having 300+fps.
...

I hardly can convince myself on this. Mostly because you can connect to a server at 20.000 km away from you, then have a 200 ms ping (so having a 5 Hz rate) and still keep 300 fps.
Fps and global response time seems rather unrelated.
Also a cheap mouse can deliver some data no less than 125 Hz. So higher than a confortable fps. So mouse signal should'nt cap fps rate neither.

If someone have precise inputs on this, i take them.

Most screens can't even display anything over 60FPS, so why bother?

Again, because if you're GPU is able to process 60 fps _on average_, you will fall under this rate over complex scenes. So you'd better have extra fps to face these cases.

I hardly can convince myself on this. Mostly because you can connect to a server at 20.000 km away from you, then have a 200 ms ping (so having a 5 Hz rate) and still keep 300 fps.

Sure, but ping is unrelated to this, we're talking about input responsiveness from between your hands and your computer. Ping would be an entirely different subject (which would be interesting too).

But try to imagine if your monitor had a refresh rate of 125hz. Then if you capped your fps to 60, you can suppose that visually it not look as smooth as if it was capped to 125fps.

Your monitor shows you the maximum frames per seconds it can, but your computer still allows more behind. If your hardware allows it, then input from mouse can "fill" all those frames, hence more responsiveness. Obviously it's more complicated than that, but I'm trying to give a better explanation. Basically, fps is not just frames shown on your monitor, it's also frames containing packets of data from input you're sending to the game.

In real life when you physically move your mouse on your desk, you can consider it infinite fps (unless someone wants to calculate that? ). So with low fps, the game will approximate the "cursor" position as close as possible to what it is physically doing, but when doing very fast motion then it simply cannot be as precise as high fps, it'll just try to fill in the blanks and smooth it out as best as it can. It's not even only about latency but precision.

Another way to explain it, watch a movie, press pause during a fast camera movement. It looks fine when played back, but when on pause the picture is blurred because there is not enough frames per second to have all the information to have a sharp picture (bare with me, I know about shutter speed and blahblahblah, I'm just trying to give another example to explain this without getting too technical). The idea is similar, but replace images with data.

I'm very tired, hope this makes sense at all.

EDIT: btw as with everything, there's nothing wrong with being skeptical, that's how you figure things for yourself. If you cannot ever "agree" with this because you personally cannot tell the difference, then more power to you, that's one less thing you have to be bothered about I can agree this is overkill anyway, it's mostly hardcore gamers that will play a game for hours everyday until they know it so well that they can tell tiny differences that others might not even if they play regularly. I mean, some players will go as far as figure out the perfect room temperature for their hands (with hours of experimenting) to have the best performance possible when playing a serious match, so... yeah.

Sure, but ping is unrelated to this, we're talking about input responsiveness from between your hands and your computer. Ping would be an entirely different subject (which would be interesting too).

But try to imagine if your monitor had a refresh rate of 125hz. Then if you capped your fps to 60, you can suppose that visually it not look as smooth as if it was capped to 125fps.

Your monitor shows you the maximum frames per seconds it can, but your computer still allows more behind. If your hardware allows it, then input from mouse can "fill" all those frames, hence more responsiveness. Obviously it's more complicated than that, but I'm trying to give a better explanation. Basically, fps is not just frames shown on your monitor, it's also frames containing packets of data from input you're sending to the game.

....

I'm very tired, hope this makes sense at all.

Sure, it does Hope you slept well
This problem is known as asynchronous multi-sensing.
Typically, the slowest sensor caps the overall effective system rate.

What i tried to point out is that the only factor which caps the fps seems to be the GPU.
If you use a 80's mouse (let's assume here they had a 3 Hz rate), i think (correct me if you know i'm wrong) you still could reach 300 fps. But, a lot of these 300 frames will use the same (or poorly interpolated) mouse signal (1 mouse sample for 100 frames in my example on average).
Of course, i'm not saying that you would have a smooth gameplay with such a mouse.
But I just feel that the fps counter is rather independent from the mouse signal quality or any other signal than the graphical unit capability itself. Again, correct me if i'm wrong.

EDIT: btw as with everything, there's nothing wrong with being skeptical, that's how you figure things for yourself. If you cannot ever "agree" with this because you personally cannot tell the difference, then more power to you, that's one less thing you have to be bothered about

Trying to humbly understand is not being skeptical btw.
I just everyday try to go to bed a little bit less dumb than yesterday.
When i will become skeptical, you will notice it

I can't remember the science behind this right now, but:

Your monitors refresh rate*2+1 is the best fps-cap.
If your monitor is 60hz, the ultimate fps would be 120. And then when you cap you add 1.
This is the ultimate, of course, your refresh rate*4+1 would probably be even better, but that's exaggerating.
I should also mention, that 60 fps on a 60 hz screen is better than 80 fps on a 60 hz screen. (You should look out for fps-drops, they're the worst)

How important your fps is is as many things highly subjective. I can not enjoy any game if my fps is below 50.

So people that have a 60hz screen but are running at 500 fps, try capping it at 121. In theory, that should be better. Worst case scenario is you'll get rid of fps-drops. And those are bad for you. They mess up your input.