Page 2 of 7

Re: The performance of the Spring engine

Posted: 16 Apr 2008, 23:40
by smoth
I have tried everything I can. Including getting a new video card, dual core processor, setting spring to 1 core and doing that settings crap in the nvidia control panel. I maed several threads about it :(

Smoth is a sad panda
Image

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 00:23
by Ixoran
14 tanks, 3 cars, 11 hovercrafts moving with AI + 22 tanks onscreen standing standing still. FPS = 37
reivanen wrote:And no, it won't start lag even if i set 100 tanks in there.. probably not with 1000 either. There was no performance difference before or after placing the models or making them move. If you will keep missing the point, i could redo the test with a few hundred tanks...
Wait. Since when were FPS games designed to run at that framrate? Last I checked FPS and Racing Games (and games in general) were supporting 80 Frames a second.
Like... the average the human eye can keep up with.

So unless Crysis had some really weird design goals, last I checked, 37/80 IS LAG.
Now, Crysis (with it's 27 million dollar budget) might no let you know it's lag, By chopping out some frames, smoothing out the differences and not letting the input catcher fall behind, but 37fps is terrible for a fast paced game last I checked. I mean F-zero GX ran at 80-100 fps.

Spring is in the process of leaving it's hacked abomination phase and (I hope) entering a phase where it gets optimized and generalized to not be designed for TA.
That's my answer for the performance gap.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 00:29
by smoth
refactoring hackish code is a nightmare. That is the reason it is taking soo long. However, at least we have something. I am pleased with it at least being here. IF ONLY I could run it at speeds faster then old people fucking in between their favorite soap operas.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 00:33
by Hobo Joe
Ixoran wrote:
14 tanks, 3 cars, 11 hovercrafts moving with AI + 22 tanks onscreen standing standing still. FPS = 37
reivanen wrote:And no, it won't start lag even if i set 100 tanks in there.. probably not with 1000 either. There was no performance difference before or after placing the models or making them move. If you will keep missing the point, i could redo the test with a few hundred tanks...
Wait. Since when were FPS games designed to run at that framrate? Last I checked FPS and Racing Games (and games in general) were supporting 80 Frames a second.
Like... the average the human eye can keep up with.

So unless Crysis had some really weird design goals, last I checked, 37/80 IS LAG.
Now, Crysis (with it's 27 million dollar budget) might no let you know it's lag, By chopping out some frames, smoothing out the differences and not letting the input catcher fall behind, but 37fps is terrible for a fast paced game last I checked. I mean F-zero GX ran at 80-100 fps.

Spring is in the process of leaving it's hacked abomination phase and (I hope) entering a phase where it gets optimized and generalized to not be designed for TA.
That's my answer for the performance gap.
Basically no consumer-level PC can run Crysis at 80 FPS at medium settings, much less high settings.

As for the 'standards' of FPS framerates, it's nothing solid. 60 FPS is the standard refresh rate for LCD monitors, and it's a rare CRT that goes over 75. Your eye won't notice a difference between 50 and 80, anyway. (And before people go crazy on me, I KNOW that they can see a frame that lasts less than an 80th of a second, but when they're all strung together you won't notice the difference).

On top of that, Crysis uses motion blur, which reduces the 'stutter' between frames. And besides, Crysis isn't a twitch multiplayer shooter like Counter Strike, for example, where you really WILL want to have every possible frame in order to get perfect precision and response.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 00:45
by Felix the Cat
80 rendered frames per second is a waste of processing power. As Hobo Joe pointed out, you will not be seeing 80 frames per second, you'll be seeing 60 or 75.

For comparison, 35mm movies are displayed at 24 frames per second, European TV is displayed at 50 frames per second, and American TV is displayed at 60 frames per second. (Movies are actually displayed at 24 FPS but have a refresh rate of 48 or 72 hertz. Each frame is illuminated 2-3 separate times by the backlight, but during those 2-3 illuminations the frame doesn't change.)

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 00:56
by SwiftSpear
Felix the Cat wrote:80 rendered frames per second is a waste of processing power. As Hobo Joe pointed out, you will not be seeing 80 frames per second, you'll be seeing 60 or 75.

For comparison, 35mm movies are displayed at 24 frames per second, European TV is displayed at 50 frames per second, and American TV is displayed at 60 frames per second. (Movies are actually displayed at 24 FPS but have a refresh rate of 48 or 72 hertz. Each frame is illuminated 2-3 separate times by the backlight, but during those 2-3 illuminations the frame doesn't change.)
Every film and TV technology uses motion bluring to mask the noticeable difference between moving frames. Games don't. motion blurring is expensive and unfeasible for things rendered in real time. The human eye can barely conciously pick up the difference between running 50 fps and 80 fps... but the instinct centers use visual information more intimately than the conscious does. High end FPS players will chock down settings if it means the difference between 80fps and 100fps. I can speak from experiance, that difference is massive when you're doing something heavily relying on instinct. I used to get pissed off while KZ jumping if my FPS dropped to 85, because it significantly degraded my ability to perform the required motions.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 00:59
by LordMatt
Could you see a frame counter at the time? I bet it was placebo effect :P

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:12
by imbaczek
The effect is real and noticeable. Human eye doesn't work like computer displays, it's asynchronous, so saying that one can't tell the difference between 60 and 120 fps isn't quite true. Maybe it looks the same for the untrained eye, but definitely not if you know what to expect. Ask anyone who's ever played Quake competitively.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:18
by Ixoran
and it's a rare CRT that goes over 75.
My old ass KDS CRT goes up to a refresh rate of 150.
I think it tops out at 80 for 1940x1440 Though.

There was a reason I used F-zero as an example, Because F-zero play fast. It doesn't just look fast.
Which is why People trying to emulate F-zero tend to BAWWWWWWWWWW all day.
Because it can make a huge difference.

My main point was that Why are we even having a $27,000,000 v $0, Bleeding Edge v Borderline Hack, Professional Team v Freetime team debate?
The effect is real and noticeable. Human eye doesn't work like computer displays, it's asynchronous, so saying that one can't tell the difference between 60 and 120 fps isn't quite true. Maybe it looks the same for the untrained eye, but definitely not if you know what to expect. Ask anyone who's ever played Quake competitively.
That's basically what I'm talking about. Fast FPS is way more noticeable with FAST play-style Games. I would find it hard to justify any RTS as needeing the incredible reflexes of Burnout, F-Zero, or a good chunk of the FPS's out there.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:22
by smoth
Ixoran wrote: I would find it hard to justify any RTS as needeing the incredible reflexes of Burnout, F-Zero, or a good chunk of the FPS's out there.
RTS Gamers can fire off a lot of commands, in many mods/games MICRO is a huge part of what it takes to win. If you have ever played with the uber-ninja-god-fuckwin players you would see this sick amount of commands in action.

everything from camera movement to attack orders is important in an rts.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:26
by Ixoran
smoth wrote:
Ixoran wrote: I would find it hard to justify any RTS as needeing the incredible reflexes of Burnout, F-Zero, or a good chunk of the FPS's out there.
RTS Gamers can fire off a lot of commands, in many mods/games MICRO is a huge part of what it takes to win. If you have ever played with the uber-ninja-god-fuckwin players you would see this sick amount of commands in action.

everything from camera movement to attack orders is important in an rts.
I 'm not saying that isn't true.
But think for a second. Due to the way RTS combat is done, You can get all the command you need into 30FPS or around that.
For a lot of FPS and racers, Input is being checked just as often as graphics, so a 1/80ths of a second joystick is totally different from a 2/80ths(or 1/40th mathfags) of a second joystick twitch.

Now I know how those Koreans play Starcraft, But it's a totally different ballgame than watching a hardcore FPS or racing play and the seizures them have over their controllers.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:35
by smoth
nope it isn't, rts gamers need high fps also. I doubt fps gamers need more kps then an rts gamer. When you are moving a SEVERAL units to dodge a laser or cluster bombs it does matter.

While I do not feel it equates to skill but KPS of a player is one of the things people rate each other by.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:41
by Ixoran
Most FPS players should have 0 kps.
They don't use the keybaord.
I think you might be confusing input speed and graphics speed.
I mean, if the keyboard is being check 80 times a second, but the game only runs at 30 fps, Thats really all you need for an rts.

You don't think so?

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:52
by smoth
I am sorry for questioning your wisdom how dare I suggest a fps player use a keyboard let alone a mouse :P.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 01:54
by Felix the Cat
SwiftSpear wrote:
Felix the Cat wrote:80 rendered frames per second is a waste of processing power. As Hobo Joe pointed out, you will not be seeing 80 frames per second, you'll be seeing 60 or 75.

For comparison, 35mm movies are displayed at 24 frames per second, European TV is displayed at 50 frames per second, and American TV is displayed at 60 frames per second. (Movies are actually displayed at 24 FPS but have a refresh rate of 48 or 72 hertz. Each frame is illuminated 2-3 separate times by the backlight, but during those 2-3 illuminations the frame doesn't change.)
Every film and TV technology uses motion bluring to mask the noticeable difference between moving frames. Games don't. motion blurring is expensive and unfeasible for things rendered in real time. The human eye can barely conciously pick up the difference between running 50 fps and 80 fps... but the instinct centers use visual information more intimately than the conscious does. High end FPS players will chock down settings if it means the difference between 80fps and 100fps. I can speak from experiance, that difference is massive when you're doing something heavily relying on instinct. I used to get pissed off while KZ jumping if my FPS dropped to 85, because it significantly degraded my ability to perform the required motions.
Unless you had a non-standard monitor, your monitor only refreshed at most 75 times per second.

Hence, it would be impossible to see more than 75 refreshes per second.

edit. FPS players use psychic powers, everyone knows that.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 02:10
by Ixoran
Hence, it would be impossible to see more than 75 refreshes per second.

edit. FPS players use psychic powers, everyone knows that.
1. All I know is my monitors display config button will go all the way up to 150 at 800x600
2. I agree!
/thread.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 04:49
by Hobo Joe
SwiftSpear wrote:
Felix the Cat wrote:80 rendered frames per second is a waste of processing power. As Hobo Joe pointed out, you will not be seeing 80 frames per second, you'll be seeing 60 or 75.

For comparison, 35mm movies are displayed at 24 frames per second, European TV is displayed at 50 frames per second, and American TV is displayed at 60 frames per second. (Movies are actually displayed at 24 FPS but have a refresh rate of 48 or 72 hertz. Each frame is illuminated 2-3 separate times by the backlight, but during those 2-3 illuminations the frame doesn't change.)
Every film and TV technology uses motion bluring to mask the noticeable difference between moving frames. Games don't. motion blurring is expensive and unfeasible for things rendered in real time. The human eye can barely conciously pick up the difference between running 50 fps and 80 fps... but the instinct centers use visual information more intimately than the conscious does. High end FPS players will chock down settings if it means the difference between 80fps and 100fps. I can speak from experiance, that difference is massive when you're doing something heavily relying on instinct. I used to get pissed off while KZ jumping if my FPS dropped to 85, because it significantly degraded my ability to perform the required motions.
Motion blur is an increasingly common thing to see in new games. It's been added in the Source engine for all the newest games like Portal and TF2 and EP3, and is supposedly going to be added to all the older Source games. Crysis has it as well, and I believe COD4 does.

As for the difference between 80-100FPS... you won't notice it. If you have an FPS counter on it's just placebo. And as I mentioned before, LCD's only refresh at 60 FPS, and CRT's are usually in the 60-80hz range.
nope it isn't, rts gamers need high fps also. I doubt fps gamers need more kps then an rts gamer. When you are moving a SEVERAL units to dodge a laser or cluster bombs it does matter.
RTS gamers will of course perform better with a better framerate, but it's simply not the same as an FPS game. The difference between 40 and 50 FPS can be the difference between a kill and a death, because it takes twitch speeds and perfect precision to do that, and thus needs a good framerate to pull it off.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 05:22
by smoth
look, believe what you want. RTS gamers with high amounts of unit micro need to be precise and fast as well. Don't bother arguing it, I won't argue with you about it.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 05:53
by Ixoran
Ixoran wrote: /thread.

Re: The performance of the Spring engine

Posted: 17 Apr 2008, 07:14
by Vadi
I don't believe anyone can see a difference between 40fps and 50fps.

Someone should do a blind test to compare just to see what others see 8)