Afaik 0 defaulted to 3 in almost every game, while 1 is the lowest value

Defaults are really not that interesting imho. I clearly feel the difference between 0 and 1.

i'm saying that when you put 0 it's actually 3 in most games
Search the internet, there's lots of research about it
P.s. you can still set 0 with nvidia inspector

You are not making any sense. Only ways you can change pre rendered frames I know of is in inspector and drivers. When you change it there its actually what you changed it to.

English comprehension.

When you set the value of "Max Pre-Rendered Frames" to "0" in nVidia control panel, it has the same effect as setting it to "3", which is the default. Of course you will still feel a difference changing it from "0" to "1" because you are, in essence, changing this value from "3" to "1".

Capiche?

Thats simply not correct. In general you will get what you set in inspector, but it depends on which app.

I always used 3 pre-rendered frames, is it bad? :))

lol i have 3 too, lets see if 0 makes any difference

Depends on game. In d3 I get stuttering with 0. 3 works fine. Also you can get slightly better performance with higher pre-rendering of frames but its small it shouldnt matter with a modern system.

1000/125 = 8

8 ms frame time = 8 ms delay on 125 fps, even assuming you're right.

5th grade maths...

Fact is, you have to "pre-render" a single frame before displaying. So you're suggesting that this option is "how many _more_ frames to pre-render" and 1 is keeping a single frame always in the buffer while previous one is displayed and next one is being rendered? Does it do it even with vsync off? I was under impression this setting is for vsync users...

The thread you're linking... lack of words, people saying lower pre-rendered frames cause performance drops? And you want serious ESR citizens discuss with that? No thanks.

watch_out_everyone_weve_got_a_bad_ass_that_knows_everything_over_here.jpg

To the contrary, I've asked for explanation of this setting and confirmation if it works with vsync off.

Besides, h8m3 also seems to think 0 was ineffectual before.

Unless you want to dispute with basic maths? Or you think pre-rendered frames increase fps for games like Quake which can render 500 fps on my hardware?

watch_out_everyone_weve_got_a_bad_ass_that_knows_everything_over_here.jpg

If you got a proper answer, feel free to share it.

watch_out_everyone_weve_got_a_bad_ass_that_knows_everything_over_here.jpg

only 500? I hope your resolution is over 9000 both y and x to get that low. Esssh :S #gameboycolorlyfe

:)

Always funny someone is whining about something he is doing himself (being misinformed in this case).

"The thread you're linking... lack of words, people saying lower pre-rendered frames cause performance drops? And you want serious ESR citizens discuss with that? No thanks. "



You obviously haven't done your homework, "serious ESR citizen".

Somewhere in your "5th grade maths" riddled mind you must have realized that actually having experience or testing something might give you an informed opinion on the subject, instead of just assuming "serious ESR citizens" all think, and are as quick to ignore fact and research, as you do.

but isn't this only for d3d games?

It is

Link please.

you dumbass

dat placebo effect

But believing everything that you cant experience isnt a placebo effect?

no, by definition it isn't.

I can choose 0 for ogl games too, so that doesnt prove anything.

You can, but it won't change anything since it only works with DirectX games.

So...ql is OGL ? :D

QL uses OpenGL.

so most of the settings like anisotropic filtering etc are practically useless for quakelive? since its open gl i mean?

I don't know if things changed since then. That's quite the necro you pulled off here.

quakelive still uses openGL 1.4 which doesn't support this unless they use the extension for it.

no idea why you would bother to use those in QL anyways...

I simply bought a computer and shouldn't of because i dont understand how to use it fully.......

its only ever as capable as the people who develop for it- when quake was made for 1999 hardware it could not support all these new things that we have today

Some good info on input lag and max pre-rendered frames

http://www.anandtech.com/show/2803/7

good thing ql runs at 120fps and you can disable v-sync ;)

Forcing this setting to 1 gives me a tiny bit of jitter... i've had it set to 2 for a couple of years now. I don't notice a significant difference between 2 and 3 respectively.

cant see any difference between 0, 1, 2 or 3

i heard rapha uses 5, so he can be 5 steps ahead

I heard fox uses a motherboard that allows him to kill people in other games.

Rumors say there is a game out there that "1" defaults to a picture of a chair

this setting only affect DirectX games. Not OpenGL games...

And again I ask for proof.

prove that it works with opengl

Why would I, theres a setting in the drivers with pre rendered frames. It doesnt say for which type of games. Unless you can actually prove your claims (and Im not saying you are wrong) logic dictates what you are claiming is pure fiction.


And on that note, I hope they will end anonymous posting on this site. It seems to bring out trolls into the light :/

Lol, didn't you look at the picture posted by comrade in comment #19?! The last point in the screenshot says "this feature only works with games using DirectX". Quake is a OpenGL game so this setting won't have any effect on it. You can set it but it won't do anything.

Yes I did miss that.

the picture posted above states it's for directx games, you're the one with burden of proof

How many pre-rendered frames does opengl (quake live) use?

nVidia's OpenGL implementation has triple-buffering but:
1. it only works with vsync iirc,
2. it doesn't add any output delay above the vsync delay (it discards frames that it doesn't need to render). This is unlike ATI.
So it doesn't affect people around here because we love tearing and we frown at any notion of "vsync on".

So that means ql has 0 pre rendered frames? (sorry too dumb to understand)

It is true that even in DirectX games, setting pre-rendered frames to "0" in the nvidia control panel has the same effect as it being set at 3.

I believe they are removing the option for 0 because it was redundant.

You can test the 4 settings (0,1,2,3) yourself and see what difference it makes, if you need proof.. your FPS is impacted by this setting as well, so it is easy to see 0 reacting much in the same way as 3.

I have tested fps on different settings and its not affected as far as I can tell (diablo bfbc2). Some games override it when you set it to 0 and vice versa.

yes, bad company 2 has a setting that it uses in it's config file for this, probably diablow as well.

test it in a game where you get high fps and you will see the small differences that the different settings make. 0 and 3 react pretty much exactly the same.

Over the 4 tests I conducted I noticed lower fps at the start of the game but after ~30-60 secs they stayed about the same. I could have been more thorough but I dont consider > 5 % noteable. Cause its mostly not humanly noticable and therefor not practicaly usable.

Remember, this test was not for optimization, it was to prove a theory. The whole 0 and 3 thing.. do you not remember?

You obviously have issues with properly testing and recording data... and if's that's the case: either learn how to properly.. or don't even try.

kthx

it took him days to notice the attached picture. It'll take him weeks to get the whole 0 and 3 thing and that all this time it was placebo on directx games.

I tested it and felt the difference. I guess you didnt notice it. I guess calling it placebo cause you might or mihgt not feel the difference doesnt make it so. Just like I can easily feel the difference between 500 or 1k hz on the mouse. But hey, if I want some more whining and bitching at least I know whom to ask!

I dont know why you dumbass keep insulting me. You have contributed nothing at all to this thread except whining at people.

So let me explain this to you. Im too tired to explain this at a 2 year old level so lets see if you can make sense of the following.

The guy im replying to says "You can test the 4 settings (0,1,2,3) yourself and see what difference it makes, if you need proof.. your FPS is impacted by this setting as well, so it is easy to see 0 reacting much in the same way as 3"
He is claiming 2 things here. However they are in opposition of eachother, what gregory bateson might claim to be a double bind. "if you need proof.. your FPS is impacted by this setting as well, so it is easy to see 0 reacting much in the same way as 3". Of course I could mindread this and try to intepret it as if that sentence would make sense any other way than it says. But I dont take stuff like that for granted.
I however have tested this (actually with a shitload (150-250 in both games) of fps, I have no idea why you think that might make a difference or why you ass-ume I dont have that to begin with) and say that fps is not really affected by pre-rendered frames. If you however have actual data showing that pre-rendered frames affect fps, please show us instead of just more what you are good at, whining and speculating. Also I cant remember seeing I told you guys how I measured, so how the hell can you bitch about my data?

You claim that diablo 3 "probably" has an inifile to set pre-rendered frames. If you are going to continue speculating maybe you should at least stop insulting me for not handling data correctly.

jesus you are insufferable. crawl back into your hole already. <3

You keep talking about how you need to explain things in a "2 year old level" and perhaps one could read this a bit further and surmise that this is how you explain things to *yourself* to help understand the world, as well. Only then does your confusion with the english language make sense.

Anyone who has read this thread can clearly see the extent of your ability to comprehend the english language, yet you use this language to tell people *they're wrong* so much. I am not going to waste any more time explaining anything to you and I am going to stop posting here now. Good luck with your shit.

Thanks for yet again proving you are your name, and nothing more.

afaik there's an option for this at opengl games in nvidia inspector starting with 3**.** drivers, but i haven't tried them yet

In the 300-series drivers you can set this on opengl as well. I havent been able to test this but it has been confirmed from several sources.

Also ManuelG from Nvidia has acknowledged that there is a difference between 0 and 1.

A difference of 1?

There has been some speculation that 0 either defaults to the games setting or that the minimum is 1. Oh I see now that h8m3 has already mentioned this. Oh well.

I'm using 301.42 and that section no longer says it only works for directx games. Also only 1-4 and app default.

what happens when u set it to "app controlled" 1? 4?

so the minimum option being 1, this means it would add 1 frame of input lag by default?

that's unacceptable for fighting games with 1-2 frame link combos, especially with the added delay from online play. remind me not to buy nvidia when i upgrade my PC.

Amd has had minimum 1 for several years.

where? can't seem to find the setting in ATI Tray Tools..

exactly

It's under 3D, it's called Flip Queue SIze for ATI.

it's set to 0.. hopefully it's for the best

Mine was undefined, now 0. Thanks, having 1 frame forced is a real step backwards :(

Set it to 0 reboot and check what the number is now. It just sets a reg entry Im not sure it actually does set it to 0. But I havent checked amd for years I guess they might have changed it.

welcome to the world of input lag. Every single device involved has it, you are just used to it.

:<

This is were we need syncerror/id to come in and go 'this is how it is blah blah blah /thread"

Then we copy that shit to the wiki.

drivers for hardware is actually one of the topics that I know nothing about :D

"...we need syncerror/id..."

"By Netherlands Weird - Reply to #60
drivers for hardware is actually one of the topics that I know nothing about :D"

...eh?

it's funny if you know some previous threads ;)

"A real competitive FPS gamer doesn't whine about input lag."

That's when I closed that page. FYI, that's the 2nd reply to the OP in that link.

if prerenderep frames was possible you would already have the input lag of the original first buffered frames and 10mhz gpu less while it reads stored frames and displays the current

Are you working to fix the wsw server browser bug?

no, i dont play the game, if i did, its still a no, they can do it themselves

do you play quake4?

no, dont play no online games, i only have 3g connection, no landline

So you moded q4 heavily, while not playing. But the reason to not fix the wsw server bug is that you dont play it?

tv is shit, i have seen all the documentries, there is 3 or 4 hours worth watching here and there and a movie, Q43A killed the rest of the bore time

most games of today are shit off or online, so it does`nt matter, putting up with BT definately aint worth the hassle just to play some wack game

wipeout is dead, EA should release a proper track editor for need for speed games with aload of prefab buildings since carbon/most wanted, and option to make your own lwo/obj building and import them

rocking http://www.youtube.com/watch?v=f7WJgrJ2oWk

what pro wouldnt bash together a track in a few hours here and there for the sake of it

http://www.youtube.com/watch?v=V2WeeL8EFrA
http://www.youtube.com/watch?v=DvQi2LzUouY

just simple a -> b tracks, no city blocks etc like final races etc, or they would never need to release another retail game for 10 years with a massive engine update

agreed

they have http://world.needforspeed.com/ free 2 play MMO, EA failed with the original engine, and all games only had night tracks, they use battlefield 3 engine for latest good photo realisic looking games...

they need to make a editor, dropping together 3096 square grids, to make the base run/track layout, raise and lower grids if your a noob, or export to obj for greater detail, import the obj again, and make your monster A -> B rural run/tracks, with sweet racing lines, manual gear bumps, specific to certain cars and time trails

Maybe a little OT but:

I use texturemode GL_NEAREST instead of the original. It may be placebo but the aiming seems to be more crisp than the origina. And I like the lego look with picmip 6...

Is there any good practical way of measure this kind of effect?( without high speed camera)

you might just have smoother fps with that setting. if you select "fastest" under graphics settings in the ql menu it uses that, so i guess it's lighter.

Apart from 1-4 pre rendered frames there is also an option "defaulted by app" or smth. Can anyone bring the light on how this one works? From what i heard if a game has some sort of its own option for prereredering frames it works separately from nvidia values. So just curious how they affect each other if specified by apps is set..

value of "1" is a safe choice anyways since most game-engines cant read value of "0" and force to render 4 frames automatically (default). I have readed about it once, cant find the source though. but I would say that you can quite feel the results ingame if you have very high overall input lag by your monitor

Found this thread - https://forums.geforce.com/default/topic/6288.../?offset=9 which links to http://www.anandtech.com/show/2803/6 Anandtech reported a difference of 13ms between prerendered frames 0 and 1.

As for whether it works in xp i have no proof for that besides feelings but i'm pretty damn sure before nvidia removed the 0 option in xp it never said anything about directx only. when they removed it and limited it to 3, it said directx only then when they brought back prerendered frames 1 it removed the line of text that said directx only so that's pretty strange. I am currently trying to get in contact with an nvidia graphics driver developer to ask if 1 is effective in opengl so if it does maybe it did used to work in xp.

Interesting,keep us informed if u would recieve any feedback

i asked live chat and they were useless so i asked the forums but nobody answered. I was thinking about contacting the radeon pro developer but i already found a thread from him saying he doesn't know whether or not it has any effect and amd won't answer him because it is an unsupported feature.Nothing i can do i'm afraid but one day i will buy a gtx260 with old drivers and test things. I can't right now because i completely changed my setup and unless i feel very confident with my aim it's a bad testing methodology and regardless no one here will consider it proof but i just want to convince myself if it is placebo or not for fun.