Frame (input lag) differences: PC to X360


#1

I played on Xbox360 for years, and have played on PC exclusively for the last year or so. I use Evo monitor for both versions. ASUS VH236H

PC is hooked up via DVI. Fixed framerate, no vsync, my computers runs it at some obscenely high FPS.

Xbox uses same monitor hooked up via VGA. Runs at native monitor resolution (1080p) inside Xbox settings. Seemed to get even worse if I change it to 720p inside xbox settings.

All offline, all training mode, same joystick

There is a distinct difference in link timing. So much so that it is hard to combo jabs, much less 1 frame links. I main Abel and his main F-Mk -> St.hp xx COD is a 1 Frame link. The timing is probably 2-3 frames different AT THE VERY LEAST, causing me to hit the buttons WAY too early on the Xbox. Even with looser links, I still have to adjust my timing to hit the buttons noticeably later on xbox.

I can’t tell which game is improperly configured but the PC sure seems to feel more responsive. You would think that with the same monitor, playing in training mode with no lag, and a beefy PC I wouldn’t have these issues.

Thoughts?


#2

The PC version is based on PS3 timings I believe. It’s been known since the game came out that SF4 runs at different speeds depending on the system. This is why people practice purely on PS3 and not 360 when training for EVO and similar.

I’ve always found it easier to hit links on 360 myself, but it’s what I started on so I got used to it.


#3

Having both consoles and a PC setup as well I have always felt the 360 version is much more responsive (also using a evo monitor with hdmi connections). I swear I’ve read somewhere that it has to do with how the ps3 handles the wired USB controllers.


#4

Can we agree that by our subjective description of “more responsive” we are saying that moves come out faster after button press? Therefore it should follow that if the visuals shown on screen are synced up, the machine you are able to hit the button on LATER and still combo is in fact the more responsive? (time from press to move execution is faster)

So if I am hitting buttons too early on Xbox, chances are my timing for PC is based on some kind of input lag causing me to slow things down.

I will be testing tonight and recording the results.

Additionally, my dual modded stick has a hack xbox360 pad in addion to the Chimp board (which can work in xbox or ps3 mode when attached to a computer. I will be testing both modes.


#5

With Super Turbo, average input lag is 2.5 frames on CPS2, 3.5 frames on Xbox 360, 4.1 frames on Dreamcast and 4.3 frames on PS3.

The extra frame of lag on the PS3 has - I think - more to do with how the PS3 handles vsync than with its bus.

FWIW: In Vanilla SF4, on the Xbox 360, using VGA out, there are subtle game speed differences caused by players holding down buttons, and VGA out runs ever so slightly faster than NTSC.


#6

I wanna clear things out a bit, since I believe every “lagless” discussion should have these points taken into account:
[list]
[] There is no real “lagless” hardware and/or software configuration and every config has a delay between input and response.
[
] There’s different delay response between configurations. Any pair of different configs with the same delay is just coincidence. This explains how difficult (almost imposible) it is to have 100% arcade perfect ports. Differences may be difficult to spot out, but there is no such a thing as an arcade perfect port. The variables are just too mucho to control (process architecture, software, build platform, operating system…)
[] Whether one configuration has more/less delay doesn’t make it better/worse. It’s just different
[
] All tournaments are anounced with their configuration, so it can be reproduced by players. There’s no sustainable argument for “this config lags”. You can just say “it’s different from what I’m used to” or “it’s unreliable” (more on unreliable on next items)
[] A system configuration may be unreliable. Examples include mvc3/ps3 or kof13/ps3 (under specific scenarios) or any game on PC. The later is because Windows (and most operating systems) are not engineered to have only one process running and because they run under different hardware, so comparing same software and Windows version will probably have spottable differences. We all know about games that don’t work well with certain graphic cards.
[
] The same goes with monitor. Every pair of different models have different delay. Even if they’re very difficult to spot.
[*] Since there’s no “lagless” config, we have to agree on what is “lagless to the human eye”. See more about this on displaylag.com[/list]

How this answers your post?

They are different platforms and, as such, are supposed to have different delay response; you should get hold of conventional config (x360 + evo monitor) so you won’t be uncomfortable in different platforms.

Stories of people having to readjust to platform change have been too many


#7

Scientifically there no such thing as “lagless”. But in the realm of human perception there are plenty that are beyond our notice.
So were looking for a board with less than 1 ms. Human reflex response is anywhere from 18 to 300 ms.

Too subjective. I would disagree with you but before we can argue the issue we first need a set of judging criteria. That said the minor differences of the lack on insert coins and console only features I don’t think should be a factor.

Agreed

Same as the above


#8

You seem to be mixing what can be “perceivable by humans” and facts.

Computer/electronic geeks such as myself know for a fact that two different circuit boards/operating systems/platforms are supposed to behave different, whether we can perceive it or not.

Under what can be perceived by humans: yes, two different configs can behave the same. I was told this when I received my chinese NES clone, even though I could tell the difference.

If you want two configurations to behave the same, you can find a criteria that satisfies you, but the fact is that they are going to behave differently.
This mentality has to be kept in mind when developing industry products, and even so, sometimes we (yes, myself included) miss a factor and end up delivering “faulty products” or “products with unexpected behaviour”.

And, yes, we have to trust the reproducibility of products. If not, see the “reliability” section I wrote.


#9

Perception of lag is around the 100 ms mark. That’s where humans notice it most, at least in a visual sense.

CPS2 is 4 frames, DC is 5, Xbox is about 5 or 6, and PS3 ranges from 7 to 8. So, the worst case scenario of PS3 vs arcade CPS2 is a 4 frame difference.

Maybe but I think it’s a little more than that. Probably a combination of the frame buffer, vsync, and a few other odds and ends. IMO, I would mostly say it boils down to completely shit programming in the case of HDR on PS3, that game is chock full of crap code and even the menus have a bunch of slow down. It’s ridiculous. It’s the sole reason I will never buy digital PS3 games from Capcom ever again. No fixing it, and no getting your money back.


#10

Have there anything I posted that isn’t fact?


#11

“diff configs => diff response” is not being too subjective.

Whether we’re OK with different response on different configs is a whole different story.

What I’m trying to say is “using different ports is useful for leveling up, but you shouldn’t trust them for precise timing” backing it up with facts.