How could we determine whether CCC2 ST has input lag?

talk about how great training mode is
Maj
Posts: 6753
Joined: Wed Dec 13, 2006 12:53 am
Location: Los Angeles
Contact:

How could we determine whether CCC2 ST has input lag?

Post by Maj »

Everyone knows that CCC2 ST is imperfectly emulated because the speeds are off. That's been tested and verified. Another issue that keeps coming up is the question of input lag. To be clear, i have no reason or desire to get involved in that argument other than wanting to defend NKI from the helluva random suckas disrespectin' the people's champion.

However, i am mildly curious about whether the input lag claims are indeed true. NKI said that it would be possible to test with a programmable pad, but i don't think it's that easy. Maybe having a ppad would help, but you'd need something more.

Simplest way to test it would be to get a splitter that lets you hook one controller up to a PS2 as well as a supergun, then use a fancy camcorder to record both TV screens. I don't have any of that stuff though, so can anyone come up with a more elegant experiment?
ShinjiGohan
Posts: 374
Joined: Sat Feb 03, 2007 7:42 pm
Location: chicago
Contact:

Post by ShinjiGohan »

in all honesty, thats really the only way that I can see it being tested with any degree of accuracy...

Well there is one other thing, NKI said something about the sound lagging as well. ie you press an attack and a couple frames go by then the sound plays.

So what you could do is set up a situation in both the supergun and CCC2 and record the exact same input. IE In ryus stage at time 85 Ryu does a standing fierce.

Then look at an advance video editor that shows the audio levels and see where it peaks from the sound of the attack.... oh wait that doesn't take into the time the button is pressed...
laugh
Posts: 118
Joined: Mon Jan 22, 2007 2:25 am

Post by laugh »

I've actually been thinking about how to PROVE the input lag on CCC2 (cuz it's so obvious it's there) since a couple months ago, and I've already posted how in the CCC2 info thread started by Sirlin. I'll copy it over here.
Somebody with an ST board+supergun and PS2 CCC2 and two of the same TVs could do a scientific input lag comparison with 1 programmable pad that sends out inputs at the same time to both setups through a converter (converter doesn't affect the lag, because both systems are fed by a single converter) and record the two TV screens with a camera that can do 60fps. The speed settings should be selected to best match each other, and the moves to use for accurate testing are the 0-frame-startup moves like normal throws and ken's fierce DP. Since they have 0 frame startup, even if there's a giant speed difference, the throw should show its first animation frame, or in the case of Ken's fierce DP, it should hit the opponent on the very next frame on a lagless system. By doing 0-frame moves, we can eliminate the speed difference factor as well because it can't skip any frames before the move is in effect.
I have everything needed to do that minus another TV and an ST board (I have one but is on a loan at an arcade right now)

Basically you just do a move that has 0-frame start up so that no matter what the speed difference is, it happens on the very next frame. The test would be even better if there's an LED that lights up when the button is pressed so that when you slow down the captured footage later you can tell on exactly what frame the button was pressed by seeing the LED turn on for a solid reference point and when the action happens on the screen. The test actually doesn't even need a programmable controller or a converter, you can just do a fierce throw with an LED that goes off the moment that button is pressed and just record that and the TV screens at 60fps.

I think there's more hope in finding the right dipswitch combination on the DC ST to match the arcade version (hopefully the round start delay can also be fixed through some magical dipswitch combination) than trying to adapt to input lag on CCC2 ST.
CC that shit
NKI
Posts: 215
Joined: Wed Dec 13, 2006 11:18 pm
Contact:

Post by NKI »

I had been thinking about this myself, and I had come up with the same idea of splitting the input to CCC2 and to a SuperNOVA. I can easily do that, but the problem I'd run into is that I don't have two of the same TV's. And I'm sure after I went through all the trouble of setting up and recording, some wise-ass would say, "TV's aren't the same. Your test is bogus."

I also thought about laugh's idea of recording the input at 60 FPS using an LED to show on what frame the circuit was completed. This is actually seeming like a more realistic test to me.

So I guess I'll get started on that.
Xenozip.
Posts: 1574
Joined: Mon Dec 18, 2006 2:24 pm
Location: N.EC
Contact:

Post by Xenozip. »

If you have two different TV's, couldn't you perform the test as-is. Then perform the same test, but with the TV's flipped around?
Looks like Jolly Ranchers & Baskin's Sherbet.
NKI
Posts: 215
Joined: Wed Dec 13, 2006 11:18 pm
Contact:

Post by NKI »

Yeah, could do that too.

Maybe I'll put both tests in the vid, just to remove any doubt.
laugh
Posts: 118
Joined: Mon Jan 22, 2007 2:25 am

Post by laugh »

I saw your post on srk before I saw this and thought we both coincidentally had the same idea. lol

The easiest way to rig this up IMO would be to use a couple of CD4066 switches. These switches could provide the most accurate simultaneous input on both CCC2 ST, Supergun, and the LED with the push of a single button without having to mod things much.
CC that shit
NKI
Posts: 215
Joined: Wed Dec 13, 2006 11:18 pm
Contact:

Post by NKI »

I actually wired up the whole thing already. It wasn't all that hard. Took less than an hour.

I just took a standard pushbutton (like you'd find in a regular American joystick), wired up a Y-splitter, one end feeding to a PS2 controller, the other end feeding to one of my SuperNOVA joysticks. Then I piggy-backed a small LED on top of that, and it works great.

Now it's just a matter of recording and editing the video.
desk
Posts: 416
Joined: Sun Dec 17, 2006 12:52 am
Location: UK
Contact:

Post by desk »

is it wrong that i'm really looking forward to the vid? lol. I don't even really care about the result as I've only played CCC2 once (I bought it just for ST). I'm just hella interested to see how people react.
laugh
Posts: 118
Joined: Mon Jan 22, 2007 2:25 am

Post by laugh »

desk wrote:is it wrong that i'm really looking forward to the vid? lol. I don't even really care about the result as I've only played CCC2 once (I bought it just for ST). I'm just hella interested to see how people react.
I'm hella anxious to see the video too for the same reason.
CC that shit
omni
Posts: 9
Joined: Wed May 02, 2007 11:01 pm

Post by omni »

NKI - stop being a jew and post what you found.
NKI
Posts: 215
Joined: Wed Dec 13, 2006 11:18 pm
Contact:

Post by NKI »

Got all the video captured. Editing is next, but it probably won't be done before next week.

By the way, if you just want to know the result now:
As expected, CCC2 has input lag. The lag varies in length, but it's always between 2 and 4 frames, most often being 4, sometimes being 3, almost never as low as 2.
Xenozip.
Posts: 1574
Joined: Mon Dec 18, 2006 2:24 pm
Location: N.EC
Contact:

Post by Xenozip. »

Very interesting. That's even more delay than a normal windows-based emulator. Four-frame input delay is totally unacceptable for tournament play, IMO.

Too bad the test was already completed, I had some additional suggestions for it.

Out of random curiosity, did you complete the test on various stages as well?

Also, random side note: http://forums.shoryuken.com/showthread. ... ost4503547

I don't know if it's possible to access the dipswitches or disable the audio in CCC2, though (alas, I apologize for not finding out myself).

I hope in the video you freeze-frame and then advance a few frames to fully emphasize the delay, visually. I was going to suggest having Dhalsim or some other character jumping during the action so that frame advancing would be visible (the character would move position during the frame step).
Looks like Jolly Ranchers & Baskin's Sherbet.
desk
Posts: 416
Joined: Sun Dec 17, 2006 12:52 am
Location: UK
Contact:

Post by desk »

So it has random lag? like, in a single fight the lag varies? or is it stage/character dependent and remains constant for the whole round?

well done for going to the trouble to sort it all out, look forward to the vid.
laugh
Posts: 118
Joined: Mon Jan 22, 2007 2:25 am

Post by laugh »

Awesome news! I wanna see this video as much as I wanna watch MagMan vs Justin MM. Thanks for doing the test, NKI.

Four frames of input lag is just unacceptable for any kind of versus play. I'm also very surprised that a player like Sirlin wouldn't notice such an obvious lag in input while many others felt and noticed it right away. No wonder it was next to impossible to reversal anything.

Also, thanks to Xenozip, I got interested in finding the absolutely correct frame data for ST using FBA's frame step (which I just found out today. Been searching for this function in FBA for YEARS. Thanks again Xeno). For starters, ST Ken's DPs have 1 frame of start up, all of them including the fierce DP. The common knowledge is that Ken's fierce DP has 0-frame startup, but after testing, I found out that Ken's DPs all have 1 frame startup, it just seems like it's 0-frame most of the time because of frame skip. I'm using SSF2 where ever I can to get frameskip-free figures.
CC that shit
Xenozip.
Posts: 1574
Joined: Mon Dec 18, 2006 2:24 pm
Location: N.EC
Contact:

Post by Xenozip. »

Yeah, I can confirm that. If you input a throw, the opponent will enter hit animation on the next frame (instantly). So, I'm assuming throws are instant and Ken's DP has 1F startup.
Looks like Jolly Ranchers & Baskin's Sherbet.
NKI
Posts: 215
Joined: Wed Dec 13, 2006 11:18 pm
Contact:

Post by NKI »

Xenozip. wrote:Out of random curiosity, did you complete the test on various stages as well?
Yes, and the results were the same: a random amount of lag (between 2 and 4 frames) on everything.
I don't know if it's possible to access the dipswitches or disable the audio in CCC2, though (alas, I apologize for not finding out myself).
CCC2 does not have an option to turn off the music.
I hope in the video you freeze-frame and then advance a few frames to fully emphasize the delay, visually.
Of course. :smile:
desk wrote:So it has random lag? like, in a single fight the lag varies? or is it stage/character dependent and remains constant for the whole round?
It's random on every single move. Could be 4 frames on your first move, then 2 frames on your next move, then 4 frames again on the next move, then 3 frames on the next move, etc...
desk
Posts: 416
Joined: Sun Dec 17, 2006 12:52 am
Location: UK
Contact:

Post by desk »

It's random on every single move. Could be 4 frames on your first move, then 2 frames on your next move, then 4 frames again on the next move, then 3 frames on the next move, etc...
:shock: That is crazy!
Xenozip.
Posts: 1574
Joined: Mon Dec 18, 2006 2:24 pm
Location: N.EC
Contact:

Post by Xenozip. »

Wow. The only thing worse than lag is inconsistent lag. D:
Looks like Jolly Ranchers & Baskin's Sherbet.
Mike Z
Posts: 198
Joined: Wed Dec 20, 2006 2:48 am

Post by Mike Z »

Doesn't that make combos pretty much impossible?

Mike Z
Maj
Posts: 6753
Joined: Wed Dec 13, 2006 12:53 am
Location: Los Angeles
Contact:

Post by Maj »

It makes tight links nearly impossible, but comboing off a deep jumpin or canceling into special moves is still easy. Though on the flipside, timing a deep jumping becomes a hell of a lot harder.

Cancels are easy though, cuz you get like a 13-frame window to interrupt a normal move with a special move. Canceling into supers is another story though.
ShinjiGohan
Posts: 374
Joined: Sat Feb 03, 2007 7:42 pm
Location: chicago
Contact:

Post by ShinjiGohan »

As long as the input lag is simliar to that of an HD TV, it shouldn't be too bad for combos because you're muscles know when to press the buttons whether or not the screen is lagging behind or not. I believe that was the case at ECC last year for evo quals. Where James Chen did a quick test on some LCD TVs that they got and he did a few combos to test for lag and he didn't find any. But when people started playing against eachother attempting reversals and such it became noticable.
Maj
Posts: 6753
Joined: Wed Dec 13, 2006 12:53 am
Location: Los Angeles
Contact:

Post by Maj »

Depends on too many things. If you firmly believe there's no input lag and you never allow yourself to doubt it, then maybe you can preserve your muscle memory and continue playing well. However, if someone tells you to watch out for lag (because it exists), then hand-eye-coordination will kick in and your muscle memory will slowly erode. It doesn't take much for any ST player to notice that something is off. Ignorance won't last and it won't save anyone forever. NKI had high hopes for CCC2 but as soon as he noticed his Bison crossups whiffing, he started to get suspicious.

To be honest Shinji, i'm totally confused by your reaction to this whole ordeal. Why are you putting Sirlin on a pedestal and blaming people like zass for questioning him? If anything, NKI and zass should be considered untouchable because they are volunteering obscene amounts of time to get to the bottom of these issues. Don't you remember how NKI put together those ST tutorial videos for CCC2? Sure, Sirlin was involved too, but it's his fucking job. Everyone else who contributes their time to these projects does it for free.

The real question is: To what extent should we absolve Capcom developers of errors and shortcomings? Yes, there have been 7 imperfect console versions of ST. Does that mean we as Street Fighter players (and buyers!) don't deserve an arcade perfect port of ST? I don't see how it's a factor at all.

Yeah, it's cool that "our people" now work at/for Capcom. Does that mean we should let them get away with whatever they want? The whole benefit of having people on the inside who understand our concerns, is to see those concerns addressed. If all we get is another flawed version of ST and a whole bunch of empty hype, then what's the point?

We need to figure out what we have a right to demand and what we're willing to concede. I do support Sirlin's efforts to get things done. It's ok if he needs more than one try. As long as he understands what we want and why we're not happy with the halfway point, then i'm very hopeful that STHD will turn out great.

Still, is it ok for Sirlin to act like he did nothing wrong? Is it ok for him to shun all criticism on the grounds that "we don't know cuz we weren't there"? Personally, what offends me is that after failing to deliver on his promises, he still has the audacity to make an "ok kids, shut the fuck up now" thread. How the hell did he expect to solve anything by talking down to the entirety of his following?

Also, let's be clear on this: CCC2 is not arcade perfect and no amount of semantical maneuvering is going to change that. In order for it to be arcade perfect, both the software and the hardware have to be accurate. In this case, the emulator represents the arcade hardware and CCC2's emulator sucks. If you're willing to call that arcade perfect, then you're letting developers get away with a whole boatload of half-assed bullshit.

Don't be so masochistic dude. We deserve better. Maybe there's more to the story than we know, but if that's the case then Sirlin sure has a funny way of asking for the benefit of the doubt.
Xenozip.
Posts: 1574
Joined: Mon Dec 18, 2006 2:24 pm
Location: N.EC
Contact:

Post by Xenozip. »

ShinjiGohan wrote:As long as the input lag is simliar to that of an HD TV, it shouldn't be too bad for combos because you're muscles know when to press the buttons whether or not the screen is lagging behind or not.
Yes and no.

I play a lot of online games with input delay, MB:AC and IaMP in particular which have netplay clients that allow players to set an input delay buffer (usually anywhere from 2-5 frames for most matches) in order to compensate for latency.

I can tell you that from my experience 2F input delay is easy to adjust to, but 3F and higher things get a lot more complicated. Stuff that requires tight-frame links, just-frame (reversal timing) inputs, even stuff like IAD's and and other things become very flaky and do not feel right or are not executed right in general.

And going from 0F to 2F delay is not simple at all. It screws up a lot of timing like in the aforementioned situations, such as reversals, links, IAD's or jump-ins in general, tick throws, and even things like tigerknee's, etc.

Also adjusted from 2F to 4-5F and back to 0F is ridiculous, which is why most players look down on practicing online for offline-tournaments. Unless you just happen to be particularly good at adjusting, it's generally not a good idea.

What NKI is saying, however, is that ST on CCC2 randomly(?) jumps around from 4F to 2F. That's absurd. That's downright unplayable IMO.

If my online matches skipped around between input delay like that it would be total garbage and I would not play it, mainly because consistent delay/latency is much better than random or inconsistent latency.

Who cares if the sprites, hitboxes, etc are accurate if the inputs are being read at inconsistent speeds? That's absolutely unacceptable for tournament use.

"I push a button, my attack doesn't come out when it should", therefor it is not tournament acceptable or arcade perfect. The internal data might be perfect, but the actual gameplay is not perfect.
Looks like Jolly Ranchers & Baskin's Sherbet.
Magnetro
Posts: 1126
Joined: Fri Dec 15, 2006 5:51 am
Location: CA
Contact:

Post by Magnetro »

NKI: Why do you keep answering power333's posts????????????????

did you get the email?
ShinjiGohan
Posts: 374
Joined: Sat Feb 03, 2007 7:42 pm
Location: chicago
Contact:

Post by ShinjiGohan »

I admit that there are problems with CCC2. However openly insulting one of the people that worked to get yet another port of ST made is not the right path. Whether we like it or not, we are at capcoms disposal of games. Its their choice whether to make another 10 ports of ST, to make SF4, 5 and whatever. But if we piss off capcom and what little supporters we have within the company. Then it'll be another 10 years before another SF game gets released. And I don't want that.

Its still fine to give our opinions but we have to do it in constructive ways. If we don't then we may scare off the only hope we have for an arcade perfect ST, or a brand new SF game. And that won't be to the satisfaction of any SF player.

For backing Sirlin rather than NKI or Beasley. It really isn't about that. I'm willing to give Sirlin the benefit of the doubt. As much as all of us wanted CCC2 to be perfect, he wanted it that much more. Why? Because he knows that the entire ST community was waiting for this. He didn't want to mess up because he was our ray of hope. So I'm willing to believe that he tried everything that he could to get this port right. And how seemingly most of the people bashed him after Evo worlds, that wasn't right. He didn't have to convince capcom to include ST in CCC2. There could have been gigawing on that disk instead (great game BTW lol). Sirlin doesn't owe us anything and we shouldn't be so hastey with our negativities. He's heard enough bad things about his work and I'm sure telling him the good things will help him far more than insults will.
Maj
Posts: 6753
Joined: Wed Dec 13, 2006 12:53 am
Location: Los Angeles
Contact:

Post by Maj »

I agree with some of what you said but i strongly disagree with your overall viewpoint. Do we really need more crappy versions of Street Fighter? Isn't that what diluted the value of the brand in the first place?

There's no need to act meek and vulnerable. This whole thing is about our money and our support. If we're so eager to give it away freely every time Capcom promises something we want or gives out SF4 stickers, they'll have no incentive to deliver on those promises.

CCC2 was supposed to have an arcade-perfect version of ST. It's important to us because every year we run into the same dilemma of which version to use for Evo. Whether or not it was Sirlin's fault, CCC2 failed to deliver. It was marketed as the answer to our demands and that turned out to be a lie. Why should we hold back our criticism? Why should we put up with idiots who keep telling us to get over it? Pretending the problem doesn't exist won't solve shit.

Now if Sirlin is gonna come around acting like everyone should stop blaming him, then he needs to tell us who to blame. Because somebody is at fault. It makes no sense for Sirlin to demand respect for claiming he created something that he ultimately failed to deliver. And he sure as hell can't do this in a condescending tone.

If the dude wants to shoot himself in the foot, that's his call. I can't understand why anyone would want to jump in the way of the bullet.
ShinjiGohan
Posts: 374
Joined: Sat Feb 03, 2007 7:42 pm
Location: chicago
Contact:

Post by ShinjiGohan »

The problem with making ourselfs vocal in a negative way is the very fact that Sirlin knows exactly what we want. We've all known that since the moment when ST tournaments went console.

Do we wanna keep on buying crappy ports? No. Sirlin knows we don't want that and I'm sure that as long as he gets a say he'll try and make it as close as possible. However thats why we're able to rent games. So see if we like them before we buy them.

I feel as long as we express our critism in a acontructive way it'll only make things better.

Plus we don't have to spend our money for a port, or as much money. Go over to a friends house, DL the game on bittorrent, or rent the game. Maybe if you're lucky they'll have it set up as a playable game at best buy or gamestop and you can try it out there.

We've been trying to yell and scream at capcom for years every time that they mess up. Why not try something else instead to get our message across.
Xenozip.
Posts: 1574
Joined: Mon Dec 18, 2006 2:24 pm
Location: N.EC
Contact:

Post by Xenozip. »

Hooo, boy.

"The arcade version isn't console perfect anyway, so it's not that good of a version anymore." -Sirlin
Looks like Jolly Ranchers & Baskin's Sherbet.
desk
Posts: 416
Joined: Sun Dec 17, 2006 12:52 am
Location: UK
Contact:

Post by desk »

shinji, the point you make about renting games in somewhat naive. I bought CCC2 on day1 because it was going to be arcade perfect and to support SF. There was no reason to rent it out first. The quality of the port wasn't up in the air, it was definite. And I don't even care, it might seem like I'm bitter about it (everyone in the UK plays hyper anyway). The thing that is actually pissing me off is everyone defending it and making out other people are just being petty.
We've been trying to yell and scream at capcom for years every time that they mess up. Why not try something else instead to get our message across.
Most people aren't deliberately being dicks about the whole affair (which seems to be what you're implying) There are very valid points being made that shouldn't be just skipped over and ignored.

I also have to point out that the fucking CCC2 thread over at SRK has me pulling my hair out. Every post makes me more annoyed than the last. I don't think there's any point in anyone making a sensible argument over there. Even when NKI releases that vid, nothing is going to get sorted and no one will admit anything. We'll just 'know' that we shouldn't hold tournaments on that version. how's that vid coming along by the way? :)
Post Reply