if I understand correctly, he’s not emulating anything. it sounds like arcade SF4 is just a PC program and you can run it on any compatible hardware.
here’s my beefs, though:
there’s no guarantee your hardware runs the game with the same amount of lag as the real hardware. they run SF4 on a modified OS right? so for all you know there’s a different implementation of vsync or different ways inputs are handled in the buffer etc. I’m not an expert in this area, but I don’t think you should assume a copy of Ultra SF4 running on your machine matches the conditions of the arcade setup.
there are some parts of an arcade setup you can’t replicate realistically - how laggy their monitor is vs how laggy the ASUS monitor is for instance. their monitor might lag an extra frame, who knows. in which case it wouldn’t be the same in the overall picture.
I think it’s a big leap of faith to assume it’s the same. especially when you’re making a claim that PS3 and vsync PC version are the same as arcade. you can’t know that unless you sit down next to a cabinet and test the real thing.
also this is just an aside (not relevant to the test) but I think you’ll find that players of modern games won’t care as much about this. if you tested ST or 3s for console and found X version was closest to arcade, people would be really happy. arcade is the gold standard for those games. for SF4 though, it’s been console-centric everywhere outside Japan. so whatever standard the community agrees on is the standard. so even if you prove PS3 is exactly the same as arcade, most players will likely prefer playing 360 since that’s what has become the standard, and people generally prefer less input lag in their setups if they can help it.
He’s using a program that emulates the I/O board (which handles the inputs) on the real arcade hardware. Without the I/O board, the game refuses to boot.
ah I see. in that case, yeah I think that’s even more pointing towards “you gotta try this test out on real arcade hardware.” will be a lot of skepticism without that.
The X box is the culprit not the ps3/arcade in difference of timing. It would explain why when I went to my first 2 tournies which ran on xbox on Asus monitors, and I was coming from ps3 on a crt, that having to use my ps3 controller +cronus on an xbox on a monitor threw everything off for me. Hitting Kens stand jab (+7) into fierce dp (3 frame) was a chore. …
But what did I do? I said fuck it and bought an xbox ($30), bought ae ($10) (and now ultra +$15). Bought live ($60), bought a 100ft cat 6 cable ($20) because wifi is trash, and now prepping as best as I can and adjusting so that the next time I go to a tournament I fare much better.
As much as I love my ps3, and the dual shock controller design, everyone plays with wifi, the overall player base is smaller, it’s filled with much more casuals, and the horrible delays and lag online create it’s own online meta game bull shit lol. I can throw out an ex lmt (flip kick) with poison offline/xbox and if blocked use my +2 frame advantage to the counter hit my opponent. On PSN, I’ll get thrown or jabbed right out of being +2 every time. Kens step kick> cr.mk xx hadoken on PSN is a true block string, adons and onis will have their jag kicks / oni slash blocked and throw you after it every time for pressing a 3 frame normal. Etc etc
In-depth Youtube video showing my findings here https://youtube.com/watch?v=QQNYkl4WVWs for those interested. It’s a little long and wordy but for those interested they can check it out. I will make a shorter version with simultaneous split screen in the near future.
I have never seen any actual live evidence of slowdown on a ps3 (offline). Many people have claimed this but I have never seen proof and can’t replicate any offline slow down in my own tests besides a very occasional amount during ultra cinematic transitions which is during non gameplay moments. I would love for someone to show me a concrete example of this though.
Yup Taito Type X-3 which is just a more porwerful version of Type X-2 ( Intel core I-5, Nvidia Geforce 560ti), the architecture is the same (pc) and the input timing should be the same as the Type X-2 version. The only difference should be graphics that are closer to the pc versions of sf4 at high/max settings.
@seph13 xbox 360 is a popular console in the U.S. and much more true for street fighter. Bear in mind that sometimes the console fanboys tend to have a bias
I am very glad you did this as I was wondering the very same thing. I own a ps3 and play on pc and played on xbox, I turned off v-sync on my PC and I noticed a difference. This always bugged me very much and its wonderful someone took an approach. If you google input lag for this matter all you get is statements from various forums saying phrases like “I think”, “common knowledge”, and other statements without any solid proof. There was a video a few years back for a ps3 vs 360 for the “input lag”. As it turns out from your data it isn’t input lag at all!
PS3 might not have extra input lag as people thought but it was just the graphical settings causing this.
If the PC can reproduce the SAME results as the PS3 with a given setting on V-Sync then either the PC miraculously has the same input lag as the PS3 or the PS3 doesn’t have the input lag people thought it had compared to the Xbox
Turning on the replay saving feature along with any non-training mode stage will cause lag. The big factor here is the replay saving, that causes huge lag both in the game and even more when the match is over. Decapre’s ultra 1 is also very laggy but other than that I don’t see my PS3 lag any more than when I play on PC.
Seph - you’re 100% correct, and all your findings are great and very useful. This should let everyone know that playing street fighter on ps3 or arcade isn’t optimal. You either play on 360, or you play on PC with v-sync disabled.
Tournament organizers have already shifted from ps3 to 360. It’s a done deal. The community has all bought xboxs to play on. All of the top players have Xbox Live subscriptions. There really isn’t any going back now, regardless of which platform it **should **be run on.
yes this is true for any combos with a 2-frame link and above. It is very easy to do them on either. However some players rely on visual cues to land their combos, and since one system gives visual cue sooner than the other you end up doing the combo earlier. Although if you rely on muscle memory and memory of the sound of the buttons on the stick more then you will land the combo no matter how much input lag or what monitor you use.
But this isn’t the issue, this happens online when there is lag, people still are able to do combos even though there is rough internet lag. The problem here is that the xbox 360 is inconsistent, without v-sync the game may or may not have a delay of 3 or 4 frames while the ps3 consistently has the same lag. When there is such an inconsistency of 1 frame, there isn’t an issue at all with 2-frame links or higher which MOST people do. Not many people actually go for 1-frame links. When Daigo or Sako comes along playing E. Ryu they do 1-frame links all the time. They have to do cr.LP, cl.st.HP which is a one frame link. Then they have to do the MK Hammerkick into cr.MP or Hammerkick into cr.LP into st.MP. These are all one frame links and when you do 4 or 5 of them chained together with FADC’s in ONE whole combo then messing up that one combo can cost you the round. This is why the inconsistency of speed on the xbox is SERIOUS!
V-sync keeps it consistent which is more important than a negligible one or two frames of delay as compared to inconsistent game frame calculation.
Also street fighter holds its origins in the arcades and it holds a high place in the FGC, it’s not like arcade machines are slow, they have good hardware but they have V-sync on.
Actually, if you’re targeting a specific, consistent frame rate, enabling v-sync can actually increase the chance of dropped frames. V-sync has nothing to do with consistency. V-sync means the game waits for the video card to be in a state where it’s safe to flip display buffers without causing screen tearing before flipping to the next frame.
Normally, this is a good thing, but if your system can’t run fast enough, it’ll drop a few frames here and there, much more than with v-sync disabled.
That said, I like to see someone else do an independent test using the same setup and possible with an actual arcade board, just to put any doubts to rest.
While technically what you say is true, your conclusions are not entirely. V-Sync does have to do with consistency of visual input timing to a degree. Without V-Sync your monitor and gpu are never in a consistent from day to day or minute to minute sync. This means that one day your 360 could have a 4 frame input to display delay (visual that is, the game is always running internally at the same speed) and the next day you could have a 3 frame input to display delay, and it could change any second or any minute. On the contrary your ps3,pc, and arcade versions have v-sync on (pc default) and will always have exactly 6 frames input to display delay. If you look up input delay tests comparing v-sync on to v-sync off you will always find more variation with v-sync off (without v-sync it’s slightly random exactly which frame the display is currently showing). Usually any variation with v-sync on is so rare or minor (sub-1frame) that it falls within the margin of error.
Ideally we would all be playing with v-synced versions of this game but people jumped to conclusions when we saw that the 360 had faster input than the ps3 without looking into why or comparing it to all the other versions.
you still have no data about the actual arcade version of the game. you can assume your computer running the arcade version will be the same, but until it’s actually tested we have no way of knowing. considering you are emulating one of the arcade components in software, there’s no reason you should be assuming it’s the same. and since your whole argument is based on what matches the arcade version, it’s kinda important to actually test the arcade version.
you’ve ignored the last few posts of people mentioning this, but you should know that one of the big reasons this topic is getting not much interest is because of that.
Why would any prefer two frames I’d lag over 0? Or is it just because they don’t have a 360, that’s usually the case.
Marvel and sf4 are both known for dropping frames on ps3, so people bypass it by picking certain stages, just play the version of the game without issues
