New discovery: Xbox 360 version of sf4 series has different timing from arcade/ps3/pc


#1

I will make a short announcement here and will post videos with in-depth analysis and additional proof of concept videos in the near future.

After spending a long time looking into and researching into the different input timings of all versions of Street Fighter 4 (ARCADE, PS3, XBOX 360, PC) and testing over the last 24 hours frame by frame video footage I can announce with almost 100% certainty that the XBOX 360 versions of Street Fighter 4 are 2 frames faster than all the other versions. This is most likely due to either non-existent V-Sync on xbox version (my personal guess), non-standard V-Sync implementation, or some other technical reason. I have frame by frame video footage showing ps3 version, pc version, and arcade Taito Type X 2 all running with the EXACT same input lag.

XBOX 360 input timing (2 frames faster than other versions) only matches one other version…PC with V-Sync turned off. V-Sync off is NOT the Arcade standard or the proper way to play. V-Sync greatly affects the visual input timing and with it off or not working properly (360) we get random timings for doing our visual links. So basically the Xbox runs a little faster than the other versions but it pays for it with inconsistent timing (especially 1 frame links) and throwing people coming from Arcade, PS3, and PC off.

I have made a somewhat long analytical YouTube video demonstrating my findings that can be found here https://youtube.com/watch?v=QQNYkl4WVWs.
I will make a shorter more condensed version with a catchy song and slit screen simultaneous comparison at some point in the future.

If you want proof of the botched Xbox v-sync look no further than this video of sf x tek which runs on the same engine as Sf4. https://www.youtube.com/watch?v=0kZv_eme1TM

Notice how the tearing % counter under the frame rate is jumping all over the place on the 360 yet remains locked at 0% on ps3 (proper v-sync). Keep in mind that street fighter x tekken runs on a modified engine of sf4. Also notice that the ps3 frame rates matches or exceeds the 360 (basically locked at 60fps) with 4 characters on the screen at the same time, this also makes it appear unlikely that the ps3 drops the frame rate with 2 characters on screen on sf4. The only exception could be overheating issues after extended play in hot enviroments which also affect equally (if not more) the 360.

I decided to research and experiment this all out when I noticed occasional screen tearing on the xbox version and during EVO 2014 players Daigo and Sako drop numerous common combos with Evil Ryu at a staggering consistency. These are combos that I can do with 90% efficiency that make no sense for players at that level to be fumbling. This fell into a suspicion I’ve had for awhile that it’s not the ps3 version that’s off but actually the XBOX. It makes sense as I remember older footage from EVO 2009 sf4 played on ps3 where Daigo CONSISTENTLY hit 1 frame links (and he came from playing exclusively on Arcade) while I now see him, Sako, and other Japanese players dropping easy 2 frame links left and right. It seems that Japan is also unaware of this difference as they also believe the Xbox version to be superior because it’s faster and probably don’t realize that it’s actually running too fast.

The consequences of this information must bring into question MANY tournament results over the last few years (including this years EVO) when players that have been playing almost exclusively on Arcade cabinets (mostly Japanese pros) come to USA or other tournaments and perform less well than expected. On the flip side this also explains why many of the best American (and EU) players get slaughtered when they go to Japanese tournaments that are frequently played on Arcade.

Finally let me bring to attention that this has nothing to do with online play (netcode) which is overall better on Xbox. I do have suspicions though that perhaps the xbox version only feels better online due to the reduction in input lag helping in this scenario and the greater number of hard wired players. In this regard the xbox version remains superior.

FOR OFFLINE AND TOURNAMENT PLAY PS3 OR PC SHOULD BE THE STANDARD IN ORDER TO MATCH THE INPUT TIMING OF THE ORIGINAL ARCADE VERSION WITH PROPER V SYNC IMPLEMENTATION.

I will post more videos with additional proof of concepts in the near future.


#2

yeah but ps3 drops frames.

I’d rather the game be 2f faster than randomly kill my inputs, start running at 30fps, or chug during my mixup making it easier to block.


My execution seems worse when I play at the arcade?
Help Pintpointing source of input lag on pc?
#3

I’ve never seen anything showing ps3 dropping inputs (have logged hundreds of hours) and the frame issue could be alleviated with the standard selection of the training stage (as much as I hate it). The 2f faster on Xbox is a 100% constant that makes it’s timing different from all other versions.

Ideally we would play on pc’s but that would be very cumbersome to set up properly in tournaments. I guess the best solution would be a ps4 version (EDIT: Nevermind, the Ps4 also runs on an AMD GPU and would probably still have the same issues).


#4

I guess Xbox 360 should replace any arcades still used in tournaments then. Btw, I thought it was common knowledge that 360 ran faster than arcades?


#5

@seph13‌ i own all 3 versions myself. I definitely feel the 2 frame difference when trying to land one frame links when i switch between versions. I’ve always felt the xbox360 version was the most responsive version and almost everyone i know prefers playing on it. I disagree with your statement about PS3 being the definitive version for tournament play based on the fact that it has the same timings as the Taito Type X 2. Think about it this way… in the grand scheme of things there are far far more console players than arcade players out there. And in this day and age console tournaments are now the standard… I’m with @Doopliss‌ - we should be using the superior console version. Nonetheless, looking forward to seeing the videos…


#6

http://legacy-cdn.smosh.com/smosh-pit/122010/kanye-serious.gif

Yeah good luck with that.


#7

While I can understand peoples thinking that we should just consider the xbox version the standard since it is the most responsive and is what we’ve all gotten use to we should take into consideration that it is the ONLY version with this faster timing and that this is most likely due to a technical limitation rather than an enhancement. It appears to me to be more of a bug because the xbox uses an AMD gpu and I’ve seen AMD GPU’s on PC version of Street Fighter 4 series have horrendous slow down/input delay with V-Sync enabled. The Taito Type X-2 and Ps3 run on Nvidia cards and that is what Street Fighter 4 was developed on. For whatever reason Street Fighter 4s V-Sync only seems to work 100% normal on Nvidia cards (though perhaps later AMD drivers fixed this). It appears that during development it was decided to just disable v-sync on X-box to avoid these issues. The other versions that run on Nvidia cards that the game was developed on can run with proper V-Sync (never any screen tearing). V-Sync always adds a minimum of 1-2 frames visual lag which would explain the difference.

I would hesitate to consider a version with disabled v-sync due to not using the same gpu as the game was developed on to be the standard.


#8

I edited out the “CAPS LOCK IS CRUISE CONTROL FOR COOL LOL11!!” to keep the SF4 forum pals’ eyes from bleeding. Please be more considerate when posting in the future. The all caps thing was fine when it was 1995 and you were posting stuff up on your geocities page, but come on man… it’s 2014.


#9

and when ps3 was the standard everyone hated it cause it plays like garbage.


#10

Apologies for the CAPS. I never use them but in my excitement I might have got carried away and wanted to make sure this subject was brought to light.


#11

The japanese players also play on xbox you know…


#12

Why would a ps4 version be the best solution if an AMD gpu is the problem?


#13

Yes I realize this. I suspect that it is because when playing online (1-2 frames added lag on the best connections) it is almost the same timing as the Arcade version. It could also be that they are just unaware of the difference.


#14

I’m so sick of people blaming the platforms for players messing up. The 2 frame input difference is so incredibly negligible, it is just a non issue, at all.
I can make all the asumptions i want why Daigo and Sako dropped their combos, but it is so easy to blame the hardware… I’ll stick with Sako dropping his stuff in the beginning because he was still cold and he didn’t exactly go for the most easy stuff, later on he was fine. Daigo, offday, not completely comfortable with E.Ryu, not known for excellent excecution etc bla bla, who knows.


#15

Great point! I totally forgot that the Ps4 runs on an AMD GPU. You’re right, it would probably still have the same issues and need V-sync turned off unless the newer AMD GPU’s/Drivers don’t have this problem anymore. Something I unfortunately can not test due to not owning an AMD graphics card. I will edit my previous comment.


#16

Or maybe ya know, Japan didn’t get a console release yet, and they were dumb enough to not import the game.


#17

So the PS3 version is the best after all


#18

Why? It’s better than the intended standard, so there’s no reason not to use it.


#19

lag test results should come with the first post. not much point in an announcement with no proof.

did you actually test to see if PS3, Arcade, and PC all lag the exact same? did you test if vsync turned off on PC is same as 360? did you account for monitor lag across the various setups?

you should post test methodology and results for everyone to pick over.

do you actually have access to an arcade setup and a high speed camera for testing this?


#20

I don’t remember ever seeing screen tearing on the Xbox version.