How much does your arcade stick lag? Arcade stick input lag testing & results thread

1232426282933

Comments

  • AriesnoAriesno Seimitsu's warrior Joined: Posts: 407
    Hrap4 is this one : ac04-300x300.jpg
    is the best Hori in Teyah tests.

    1up-game.net/blog/wp-content/uploads/2015/10/ac04-300x300.jpg

    According to this japanese test , HrapH (falcon buton and Xinput) > Hrap4/TE2> HrapV/VLX.

  • ShinMagusShinMagus Joined: Posts: 428 ✭✭✭✭✭ OG
    edited March 2016
    Hmmm. I was thinking about the HRAP4 Kai, which is similar in input lag to the HRAPV and roughly similar to the VLX Kuro, according to Teyah. You seem to be thinking about the HRAPv4, which is another (older) model (the picture you posted is of the HRAPv4). Dunno which the video was referring to.

    UPDATE: I guess there isn't even a HRAP4 other than the HRAP4 Kai. Or is there?
    Magus
    12000 B.C. - Dark Ages
    "The black wind begins to blow..."
    "Can you hear that? It's the sound of the Reaper."
  • TenshoTensho Joined: Posts: 2,873
    edited March 2016
    In that video the HrapVH is faster than the TE2 and HrapV4. So if Teyahs tests are correct the HrapVH is under 4.80ms.
  • DarksakulDarksakul Your lack of faith disturbs me Joined: Posts: 23,821
    ShinMagus wrote: »
    Hmmm. I was thinking about the HRAP4 Kai, which is similar in input lag to the HRAPV and roughly similar to the VLX Kuro, according to Teyah. You seem to be thinking about the HRAPv4, which is another (older) model (the picture you posted is of the HRAPv4). Dunno which the video was referring to.

    UPDATE: I guess there isn't even a HRAP4 other than the HRAP4 Kai. Or is there?

    There is a plain "Vanilla" Hrap 4 with Hayabusa joysticks and kuro buttons, no touch panel, button assignment switch or anything.
    “Strong people don't put others down... They lift them up.”
    - Darth Vader, Philanthropist
  • XanfellXanfell Joined: Posts: 11
    whoops. Forgot to post up the Mad Catz TE2+ test for you mad catz fans.
    https://youtube.com/watch?v=coHVlO9pH58

    and he has a brook converter test if anyone is interested.
    https://youtube.com/watch?v=Qw2oNESnMAI
  • KomatikKomatik Card demon Joined: Posts: 2,669
    Xanfell wrote: »
    whoops. Forgot to post up the Mad Catz TE2+ test for you mad catz fans.
    https://youtube.com/watch?v=coHVlO9pH58

    and he has a brook converter test if anyone is interested.
    https://youtube.com/watch?v=Qw2oNESnMAI

    These videos omg. :'D
    Steam/Fightcade:Coffeeling
    Xrd: ? | ST: Claw, Ryu, O.Sagat | 3S: Chun-Li, Akuma
  • NYCgeekNYCgeek Joined: Posts: 96
    I hope the person does a TE2+ vs HRAP H
  • NeviusNevius Joined: Posts: 8
    I wonder if anyone has a comparison on input lag of the pc and ps4 version of SFV. I wonder how it is with vsync enabled or disabled on pc. And how different it is going as a pc player to a tournament using ps4 and vice versa.
    Probly would best to measure it with a highspeed camera or something. Would be really cool if someone could find it out. Make sure to use the same monitor and input device and don't use a xinput mapper that lags.
  • DubonDubon Joined: Posts: 497
    So uh, whats going on with that update to Teyah's site that was supposed to come early last week? I'm kinda curious about it now..
  • gahrlinggahrling Better safe than Hori.. Joined: Posts: 3,968
    Teyah wrote: »
    Upon request I also tested an old PS3 Mad Catz MLG TE stick that a friend was convinced was laggy (so much so that nobody wanted to play on it). Turns out he was right - it had +21.75 ms of lag. He's working on swapping in a new PCB on that thing now.


    Is there a reason why you don't identify the pcb but only identify the arcade stick? Because AFAIK that MLG should be using the same pcb found in the licensed PS3 Fightsticks..
  • DarksakulDarksakul Your lack of faith disturbs me Joined: Posts: 23,821
    gahrling wrote: »
    Teyah wrote: »
    Upon request I also tested an old PS3 Mad Catz MLG TE stick that a friend was convinced was laggy (so much so that nobody wanted to play on it). Turns out he was right - it had +21.75 ms of lag. He's working on swapping in a new PCB on that thing now.


    Is there a reason why you don't identify the pcb but only identify the arcade stick? Because AFAIK that MLG should be using the same pcb found in the licensed PS3 Fightsticks..
    Especially since many of these sticks do have multiple PCB Revisions, the Mad Catz Street Fighter TE (Round 1) for the Xbox 360 has at least 2 PCB revisions.
    Revising a Printed circuit board for a major manufacturer is quite cheap. The Dual Shock 4 also had several hardware revisions already.

    Also if he going to quote Brook (or even the CronusMax) converters latency times, he also needs to mention what version of that converter's firmware he is using.

    “Strong people don't put others down... They lift them up.”
    - Darth Vader, Philanthropist
  • ShinMagusShinMagus Joined: Posts: 428 ✭✭✭✭✭ OG
    edited March 2016
    Hmmm. I thought the Brook Universal Fighting Board was going to be tested... :( The Xbox360-XboxOne to PS4 Brook Converter too, since people who used to play USF4 tended to have Xbox360 sticks rather than PS3 ones.

    I wonder if the Brook Universal Fighting Board has input lag similar to that of the Brook Fighting Board...
    Magus
    12000 B.C. - Dark Ages
    "The black wind begins to blow..."
    "Can you hear that? It's the sound of the Reaper."
  • gahrlinggahrling Better safe than Hori.. Joined: Posts: 3,968
    Darksakul wrote: »
    the Mad Catz Street Fighter TE (Round 1) for the Xbox 360 has at least 2 PCB revisions.


    3 revisions for the original PS3 TE, but who's counting : p
  • DarksakulDarksakul Your lack of faith disturbs me Joined: Posts: 23,821
    gahrling wrote: »
    Darksakul wrote: »
    the Mad Catz Street Fighter TE (Round 1) for the Xbox 360 has at least 2 PCB revisions.


    3 revisions for the original PS3 TE, but who's counting : p

    Also we never establish if 2 of the same identical boards, same revision could give out 2 different results.
    Example: What if variations in manufacturing produce the 2 of the same Mad Catz TE PS3 Round 1 UCHI protocol boards lest say Revision 1, version 1 could have two very different results.
    “Strong people don't put others down... They lift them up.”
    - Darth Vader, Philanthropist
  • DubonDubon Joined: Posts: 497
    Firmware version for Brook ps4 PCB should also be added
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    Jon tested the Brook converter vs a native stick on ps4 and on average had about half a frame or so of difference. Native would jump between 4 and 5 frames when done 20 or so times and the converter would average a consistent 5 frames across the board.

    This is why assigning a number down to the ms without testing via logic boards and oscilloscopes is horse shit. This is also why assinging a grade on a scale on something unscientific like this and passing it around is misleading. The brook converters are not shit tier and this asshole here is telling people it is.
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • TeyahTeyah Return of the Dragon Joined: Posts: 1,125
    Brook and CronusMax Plus are running the latest firmware, I'll add that information in on the next update.

    FYI I've retested firmware changes several times over the past 2 years on different converters and no differences in delay were noted. Not saying it isn't possible to significantly improve delay via firmware update, but I haven't seen it yet on the Cronus, CronusMax Plus, or PS360+.

    I do have PCB pictures up for some sticks on the site (HRAP4 line, Qanba Q4 PS3 line). No need to void wareanties to add pictures of PCBs for brand new sticks like the TE2+ or TE-S+ which only have one PCB.
    I see in frames IRL
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    K
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • TeyahTeyah Return of the Dragon Joined: Posts: 1,125
    edited March 2016
    Moonchilde wrote: »
    Jon tested the Brook converter vs a native stick on ps4 and on average had about half a frame or so of difference. Native would jump between 4 and 5 frames when done 20 or so times and the converter would average a consistent 5 frames across the board.

    This is why assigning a number down to the ms without testing via logic boards and oscilloscopes is horse shit. This is also why assinging a grade on a scale on something unscientific like this and passing it around is misleading. The brook converters are not shit tier and this asshole here is telling people it is.

    Was Jon testing against a PS360+ PCB? If not, then of course his results are going to seem lower when his control has more delay than the control I'm using.

    And I think you're confused about the aim for my tests. It's not to test the absolute lag of any stick. It is to test the extra lag between different types of sticks when used on a given console. An oscilloscope test wouldn't be useful for my purposes, as I'm determining the time it takes for the PS4 (or PS3/360) to read and interpret the inputs, not an external device.
    I see in frames IRL
  • gahrlinggahrling Better safe than Hori.. Joined: Posts: 3,968
    Teyah wrote: »
    Brook and CronusMax Plus are running the latest firmware, I'll add that information in on the next update

    Thank you.

    Teyah wrote: »
    I do have PCB pictures up for some sticks on the site (HRAP4 line, Qanba Q4 PS3 line). No need to void wareanties to add pictures of PCBs for brand new sticks like the TE2+ or TE-S+ which only have one PCB.


    Void warranty? Dude, the TE-S+ opens from the top giving you full access to inner parts without voiding anything. Does the TE2+ even come with a tamper-proof sticker?

    They only have one pcb each so far (that we know of), it would be wise to take note of their current print codes for future reference, along with photos.
  • VickoVicko Vegas Modder Extraordinaire! Joined: Posts: 2,830
    Warranties are horse shit anyway. I straight up called madcatz and told them I fried a pcb while modding and they sent me a new one. Please don't make these tests more laughable than they are.
    SRK Tech Talk Member of the Year 2015
    Need a stick modded? I can dual mod, repair, and customize your stick or Hitbox for PS4/PS3/XB1/360

    www.vickomods.com
    lol
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    I'm 99% sure he tested a native stick in ps4 mode and then the very same stick in ps3 mode with the adapter on it. It was what i suggested when he asked for testing setup advice.

    You don't honestly think we're stupid enough to test a different stick with an entirely different pcb vs a completely different one with a random converter on it, do you?

    If you aren't attempting to test the absolute lag of a stick, then why are you assigning absolute numbers? Why do you pass your "information" along as absolutes and buyer bewares? Why don't you step in and explain your results are not absolute when people are freaking out that their sticks might cost them wins?? Would appreciate some consistency here. FWIW a PS4 or any other console is an external device the PCB in question would be hooked up to, soooo...
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    I can't wait for all the people who were happy using their Brook converters to now whine about all of a sudden lag and start looking for alternatives because of this highly accurate and scientific test.
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    Hey guys remember when PS360+ had hundreds of happy customers and consistently sold out, then all ofna sudden they were laggy and everyone could all of a sudden feel it?
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • RBFRBF Joined: Posts: 1
    What is the matter with you? Take a deep breath and look at how you're coming across.

    Here's a video from a few posts up that you're ignoring

    If you can't watch it, use a VPN but the gist is that it shows the Brooks Converter (PS3 to PS4, and 360 to PS4) lagging compared to a HRAP4H.

    No one said that it's shit, Snake Eyez winning 3 WNF in a row with a converter shows that it's fine.

    Instead of tearing down someone's efforts because you forgot to take your meds, feel free to video your own tests and show us the results. Oscilloscopes are available on amazon.

    @Teyah, thank you for the work. Do you plan to test the Brooks Universal Fighting Board or the updated Hori RAP4 Kai(hayabusa+xinput)?
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    We've explained tons of times. I'm tired of repeating myself so maybe you go back and actually read the damn thread.
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • Smashbro29Smashbro29 Waiting for the new Framemeister... Joined: Posts: 7,199
    I don't get why people are so bent out of shape over a dude who decided to run some tests. I'm thinking that since this is tech talk, this is a really great starting point for this type of thing? shouldn't we be interested in conducting different tests in different ways since we're all pretty much here for reasons like this? Until someone else wants to take a stab at testing, I can't see why we're going to crap all over some work that someone in the community took their time to try.

    I love Teyah's work for PM but this test is inaccurate.

    Accurately doing this would cost a lot of money and this as is just gives us imperfect info.

    I love Teyah from Smash though, don't be mad!
  • MoonchildeMoonchilde Resident tech talk double poster Joined: Posts: 4,838
    You guys telling us to do a better job if we don't like teyahs work have obviously not read the thread in which we state multiple times the equipment is very expensive. Perhaps you may have also missed where many of us would be happy to test this in a much more controlled environment, but again it boils down to costs and needing a supply of pcbs. If people are willing to donate the funds to cover the costs of the equipment I'm sure there are people here who'd be more than willing to settle this scientifically.

    Unless some of you feel like putting some cash up front, then don't get mad us for pointing out the numerous problems of this testing and the inaccurate results.
    Disclaimer: I work for Paradise Arcade Shop. My posts are probably biased. Take that into consideration. Bye!
  • DarksakulDarksakul Your lack of faith disturbs me Joined: Posts: 23,821
    Jazz cat wrote: »
    It doesn't matter if someone thinks there is a flaw in his method..
    It does if all the information is incorrect, and people are making their investment choices on that incorrect information.

    We already explained time and time again all the faults in this thread, read the thread.

    I said it before this thread needs to get locked.
    There is nothing scientific about the claims here, its misinforming people and the information is based on terribly incorrect and it creates alot of misnomers for the larger community.
    “Strong people don't put others down... They lift them up.”
    - Darth Vader, Philanthropist
  • shamash.phishamash.phi Waler Master Joined: Posts: 73
    Moonchilde wrote: »
    You guys telling us to do a better job if we don't like teyahs work have obviously not read the thread in which we state multiple times the equipment is very expensive. Perhaps you may have also missed where many of us would be happy to test this in a much more controlled environment, but again it boils down to costs and needing a supply of pcbs. If people are willing to donate the funds to cover the costs of the equipment I'm sure there are people here who'd be more than willing to settle this scientifically.

    Unless some of you feel like putting some cash up front, then don't get mad us for pointing out the numerous problems of this testing and the inaccurate results.

    What tools a person will need for "making this more scientifically"? I'm curious about that...
    Sorry if I speak like shit.
  • DarksakulDarksakul Your lack of faith disturbs me Joined: Posts: 23,821
    Moonchilde wrote: »
    You guys telling us to do a better job if we don't like teyahs work have obviously not read the thread in which we state multiple times the equipment is very expensive. Perhaps you may have also missed where many of us would be happy to test this in a much more controlled environment, but again it boils down to costs and needing a supply of pcbs. If people are willing to donate the funds to cover the costs of the equipment I'm sure there are people here who'd be more than willing to settle this scientifically.

    Unless some of you feel like putting some cash up front, then don't get mad us for pointing out the numerous problems of this testing and the inaccurate results.

    What tools a person will need for "making this more scientifically"? I'm curious about that...

    Well for one there poor documentation on what actually been tested. The same stick can have many different revisions of the board inside.
    Science is on the big part documenting everything you can. Teyah admits at no time did he open up a stick and document what board is inside for fear of voiding the warranty.
    The original Xbox version of the Mad Catz TE Street Fighter IV Round 1 has at least 2 revisions of the board, the PS3 version had 3.
    Those revisions will have different Latency times.

    The equipment involved for that would be pen and paper, and maybe a magnifying glass.
    I suggested a method using a oscilloscope and a logic probe. Even if I was wrong, no real reason was ever presented how that could be wrong other than stanch opposition.

    Now we are hearing from Teyah
    Teyah wrote: »
    These aren't absolute numbers. If you read the overview you'll see that all of the results are relative lag compared to the best performing stick which is ranked at +0 ms.

    At no point on his original post is the disclaimer that these numbers aren't 100% accurate. Even with scientific testing you report your margin of error.
    And if you don't know the margin or if the results are relative it is your obligation to report that.
    “Strong people don't put others down... They lift them up.”
    - Darth Vader, Philanthropist
  • DarksakulDarksakul Your lack of faith disturbs me Joined: Posts: 23,821
    Xanfell wrote: »
    lol.

    I could say the same thing to you, do not get mad at Teyah for posting this information, as you have said you have raised your opinions on this, do you really need to discuss and repeat the same cycle.

    Lets just say Teyah used the confirmed hit count method and then charted those, the s tier would be 0 and the f tier would be 85 hits taken from the s tier.
    We would get people like you and say it is not scientific enough, add some numbers and retest.

    Then Teyah, uses his current method, "it is still not scientific enough" - we want the time the pcb takes to send data packets, not actual time it takes for the pcb + ps4 + game + display to translate that packet into an action.

    Lets just say Teyah uses the "More scientific approach" you guys just going to try and find something wrong with the data or the method he uses.

    Cannot please everyone can we?

    Mad Catz also used Teyah's information and methodology to product test the new sticks, which is pretty good, bigger companies are using this information to create products with lower latency.

    Since there is loads of videos on youtube testing product A against product B and the results matches up to this chart, I am happy with that, because one of them is going to be better than the other one.

    Mad Catz was at a position where they had to at least take a look and address the issues as one of their products was being labeled as the worst rating for latency.
    What was done was a PR decision not a engineering decision. Coming out as defensive would cost Mad Catz so they decided to be diplomatic. And I am sure their own engineers are looking into the issue on their own for product refinement (and to see if their is any validity of the claims here).

    I as a private individual and a member of this larger community decide to take a stand, if part of the information is faulty and everything is relative then how much of this info is skewed.
    I looked at the raw data and feel that the charts on page one no way reflects on the raw test data. The charts make an assumption that every stick been compared to every other stick in multiple testing, and that is far from the case. And there never been a satisfied explanation how millisecond results are given when milliseconds measurements were never made. I am sure someone at lets say a Aerospace Engineer with electronics engineering background who knows how controller latency really effects performance can but everyone here (both sides of the debate) to shame with actual citable sources.

    My saying its not scientific enough comes from people taking this testing as 100% truth and taking everything at face value which is happening and been happening rather than seeing everything being relative and more transparency was shown for the entire method.
    “Strong people don't put others down... They lift them up.”
    - Darth Vader, Philanthropist
  • devastatdevastat Joined: Posts: 23
    Wow, Mad Catz TE2+ looks to have such a small input lag. Thanks for the new tests Teyah.
Sign In or Register to comment.