The 1ms firmware vs the 10 ms firmware is only different in that when the host asks it about it’s features/details, it tells it one or the other. The actual processing speed of the PCB is the same either way. Some hosts will make use of that 1ms or 10ms specification and only check the controller for new data in those time intervals. So, though I got excellent response times on my host, I can’t speak for what the PS3 or X360 does regarding polling any of these controllers.
If this experimentation were to be air tight, it should be done using a hardware USB analyzer that can sniff the communication between the console and the controller to see the actual response time from the moment the console polls the controller to the moment the console receives a completed response. Anything outside of that depends on how the console/game handles said data, which is not guaranteed to yield the same results every time (depending on what the console/game program is busy handling at any given moment).
-ud
I wanna know why this even matters? None of its honestly noticeable to the naked eye. As long as the inputs are reading correctly at whatever given speed then no one will ever know the difference. This whole thing about being able to tell the difference from this stick to this stick is bullshit. Test like these take away money from people that mod. sticks. Honestly, I’ve used just able every stick on that list with one of those setups and never noticed the difference. If, I was to say I did I would be lying to myself and every user here.
Wrong, wrong, and so fucking wrong. Go into USF4 training mode and switch network simulation from 0, 1, 2, 3, and 4. If you can’t tell any of these apart then there is something wrong with you.
It’s meaningful in terms of the numbers I’m coming up with for comparability’s sake. Not in terms of people being affected by timing, etc.
Nobody has done this so far as far as I can see. And I’ve been following the thread for quite awhile
Also that big list you posted there … doesn’t make any sense. How is HDMI/receiver/TV lag going to affect a PCB in any way? Do you expect wires to cause more lag for one side than the other (or any lag at all for that matter)? Do you think the PS3 or 360 will favour P1 side over P2 (even so, I switched sides on my tests halfway through)? I think you need to step back and gain a better understanding of what is being tested and what controls are in place. The overview page explains it pretty well: http://teyah.net/sticklag/overview.html
With the exception of the “human mind is not fast enough” bit, everything you said there is true, yet we can still observe (for example) SF4 on a 360 delay 4 frames consistently every time, and can notice a difference when it is instead consistently 5 frames. At a one frame difference something seems wrong, two frames it is very noticeable, three frames are right out. You aren’t suggesting that it would take >4 frames delay on a monitor for a fighting game player to notice it was lagging? With a two frame delay people will consistently whiff jump-ins in games like ST or Alpha.
Excellent reply. This “lag” testing is obviously way more complicated and has way more variables than I previously thought.
I must say, that even though I feel that I don’t really care about manageable latency, if I can get the one with less, why not?
Then on the other hand a sanwa modded namco with a Cthulhu PS3 in it sounds awesome. (I assume you guys are talking about different models, he’s testing the PS3/PC version, and maybe you are talking about the MC Cthulhu with the upgradable firmware?)
Yeah, it’s not absolutely airtight. That is why several hundred trials are used to smooth out any inconsistencies, as you would expect these different results in polling and handling of data to not favour either side in the long run.
Even after 1,000 trials though, you can see some variance when 3 of the same PCBs were tested: two of them were < 0.10 ms apart, and in one case there was a ~0.30 ms difference in the results. For purposes of comparability though (which is my main goal with these tests), a 0.3 or even 0.5 ms variance isn’t very material. If the variance were higher, like 2-3 ms, then I would consider an alternative testing method. But I am willing to live with knowing that realistically I could be half a millisecond off on each result. Whether the Q4RAF Black is 2.1 or 2.6 ms and the MCZ Pro is 23.5 or 23.0 ms doesn’t really matter to me (and I’d think, to most people viewing the results).
I’ll just say that it takes a lot of time to do 1,000 trials and that it’s not always practical to take over one full setup on game night when there are people waiting to play. So I had to draw a line somewhere. Once I saw consistent results coming in, I figured 1,000 per stick would be a very good approximation while still being practical to do without taking up hours and hours of my day.
We don’t even know if he’s using the latest firmware…
I wish someone here who uses a PS360+ would speak up about it and whether or not they notice any lag. I’m playing USF4 right now with my TE-S and it seems completely fine to me. Granted, I’m not sure what it would feel like if I had something better to compare it to…
I have an Eightarc as well as a custom stick with a PS360+ and have always noticed something slightly “off” with the 360+ on my X-Box. I also play on a 55" LED. I could never quite put my finger on it, but this might explain something. I definitely have whatever the “latest” published firmware.
I just played a friend local last night and was hitting 1 frame links. No noticeable lag.
I don’t even have to do that. I’ve gone from playing on a TV with 5 frames of lag to CRT and back. Yes, I’ve noticed in those cases. Was I able to adapt? Yes.
I’ve played on a friend’s TV with 2 frames of lag. Not really noticeable, not enough to kill my game or make me drop combos or miss reacting to an opponent.
I’ve played on a CRT longer than I’ve ever played on laggy setups. I know what lag feels like and what it doesn’t feel like. 1 frame is not a big deal. There is nothing wrong with me, my brain adapts and that’s how the mind works. After a few minutes on the setup you should be familiar enough with it to be able to play. 1 frame, 2 frames. 3 frames is where things start to feel delayed but aren’t really visible. 5 frames is where things feel delayed and is slightly visible, but not enough to kill anything completely. 8 frames is the dead zone, totally feel it, totally see it. But uh, I’ve still played with 8 frames of delay on occasion just because. I actually played through Ni No Kuni with 8 frames of lag because I wanted frame interpolation since 30 fps looks like shit and the game wasn’t timing sensitive.
So yes, I know how lag feels and what is acceptable and what isn’t.
Then why have a tier list? You have S tier vs A tier and the only difference is 2 ms. Please put things in perspective because you’re going to spread paranoia that everyone is now going to be worried their PCB is lagging. If you’re going to spread information you have responsibility for the consequences of it.
What does the questions you posed about the list have anything to do with this? If people are experiencing lag in most cases it’s going to be the chain of processing from Point A in the PCB all the way to point Z in the display.
I understand what is being tested. I’ve already explained why I think it’s flawed, since these boards are not being tested in a more controlled environment and are only being tested on a single game (GGXAC) Also, your page doesn’t explain how you arrive at the numbers. Could you post a formula? Did you have to analyze hours and hours of PVR footage frame by frame? Are you recording at sub-frame speeds to get numbers in the ms range?
I think it would be far more beneficial to do this the UD way. Then, you’re getting nothing but the raw PCB. You’re cutting out all the possibilities that something else might be different. If PCB A has 2 ms of delay on his method and PCB B has 5, then that’s what it is. In the case of a PCB performing differently on a different console, that most likely has more to do with the console than the PCB, since the PCB is still sending out the information as it is. There is also no variable software layer in his method like a game possibly being messed up somewhere in the control code.
Oh ok. Well hopefully he did. I want to get a PS360 because the nI can use it on my ps2 and any other console pretty much and not have to carry around a bunch of sticks to my friends and other places.
I’ve played on my own stick with a PS360+ in it on a Xbox 360 just the other night and played on a friend’s stick with a pad hack and didn’t notice enough of a difference to blame anything other than my own execution. I was able to do the same things on both sticks.