Actually, the typical method is to tie the buttons together electrically, and that’s much much more precise than using a video recorder.
Also, I just tested the PS360 PCB’s xbox side and it was a frame ahead in 32 of 150 inputs - so it’s probably 3.5 ms faster.
actually the buttons were chained together in the video. so pressing 1 button on the RAPVX would press the same on the TE.
Thats my mistake then. I glanced at the video but saw nothing showing that he wired to two stick together to the same button.
Per chance I doubt he gave a listing as to the length of the wire used. A shorter wire for the hori would mean that its signal would have to travel a shorter distance than the MC TE. Ergo create lag.
Hence why NKI’s video method worked so well.
So Toodles,
How much lag is on your pcb?
But if the button press lit up the button but still had lag, it wouldn’t matter, would it? Also -
I’m inclined to believe Toodles cause it seems like he knows what he’s talking about in this area.
Ok so my friend (stick modder) replicated the test in the OP with HRAPVX v madcatz stick and his initial results have confirmed TE/SE PCB has a random 0-1f of inconsistent lag (sorry no vid). I think he said around 1/3-4 inputs the madcatz would drop a frame. Strangely the madcatz pad PCB also did the same but a LOT less 1/20-30 times it would drop. He’s gonna prob do more testing on pc and what not, but yeh confirmed other results more or less so far.
The timing differences we’re seeing would require hundreds of miles of length difference.
It does seem like the MadCatz sticks are the slowest of the bunch, but the difference is tiny. Comparable to the advantage you’d get by hit-confirming using the score or life bars instead of the character animation.
is this really going to be used as an excuse why players cant win matches? a random 0-1f… the lag must really be an issue to those top players using retail sticks when they compete.
Oh please. I thought blaming a loss on something other than oneself was a huge part of ‘the scene’.
A slow ass game like sfiv means this is not that big a deal, but random lag on a stick isn’t cool. Nowadays with HDTVs, ps3 built in lag, and now madcatz, we are facing an uphill battle to get everything “just right”. It’s a huge pain in the ass.
you want to make a claim like this then you need to make everything equal and not shortcut your experiment. Maybe your right and it would take hundreds of miles of wire to make that difference. Or maybe it takes the hori PCB 14 ms for the button press to get to the console and the TE 15-16ms. Ergo even with the TE stick being only 1 or 2ms slower than the hori it can create the illusion that is actually a full frame behind. You don’t know because the original tested BSed his test.
Maybe mathimatically speaking the timing difference because of the wire is moot, but if your going to be doing some “offical benchmarking”, then you need to make as everything equal as possible to get the truest results.
Even the offical evo monitors would randomly have lag that fluncuated between 0 and 20ms. On average it was under 1 frame but there are the occassional spikes that would be 1 frame of lag. Anyone using a monitor like this and not a SD CRT has no right to complain about some random few ms lag which could equal 1 frame of lag under the right circumstances since your playing with lag anyway from the monitor/tv, possibly your system, online play and now controller lag.
Guess we’ll need to start another topic and find a way to lag test all pcbs.
Indeed. But even if Sf4 is slow its a pretty link heavy game, doesnt help either
I wired buttons on my Madcatz SE and PS360 together. In 150 inputs, the Madcatz was a frame late 32 times the PS360 on zero. A frame is 16 2/3 ms, so that’s an average delay of 3.5 ms. 0.0035 light seconds is about 650 miles - cut that in half for slow conducting, and it’s still better than 300 miles. I can assure you that there wasn’t enough difference in lead lengths to account for that delay in my experiment.
Can you make a simple drawing on how you made the setup? Just want to know so everyone uses the same type of setup (I know there´s not many types you can wire it differently but still it would show how it is done so anyone can test it for themselves).
Nominated for article.
Nice track record, Madcatz. First a fightpad with no diagonals, then a fightstick SE that craps out on you, and now this.
The best way would be to use a switch separate from either board to control gates that separate the two boards. When I did the test with the wired vs wireless, I used a 4066 chip, and the separate pushbutton activated the gate to connect the two wires. The video of the TE vs HRAP testing appear to use optoisolators to do the same thing.
Just wiring the signals together to single button by itself is a very bad idea and will skew results.
People winning tourney’s in the U.S. are using TE’s, and Daigo won Evo with a TE… honestly the difference is probably obsolete.
actually, this whole conversation makes me wonder about wiring resistance - I wonder, do some people wire up the stick/buttons/pcb with thicker wire, in an attempt to reduce electrical resistance? I know some audiophiles do that for speakers and such.
So which stick doesn’t have lag?
My only beef with the entire thread is this. I would be nuts if I had to deal with an average of 35 ms of display lag. That’s a consistent input lag of 2 frames due to the display alone. Sorry, but this is not really alright for a fighting game, just not catastrophic like, say, 67 ms.
I have two monitors, a Samsung 931BW and a Samsung 2232BW. I feel very much encouraged to use the former, even though it’s 19-inch, as opposed to being 22-inch, just because of the input lag difference (the Samsung 2232BW in the worst case supposedly has an input lag of 45 ms, and an average of 28.1 ms).
Samsung 931BW
http://img1.digitalversus.com/produits/36/1625/36_152_112.jpg
X
Samsung 2232BW
http://img1.digitalversus.com/produits/36/2683/36_284_112.jpg
Source: Product Face-Offs Monitors - DigitalVersus
Aside from the above information, I have taken pictures of timers in a cloned video test myself and noticed the bigger monitor does indeed present a bigger input lag than the other.