Arcade Stick Suggestions

<blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/7169/oldschool_BR">oldschool_BR</a> said:</div>
<div class=“QuoteText”><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:</div>
<div class=“QuoteText”>When it comes to buttons, it’s all preference.  Some guys swear by them and claim that they feel better for pianoing inputs.  I like the light touch and clickiness of the Sanwas, it just feels very responsive and tactile for me.  </div>
</blockquote>

It’s virtually impossible to do frame-perfect pianoing on Sanwa buttons. I’m not aware of any non-09’er/SF4 teen who thinks otherwise. They have relatively loose springs and a very short activation distance, so it is very hard to press them soft enough so they deactivate on the next frame.<br></div>
</blockquote>

Is there any data to back this?  I haven’t encountered an issue with this type of stuff when trying to land reversals in ST.  <br><br>  <blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/28257/Born2SPD">Born2SPD</a> said:</div>
<div class=“QuoteText”><p>^ this, x10000.<br></p><p>As for the stick, its known that Its easier to do flick SPDs (a.ka.a Hajiki Screw) on LS-32. This is really good for thawk, walk in typhoons are soo easy that way. Harder to use with N.Gief though because of green hand. I also preffer the “feel” of the LS-32 over the JLF.<br></p></div>
</blockquote>

Still sounds like a preference issue.  Sanwas and Seimitsu sticks are always hotly debated, but for general all-purpose execution, there’s nothing objectively definitive that I’ve seen that proves that one stick is superior to another.  

<br><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:Is there any data to back this?  I haven’t encountered an issue with this type of stuff when trying to land reversals in ST.  <br></div></blockquote>I would like to see this as well. The switches in Sanwa are functionally identical to Seimitsu, the only difference is the force to activate the switches needed. What’s probably going on is a player’s fingers are heavy handed and not snapping off the button quick enough, leaving the button in activation mode longer. It doesn’t take much force to activate a Sanwa, however, that doesn’t mean the spring is in slow motion releasing the silver ball from the contact points in the switch. A bad switch with a nearly worn out spring maybe, but the switches are momentary contact, as soon as the ball breaks contact with the metal plates to the sides of it activation is lost, which results in a neg edge in the game. The ball has 16 ms to do this, it’s more than enough time.<br>

<br><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:<br>Is there any data to back this?  I haven’t encountered an issue with this type of stuff when trying to land reversals in ST.  <br></div></blockquote>1) Simple physics: looser spring -> less forece -> less acceleration -> less speed -> less distance covered to the square -> more time to the square;<br>2) Shorter activation distance -> longer deactivation distance -> more time to deactivate or more required hand control;<br>3) Item 1 makes item 2 harder when doing trills and multi-tapping;<br>4) I could reach 18 Hz with trills on Happ Competition buttons (which are great, if only they were a bit larger). Nowhere near on Sanwas. And around 14-15 Hz on Seimitsus;<br>5) Moonchilde, Sanwa buttons differ from Seimitsus in - at least - 2 ways: shorter activation distance and softer spring.<br><br><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:</div>
<div class=“QuoteText”>Still sounds like a preference issue.  Sanwas and Seimitsu sticks are always hotly debated, but for general all-purpose execution, there’s nothing objectively definitive that I’ve seen that proves that one stick is superior to another.  </div>
</blockquote>

I’m talking about buttons. Sticks are highly debatable, among several aspects, from ball tops, springs and throw distances to mounting structure, on lap or cab, etc. I feel several Japanese joysticks are troublesome due to being hard to reach diagonals without the square gate, which, in turn, I find it awkward. I feel I could get used to it on a cab, but it sucks trying to stay on main distances* if the stick is not anchored. I’m too used to playing with a heavy hand.<br><br>*and not slide to a diagonal<br>

<blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:</div>
<div class=“QuoteText”><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/7169/oldschool_BR">oldschool_BR</a> said:</div>
<div class=“QuoteText”><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:</div>
<div class=“QuoteText”>When it comes to buttons, it’s all preference.  Some guys swear by them and claim that they feel better for pianoing inputs.  I like the light touch and clickiness of the Sanwas, it just feels very responsive and tactile for me.  </div>
</blockquote>

It’s virtually impossible to do frame-perfect pianoing on Sanwa buttons. I’m not aware of any non-09’er/SF4 teen who thinks otherwise. They have relatively loose springs and a very short activation distance, so it is very hard to press them soft enough so they deactivate on the next frame.<br></div>
</blockquote>

Is there any data to back this?  I haven’t encountered an issue with this type of stuff when trying to land reversals in ST.  <br><br>  <blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/28257/Born2SPD">Born2SPD</a> said:</div>
<div class=“QuoteText”><p>^ this, x10000.<br></p><p>As for the stick, its known that Its easier to do flick SPDs (a.ka.a Hajiki Screw) on LS-32. This is really good for thawk, walk in typhoons are soo easy that way. Harder to use with N.Gief though because of green hand. I also preffer the “feel” of the LS-32 over the JLF.<br></p></div>
</blockquote>

Still sounds like a preference issue.  Sanwas and Seimitsu sticks are always hotly debated, but for general all-purpose execution, there’s nothing objectively definitive that I’ve seen that proves that one stick is superior to another.  </div>
</blockquote>

  1. Simple physics: looser spring => less force => less acceleration => less speed to the square => more time to the quare;<br>2) shorter activation distance => longer deactivation distance => longer deactivation time;<br>3) Moonchilde, they are quite different. The spring is harder on Seimitsus and the activation distance is larger;<br>4) The dead links on the button press thread I posted years ago had me doing trills at around 18 Hz on Happ Competition buttons. The exact same maneuver failed on Sanwas, as the button would be pressed again before deactivation. I can do around 15 Hz on Seimitsus.;<br>5) If you have the Input Watch program - or better yet, XSPR’s execution aid - it should be easy to try double and tripple tapping and checking the results. Doing frame-perfect multi-tapping on unmodded Sanwas is very hard.<br>

<blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:</div>
<div class=“QuoteText”><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/7169/oldschool_BR">oldschool_BR</a> said:</div>
<div class=“QuoteText”><blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/50728/eltrouble">eltrouble</a> said:</div>
<div class=“QuoteText”>When it comes to buttons, it’s all preference.  Some guys swear by them and claim that they feel better for pianoing inputs.  I like the light touch and clickiness of the Sanwas, it just feels very responsive and tactile for me.  </div>
</blockquote>

It’s virtually impossible to do frame-perfect pianoing on Sanwa buttons. I’m not aware of any non-09’er/SF4 teen who thinks otherwise. They have relatively loose springs and a very short activation distance, so it is very hard to press them soft enough so they deactivate on the next frame.<br></div>
</blockquote>

Is there any data to back this?  I haven’t encountered an issue with this type of stuff when trying to land reversals in ST.  <br><br>  <blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/28257/Born2SPD">Born2SPD</a> said:</div>
<div class=“QuoteText”><p>^ this, x10000.<br></p><p>As for the stick, its known that Its easier to do flick SPDs (a.ka.a Hajiki Screw) on LS-32. This is really good for thawk, walk in typhoons are soo easy that way. Harder to use with N.Gief though because of green hand. I also preffer the “feel” of the LS-32 over the JLF.<br></p></div>
</blockquote>

Still sounds like a preference issue.  Sanwas and Seimitsu sticks are always hotly debated, but for general all-purpose execution, there’s nothing objectively definitive that I’ve seen that proves that one stick is superior to another.  </div>
</blockquote>

  1. Simple physics: looser spring => less force => less acceleration => less speed to the square => more time to the quare;<br>2) shorter activation distance => longer deactivation distance => longer deactivation time;<br>3) Moonchilde, they are quite different. The spring is harder on Seimitsus and the activation distance is larger;<br>4) The dead links on the button press thread I posted years ago had me doing trills at around 18 Hz on Happ Competition buttons. The exact same maneuver failed on Sanwas, as the button would be pressed again before deactivation. I can do around 15 Hz on Seimitsus.;<br>5) If you have the Input Watch program - or better yet, XSPR’s execution aid - it should be easy to try double and tripple tapping and checking the results. Doing frame-perfect multi-tapping on unmodded Sanwas is very hard.<br>

Just to clarify things, we are talking about pressing multiple times the SAME button.<br>

I can see why you believe your theory makes sense, but I’d much prefer actual structured tests to be performed that verifies this along with consistent results.  Until such evidence shows up, I’ll try and avoid theorycraft and stick with the safe statement: “It’s all preference.”   

The theory behind it is more important! Anyone can perform controlled tests with a speed servo and either a controller (+execution aid or input watch) or buttons and a frequencimeter. And it’s not like one can not visually check than Sanwa’s return slower than other major button brands. Just release different buttons at the same time and see.<br>

<blockquote class=“Quote”>
<div class=“QuoteAuthor”><a href="/profile/7169/oldschool_BR">oldschool_BR</a> said:</div>
<div class=“QuoteText”>The theory behind it is more important! Anyone can perform controlled tests with a speed servo and either a controller (+execution aid or input watch) or buttons and a frequencimeter. And it’s not like one can not visually check than Sanwa’s return slower than other major button brands. Just release different buttons at the same time and see.<br></div>
</blockquote>

The theory behind it is nice, but it’s subject to debate and scrutiny without any tests.  I can argue while oranges are better than apples and provide solid pros and cons, but it’s pretty meaningless without any evidence to back it up.  And anyone might be able to perform controlled tests, but so far I haven’t seen anyone who’s actually bothered to do so.  As far as trying to visually identify a difference, I highly doubt anyone can eyeball it accurately and be able to discern any performance difference.  Human perception and imaging process is FAR too prone to errors and tricks. 

<strong>“1) Simple physics: looser spring => less force => less acceleration => less speed to the square => more time to the quare;<br>2) shorter activation distance => longer deactivation distance => longer deactivation time;<br>3) Moonchilde, they are quite different. The spring is harder on Seimitsus and the activation distance is larger;<br>4) The dead links on the button press thread I posted years ago had me doing trills at around 18 Hz on Happ Competition buttons. The exact same maneuver failed on Sanwas, as the button would be pressed again before deactivation. I can do around 15 Hz on Seimitsus.;”</strong><br><br>1. The spring isn’t that much looser that it would make it that much harder to deactivate. <br><p>2. If the activation travel distance is short, then the deactivation travel distance is equivalent. However, deactivation travel is much, much shorter than activation since a momentary switch needs only break contact to go back to neutral.<br></p><p>3. Having taken apart both Sanwa and Seimitsu micros, I can tell you the one main difference is spring tension. Any travel between the buttons is negligible. If there was truly a larger activation distance, then that would mean activation would be every slightly slower because more distance must be traveled to activate, although deactivation would be near instant since all that requires for deactivation is the ball break contact from the plates. However, all that is negligent, because the button activates when the ball comes in contact with the circuit plates. We’re talking very momentary contact for both types of switches, and near instant deactivation on release no matter the tension between the springs.</p><p>4. Is player error/playstyle/preference. You simply can’t accidentally keep the button activated by having your finger rested on the plunger on a set of buttons with higher neutral tension. Your input method is probably too heavy to the point that the buttons can’t force their way back to neutral fast enough. You’re relying on the buttons to pop your fingers off activation instead of quickly tapping. That’s ok, that’s why we have different button types, because it’s preference. Some people have a weaker touch and for them, Sanwa is faster than someone with a heavier handed touch.<br><br>I have a friend who consistently gets 1 frame links on Sanwa buttons. If they couldn’t handle 1 frame negative edging then they couldn’t do 1 frame links either.<br><br>Either way, ALL this is speculation until tested with a machine set to press the button down on frame 1 and release on frame 2 with an input watch to either back up or debunk your claims.<br></p>

Also, if I really have to, I can draw an illustration on why both of you have your theories wrong on Sanwa buttons.

  1. I said slower to deactivate, not harder. It takes longer for a fully pressed button to deactivate.
  2. Incorrect, in several senses. First of all, switches have hysteresis: they operate differently when activating and deactivating. Secondly, there are several types of switches, for several electrical and electronics applications, such as break-before-make and make-before-break. Finally, if the total travel distance is x, and even if there was no hysteresis, if the activation distance was a, then deactivation distance d would be given by

d = x - a.
3. The travel distance is similar: the activation distance is not. Which was part of my point.
4. But doing 1 frame links demands timing, not mashing or multi-tapping, so the button type does not matter.

  1. Harder was a poor choice of words. Slower, more difficult, whatever. Same difference. The spring isn’t that much less tense. It may take a tad bit longer to fully pop back to neutral with a spring that has less tension, but not to deactivate. See below:

As you can see, the spring is still nearly under full compression and tension when the switch is deactivating. It still has plenty of force behind it to pop the ball out of the circuit creating a negative edge.

The distance required for the switch to travel is negligent in this case. Neutral, maybe, but not deactivation.

  1. I’m sorry, I don’t understand any of that. What do those other types of switches have to do with the kind that are in Sanwa? Look at the illustration, the switch is a momentary contact switch. When deactivated, the circuit is broken, as soon as the metal ball comes into contact with the metal plates the circuit is complete and it sends the signal. As soon as you let go, contact is broken immediately and the signal breaks, along with the circuit. The game sees this as a negative edge. Regardless of the tension of the spring between a Sanwa and a Seimitsu, they’re going to deactivate at nearly identical speeds. If the spring couldn’t release that small of a distance fast enough within 16 ms (1 frame) then it wouldn’t even be able to hold the plunger up as it rests on the switch’s actuator. There is more than enough tension on the spring to release the ball from contact within that time frame. If you were to put your formula into my illustration, you’d see that d is negligible. Between a Seimitsu and a Sanwa, there isn’t going to be any difference. Popping back to neutral, there will be, which is what I should have written instead of deactivation travel (poor choice of words there :frowning: )

I believe your formula is a little flawed. Lets say that the total travel distance is 3 mm, and that activation is 2.5 mm. That means x is actually the throw, extra travel that the silver ball may slide along the contact plates. That means there is .5 mm of throw, but as you can see from the illustration, the deactivation distance is as soon as the ball loses contact with the places, which is even smaller of a distance.

  1. The activation distance shouldn’t be that different between the two switches, they’re builds are pretty similar enough with very minor differences. The main difference is the spring. If you took out the spring from a Seimitsu/Hori (manufactured by TKC) and placed it into a Sanwa, then it would feel just like a TKC switch.

Lets consider something else. I’d wager that a Sanwa is actually faster to activate, since it requires less force than a Seimitsu. So you should be able to activate and deactivate faster on a Sanwa since your fingers require less force to push the switch into activation, and if you put more force, it would do it faster since it offers less resistance. Since deactivation is nearly instantaneous, the only factor left is returning to neutral, which will be a little bit slower than Seimitsu. However, since you can simply reactivate it before it pops back into neutral, this point is moot, and a Sanwa is functionally just as fast, if not faster than a Seimitsu.

Thus I mention player error. If you can’t move your fingers fast enough, then you will have more problems on a Sanwa since slow fingers will keep it in activation longer because of the weaker spring in the switch. However, a button with a spring that has more tension may be able to force your slightly heavier touch back to neutral faster, so rapid fire presses may be easier to do in this case, and may be why you can rapid fire faster on buttons of higher tension. There is nothing wrong with admitting this, and I personally prefer buttons with higher tension for more tactile feedback and because I accidentally activate Sanwa’s when my fingers are simply resting on them. Obviously, if I really wanted to play on Sanwa, I could, it would just take more practice and self awareness to correct the issue. I don’t even like Sanwa buttons, but from a technical standpoint, with a lighter touch or forcefully snapping your fingers off them faster could potentially be faster than Seimitsu simply from the fact you can activate them faster with less force.

  1. Doing 1 frame links certainly demands timing, and most people use plinking to do so because most humans aren’t capable of consistently hitting 1 frame links without pianoing. I’m sure you’re aware this is a form of piano method to get more inputs per hit via input and neg-edge to increase your chances of landing those 1 frame links, just like how in ST you piano to increase your chances of landing the 1 frame reversal window. My friend who I play locally with can do this consistently on Sanwa so I know the buttons are perfectly capable of 1 frame neg-edging.

Anyway, unless there is some very hard data to support the idea that Sanwa can’t do 1 frame neg-edging, then I’m going to remain a disbeliever on this topic.

The images provided are just not accurate ways of displaying when buttons activate and deactivate. As a matter of fact, the travel distance before and after activation on them are quite far from the behavior presented by the most common push-buttons used in arcades (Happ, IL, Saimitsu, Sanwa). Also, one must recall the whole set (push-button + installed switch) must be considered.

The spring is less tense. Thus, it is guaranteed to generate slower button recover movement. Now one may only quantify how much slower it is, but never that it isn’t slower, despite physics and reality explaining otherwise. Any person can easily test this: the Input Watch and Execution Aid programs are around, and articles or videos on putton pressing techniques are also around, including the one I posted in the ST section, which applies to ST. Thus, assuming one has an arcade stick ready, all you need is a PC.

Activation distances depend on switches and buttons. They are different due to different part dimensions or sizes, different operating principles, and so on. Reality is buttons have very different travel distances, activation distances, hysteresis*, due to several reasons.

The idea of fast finger movement being better on Sanwas is unwarranted. That’s precisely the opposite: I did test the exact finger movement for [trill mashing](Mashing and Button Pressing in ST on 3 different button brands, and Sanwas were the ones which stopped responding first, that is, which stopped deactivating before the button was pressed again.

Negative-edge is not used in SF4’s priority linking. I am aware of the technique, even though I do not have the slightest interest in SF4 mechanics. The game works as anti-SF2 in several senses, from gameplay to moveset, from hit and hurt boxes to mechanics. The priority linking glitch does not exist in SF2.

*which we can quantify as difference between how much you must push a button for the switch to close contact and how much you can release it before it deactivates. This can be extreme in a few types of switches and joysticks.

How is the image not accurate? It’s exactly what the inside of a Sanwa/Seimitsu/MadCatz switch looks like during neutral, activation, and deactivation. Of course I’m not going to draw in detail what the spring looks like exactly, so a black line will do to get the point across, and I am not going to take a photo by photo shot of the insides, and measure the ball contact, the contact plates, the springs, the white actuator, just to prove to two people on a forum of thousands when the image already does the job fine. It’s the functionality of the switches, they’re nearly identical with the main difference being a slightly (SLIGHTLY) less tense spring. Please explain why the image is incorrect or draw one of your own and back up your claim.

The spring is slightly less tense, but under full stress the force during release will be more than enough to remove the ball from the contact plates within a frame. MORE THAN ENOUGH. Slower back to neutral doesn’t even matter, because you can activate again before the button pops back to neutral, so a slower spring is moot unless the button was manufactured in a way that required neutrality before being activated again, which isn’t the case.

Please explain why faster finger movement with more force behind the fingers on a spring with less counter force would activate slower than a spring with more counter force against the force of your fingers. It’s not possible, something with more counter force is going to be harder and slower to move than something with less.

You’re a human. You can do trill mashing techniques and all that jazz and watch inputwatch but when it comes down to it, you are not a machine with an internal timer to activate and deactivate on a frame by frame basis, nor are you capable of rapid fire without fatiguing your hand nor are you able to control the force behind your fingers so that it’s 100% consistent for all button types. The only way to test it is via a mechanical method, and your human method is currently flawed nor is it a controllable test environment.

Again, until there is hard data, I simply can’t believe you. As much as I respect you guys and what you’ve done for the SRK Wiki in the ST section, I simply can’t back your claims because they just don’t make sense. If you could provide a mechanical test with a controlled method, then I’ll happily admit I’m wrong and accept the data, HAPPILY. Or if one already exists, I would gladly admit I’m wrong.

What the fuck are you guys talking about? Are you guys just copying the encyclopedia and pasting it on here?

My content will be in an encyclopedia. Eventually. :slight_smile:

Some part of me wants to build a ‘pressure turbo’ system where the button has an eletromagnetic return so it oscillates when you push down, and, the harder you push, the faster the turbo goes.