Technical knowledge compilation (Gamecubes, controllers, TVs, recording)

[SpiderMad]

Yeah, and also wood is sturdier and less prone to deformation than rolled up thin plastic.

I look forward to your guide on that then

 

[Kadano]

I had announced to only write new tutorials for when my website is online, but it seems this will still take some time, so here’s something new that many have been requesting. (I had documented this in a video tutorial more than a year ago, but admittedly that video isn’t reallly good.)

Decreasing controller button click resistance by perforating the rubber plates (decreases finger strain during clicking) [mod B5 in my modded controller shop]

1. Open your controller with a triwing screwdriver

2. For the L and R triggers: Hold down the trigger to push the rubber plate out a bit, then grab it by its ridge and pull it out.

3. Depending on how much you want to decrease the click resistance, either perforate the plate at ~8 spots with a sharp scissor or screwdriver tip (decreases to about 70% resistance) or cut into it with small scissors (~30-50% of original resistance).

4. Put the rubber plate into its position again. I find it easiest to do this with a small screwdriver.

5. Put the controller together again and playtest whether it improved. If the resistance is still too high, take the plate out again and increase the size of the perforations.
If the resistance is too low or the click doesn’t work properly any more, the holes are too large and you need to replace the plate.

This modification can be done for all digital buttons except Z. (The Z button uses a different switch type that can’t be easily altered / replaced.)

 

[SpiderMad]

I had announced to only write new tutorials for when my website is online, but it seems this will still take some time, so here’s something new that many have been requesting. (I had documented this in a video tutorial more than a year ago, but admittedly that video isn’t reallly good.)

Decreasing controller button click resistance by perforating the rubber plates (decreases finger strain during clicking) [mod B5 in my modded controller shop]

This modification can be done for all digital buttons except Z. (The Z button uses a different switch type that can’t be easily altered / replaced.)

How many holes and where do you recommend (perform for your requests) for A/B/X/Y?

Also the way my unmodified old controller is (where hard press works but the click isn't there at all, unless you drop it hard and it comes back and then it goes away after a few games: which has happened twice now) is technically the MOST ergonomic of all: with the downside being less consistency/accuracy. Overall not worth the trade off compared to a very weak but tangible click, but still very usable.

 

[Kadano]

Just the same as for L and R.

 

[EWC]

Is there any special trick to detaching the potentiometers when removing a stickbox? I've had a couple of them break when messing around with spare controllers. I want to be able to perform maintenance on my main controller without risking damage.

 

[SpiderMad]

So at a tournament on Sunday, my controller was having the shield being held on occasion. This has happened before a couple times long ago, but it was usually the light shield. This time it was the hard shield, and slamming it against objects was the only way to get it to dissipate.
https://youtu.be/1YdMBxFgr98?t=149

Months ago I posted about how my controller normally somehow doesn't have any tangible hard press click but it still works after going past a certain point (that seemed to vary just a tad for some reason). And then if you drop it hard on accident it sometimes got the click back for a few matches and then went away.

Kadano surmised the hard press rubber was completely broken from my post.
This event Sunday made me open it up finally, and it was indeed broken. Now it all makes sense.

At first I replaced it with a normal hard press rubber. I don't like normal ones though, so I was going to put some holes into it like I've done to my "modded/2ndery" controller. But then I forgot I never realized how it got the click back was because it went behind the tube.

So I just put the broken one back in and pushed it behind the circle. This state of the hard press makes it essentially as if you put holes separating like almost all of the plastic (the weakest possible you can make it). The only problem is the click goes away, so I'm pry gonna have to try inserting a small piece of film or something to try and get it from not going past the outer circle of rubber.
UPDATE: Yeah I need to find some way to get it attached again. I may use glue. I wish I had a strong laser to just melt it together.

The question now is how the heck my casual friends used this controller hard enough to break all the rubber on everything.

Also for some reason there are no screw holes on the trigger plates (or bars either) in which case the spring just keeps it held (As you can see in the video at 2:23). Also the springs are shorter than that of my black controller.

If there were a way to slightly reduce the extreme looseness of the stick, that'd be an improvement (Kadano mentioned using some kind of glue or something in the oven, I feel this would cause more havoc than good).
EDIT: Yeah, too much work and risk to take it apart. It's also a Type 2 stick-box so that explains why I like it.

As well as using a modded face-plate to get perfectly consistent shield drops, although my rate with this controller is becoming pretty decent.

Also the Z button's "button box" isn't as easily pressed out as other controllers, but tolerable (if there's any way to fix this).

Also I'm pretty sure now either the
-Worn out control stick piece/rubber
-Worn out control stick gate on the face-plate
-Difference in control stick boxes/looseness
definitely causes me to get much better wavedash lengths with my old worn out controller than with others (especially my black which I'm trying to mod extensively).

Also with my black controller I've been modifying, I've decided to for-go having a spacer for the moment. Instead I may just remove the spring entirely for my L (which I use for everything) and just have a spring in my R (which atm I have none in and use for power-shielding). Then I can possibly try to muscle-memory my L-cancels and Light-shielding to be instead on R so that my L is very ergonomic to Wave-dash with. I have reason to believe though that doing everything with one trigger is best, solely because of the way they teach people to drive with only one foot to use between the gas/break (though this could have to do with when there was stick shift?).

 

[Pauer]

Kadano , I remember you recommended the Asus VG248QE to me due to its great response time for playing melee on an LCD. I went ahead and ordered it since I needed a new PC monitor anyways, my question now is what would you say is the best way to play on it? If I remember correctly, you play on your CRT Monitors by using a component -> vga converter, right? Back when we talked about it a while ago you said I should just get a VGA cable for the wii from ebay or so. Also, the monitor doesn't have a VGA slot but a DVI one, but I don't think there's a problem using a converter from VGA -> DVI, right?

 

[Kadano]

Kadano , I remember you recommended the Asus VG248QE to me due to its great response time for playing melee on an LCD. I went ahead and ordered it since I needed a new PC monitor anyways, my question now is what would you say is the best way to play on it? If I remember correctly, you play on your CRT Monitors by using a component -> vga converter, right? Back when we talked about it a while ago you said I should just get a VGA cable for the wii from ebay or so. Also, the monitor doesn't have a VGA slot but a DVI one, but I don't think there's a problem using a converter from VGA -> DVI, right?

I recommend the VG248QE for Dolphin setups or Wii → HDMI converter cables. For VGA from console, I’d say either a CRT monitor or the AOC G2460FQ (which also has very low lag with about 4 ms and all four standard inputs – VGA, DVI, HDMI, DP).

 

[Pauer]

I see the problem here is that I can't find any of the wii -> hdmi converters fizzy mentions in his article.

https://sewelldirect.com/wii-to-hdmi-converter
http://www.neoya.com/wii2hdmi
both sold out
the sewell one is available on amazon but doesnt ship to Austria.

I think I'll have to go with the
C&E Component RGB YPbPr to HDMI converter
In the comments of fizzy's article someone says blackouts can occur but depend on the energy socket it's plugged into. Additionally, it's rather expensive. Any other suggestions?

 

[Kadano]

I see the problem here is that I can't find any of the wii -> hdmi converters fizzy mentions in his article.

https://sewelldirect.com/wii-to-hdmi-converter
http://www.neoya.com/wii2hdmi
both sold out
the sewell one is available on amazon but doesnt ship to Austria.

I think I'll have to go with the
C&E Component RGB YPbPr to HDMI converter
In the comments of fizzy's article someone says blackouts can occur but depend on the energy socket it's plugged into. Additionally, it's rather expensive. Any other suggestions?

You can get the same device from AliExpress for much less: http://www.aliexpress.com/item/5RCA...-Audio-Converter-YPbPr-RGB-R/32236376607.html

They also have this one: http://www.aliexpress.com/item/Free...-to-HDMI-converter-adapter-box/926684756.html

But idk if a rebrand of that has ever been reviewed.

AliExpress is trustable and you can pay with bank transfer, without any additional fees. I’ve ordered lots of stuff there.

 

[Pauer]

that sounds great, actually. Thanks a lot as always!

 

[Gearitz]

So I have two original gamecube controllers one which is a platinum with the metal rods in the triggers and a black one without them or a protective covering however on the motherboard where the wood paddle receptors for the triggers are there is this black plastic I believe it is the triggers cover.

 

[KP17]

Kadano : Do you mind posting a picture of the specific tools used to puncture the rubber plates underneath the triggers? The screwdrivers I have access to put too much force on the plate, and after making 4 decent punctures at the cardinal directions, attempting to puncture a hole in between those locations led to the plate tearing. I tried cutting slits with a nice pair of scissors, and that worked fine, but only led to about 30% decrease in resistance which still is uncomfortable for my finger. Any thoughts?

EDIT: After testing with more plates, when I am successful with puncturing the plates to get the desired resistance (about %50-%70), my R trigger starts to lightshield (this trigger has a spring removed). Basically it will light shield as if there was a spring in it, and I have to shake my controller to get the R trigger to move to its unpressed position to get the shield to go away. I know I can 'trigger trick' to make this problem go away but I am wondering if this problem is caused by structural damage to the plate? Maybe after more use it'll just tear and won't be playable?

 

[Kadano]

If you can "trigger trick" it, it's not caused by the plate directly. I suggest disabling the analog input by holding down the R trigger while putting the back plate on, with the pot slider all the way up.

 

[Stride]

Concerning replacement stick caps: I have recently ordered every distinct type of stick cap I could find on DealExtreme and attempted to fit them all onto a controller. Unfortunately, none of them seem to work (for various reasons).

PlayStation type
http://www.dx.com/p/replacement-joystick-caps-for-ps3-wireless-remote-controller-pair-72284
http://www.dx.com/p/replacement-thu...or-ps3-ps2-controller-grey-black-2-pcs-237626
These fit onto the stickbox, but sit considerably higher than the GameCube stick cap; this means the bottom part of them comes into contact with the shell in some positions and creates too much friction (sometimes to the extent of getting stuck). If the stick gate were to be expanded to accommodate the raised position then these ones could work.

Xbox 360 type
http://www.dx.com/p/replacement-thu...er-for-xbox-360-controller-black-10pcs-343218
http://www.dx.com/p/replacement-thu...et-for-xbox-360-transparent-white-pair-227119
I couldn't get these to fit onto the stickbox at all.

Wii Nunchuck type
http://www.dx.com/p/replacement-plastic-3d-joystick-cap-for-ps4-black-4-pcs-291001
http://www.dx.com/p/replacement-wir...g-stick-caps-for-ps4-transparent-2-pcs-318604
These fit onto the stickbox but sit too low, which causes them to come into contact with the potentiometers when moved downwards or to the right.

Kadano , have I done something wrong, or did you only test whether the caps could be fitted and not whether they left the stick fully functional when you tried them yourself?

 

[Kadano]

I only bought some of them, and I only put them on the C-stick. Some of them fit well there, others took a bit of reshaping.

I don't have the time / energy right now to do an in-detail post on them right now.

 

[DRGN]

Concerning replacement stick caps: I have recently ordered every distinct type of stick cap I could find on DealExtreme and attempted to fit them all onto a controller. Unfortunately, none of them seem to work (for various reasons).

PlayStation type
http://www.dx.com/p/replacement-joystick-caps-for-ps3-wireless-remote-controller-pair-72284
http://www.dx.com/p/replacement-thu...or-ps3-ps2-controller-grey-black-2-pcs-237626
These fit onto the stickbox, but sit considerably higher than the GameCube stick cap; this means the bottom part of them comes into contact with the shell in some positions and creates too much friction (sometimes to the extent of getting stuck). If the stick gate were to be expanded to accommodate the raised position then these ones could work.

Xbox 360 type
http://www.dx.com/p/replacement-thu...er-for-xbox-360-controller-black-10pcs-343218
http://www.dx.com/p/replacement-thu...et-for-xbox-360-transparent-white-pair-227119
I couldn't get these to fit onto the stickbox at all.

Wii Nunchuck type
http://www.dx.com/p/replacement-plastic-3d-joystick-cap-for-ps4-black-4-pcs-291001
http://www.dx.com/p/replacement-wir...g-stick-caps-for-ps4-transparent-2-pcs-318604
These fit onto the stickbox but sit too low, which causes them to come into contact with the potentiometers when moved downwards or to the right.

Kadano , have I done something wrong, or did you only test whether the caps could be fitted and not whether they left the stick fully functional when you tried them yourself?

I searched for ages for caps in the past. Try these. I'm not quite sure, but they might wear a little faster than usual. But they feel good, and work. I might be wrong about the wear, because I'm judging by the wear on someone else's controller that I put one of these in, and they might just wear theirs faster than I do.

 

[Kadano]

Stride DRGN the same Gamecube-type stick caps that are listed for 3$ per piece on GamingGenerations can be had on AliExpress for 1$ per piece (comes in sets of 10). http://www.aliexpress.com/item/Free...Cap-for-NGC-gamecube-console/32266797099.html

I have them, they are quite good. They fit a bit loosely and the texture doesn't have as much friction as the original, and they are a bit shorter iirc. But altogether, I'm happy with them.

 

[SpiderMad]

Are the textures for Black vs Purple vs Orange etc. different? Or is it only in my head? I feel like if you tried really hard you could say feel the difference between a black and a purple. I feel like I love purple.

Also what's the best way to PURPOSELY smooth out the grooves on a control stick?

 

[BirdmanJR]

If ALL I care about is input lag is there monitors with less input lag than my XL2420Z ? Anything with less input lag than my XL2420Z ? Even CRT monitors ? Even if its just a small difference Id like to know.



http://www.cnet.com/products/aoc-ct720g-17-crt-monitor/

I would need to get a DVI to VGA conveter. Would this have much less input lag than the BenQ?

 

[Dolla Pills]

Any recommendations for a converter to connect my Wii to my CRT monitor (one VGA input port)?

EDIT:
Actually I was mistaken and that's not hard to find. What I don't know how to do is to how to get audio out of my Wii in this case.

Is there something I can connect the audio cables to that would give me the audio (preferably not desynced from the video going to my monitor)?

EDIT AGAIN:
So what I'm thinking is I'll use these cords to get VGA output from my Wii and then I'll use this adapter to connect the audio to some cheap USB speaker that I'll buy. It's slightly pricey but I can definitely afford it, but does this make sense?

 

[SpiderMad]

So what are the tried and true methods of WEARING OUT a control stick to make it looser? Should I just make aggressive rotations with it for a half hour each day?

 

[Kadano]

Any recommendations for a converter to connect my Wii to my CRT monitor (one VGA input port)?

EDIT:
Actually I was mistaken and that's not hard to find. What I don't know how to do is to how to get audio out of my Wii in this case.

Is there something I can connect the audio cables to that would give me the audio (preferably not desynced from the video going to my monitor)?

EDIT AGAIN:
So what I'm thinking is I'll use these cords to get VGA output from my Wii and then I'll use this adapter to connect the audio to some cheap USB speaker that I'll buy. It's slightly pricey but I can definitely afford it, but does this make sense?

Yeah, that'd work just fine. 26$ for the cable that looks like it's not genuine Mayflash is a bit more than usual though. It does ship from USA, so delivery will be faster, but if you want to save a bit by ordering from China: http://www.aliexpress.com/item/AV-H...intendo-Wii-Sony-PS3-Console/32355370301.html

I've ordered the same from there and it works fine.

 

[Dolla Pills]

Yeah, that'd work just fine. 26$ for the cable that looks like it's not genuine Mayflash is a bit more than usual though. It does ship from USA, so delivery will be faster, but if you want to save a bit by ordering from China: http://www.aliexpress.com/item/AV-H...intendo-Wii-Sony-PS3-Console/32355370301.html

I've ordered the same from there and it works fine.

So I managed to get everything working on Monday (I ended up having to blindly navigate my Wii to change the display setting to 480p) and I have to say it's one of the best set ups I've ever played on. However, it takes several minutes for the image to stabilize as it tends to black out at first. Have you experienced anything like this and if so is there anything I can do about it?

 

[Kadano]

So I managed to get everything working on Monday (I ended up having to blindly navigate my Wii to change the display setting to 480p) and I have to say it's one of the best set ups I've ever played on. However, it takes several minutes for the image to stabilize as it tends to black out at first. Have you experienced anything like this and if so is there anything I can do about it?

I haven't had that problem at all, so I don't know how to solve it. Buying an Extron RGB interface (164xi, for example – about 30$) and running the VGA cable through it before connecting it to the CRT monitor might fix it though.

 

[SpiderMad]

So what are the tried and true methods of WEARING OUT a control stick to make it looser [from Left/Right]? Should I just make aggressive rotations with it for a half hour each day?

No one knows?

 

[Cyberscum]

Hello, I'm glad to see technical details around here about the hardware we so appreciate, I have a couple of questions however! Lately I've been having trouble dash dancing (especially with Captain Falcon it seems? Might be just my imagination) where I tap the stick but instead of going into my dash animation I only turn around, this causes me to do a stupid dance in place every now and then instead of dash dancing which isn't very optimal, do you have any idea what might cause this? I believe I have T3 stick boxes and I'm also currently using a PS3 stick since the GC one was worn out. I've taken my controller apart plenty of times for cleaning and swapping parts but I don't feel comfortable with taking the stick box out/apart yet so I'd rather not do that if it's not needed. This is also my only controller.

Next, I'd like to ask if there is currently a good place to order OEM replacement sticks? I saw they're out of stock in the GCS gaming store. Also what do you think is the best way to keep the stick from wearing and tearing? I'd like to keep it in stock condition for as long as possible.

And for future reference.. If I ever were to replace the mechanical parts (whether on this controller or another) which are the highest quality components that are currently known of and are compatible? I know I can't put T1/2 stick boxes on mine if I have T3 but I like to think about replacing everything with the highest quality components to make sure everything is top notch :]

 

[NTRN]

I have a technical question in a slightly different vein. It is about controller polling. The GC polls the controller twice per 1/60s, but the values are read by melee once a frame. Is there any way to tell which poll the game reads? The polls are usually relatively close together compared to the 16ms period, which I think can work to my advantage to differentiate the two compared to if they were evenly spaced at every 1/120s. I think it would be the second one, but it is hard to know.

To add to this, I have access to the controller data, but not to the game data which I could use to cross reference, but it would still be difficult. I suppose I could send from the controller a single poll input like jump, then the rest of the polls send no input, then check to see if melee jumped or not, then check which poll (first or second) that occurred.

I'm a little concerned about the potential for more than 2 polls per frame messing up my algorithms though.

 

[ShrieK1295]

So to decrease resistance on the a/b/x/y buttons, you make holes in the silver circle that surround the black circle?

 

[Kadano]

So to decrease resistance on the a/b/x/y buttons, you make holes in the silver circle that surround the black circle?

Yes, although I'd rather call the rubber pad's color grey than silver.

 

[hanson933]

Windows Defender on Win10 says that stream overlay template zip is a Trojan btw
https://www.microsoft.com/security/portal/threat/encyclopedia/entry.aspx?name=Trojan:Win32/Rundas!plock&threatid=2147711527&enterprise=0

 

[[BROF]]

This might sounds like a crazy question, but has anybody ever dwelt into adding cushioning to the Triggers? As in the outside part of it, the concave section where the fingers rest.
I'm trying different things to make controllers a bit more ergonomic, and just wanted to hear if anybody here has done something like that.

 

[Kadano]

Controller snapback
Newer, more extensive guide can be found here: https://sites.google.com/view/kadanosnapback/home

What is controller snapback?
Controller snapback is caused mechanically by the mass of the control stick knob, which in flick inputs accelerates during the return from full extension to neutral, causing it to travel into over-extension to the opposite side until being slowed down by the spring cushioning its momentum. In most controllers, this makes the stick vibrate like a pendulum for a few times until resting still in the middle.

The amount of physical over-extension varies slightly from controller to controller, but most of the time, it is great enough so that it reaches the level where directional inputs beyond the dead zone are read for about 3 milliseconds.

In-game, this is a problem because inputs beyond the dead zone in Melee store the aerial neutral-B orientation flag for 20 frames. So what happens is that, while airborne and facing right, you press left strongly and let go of the stick ("flick input"), intending to perform a neutral-B to the left side with this input and moves like Falco's laser or Sheik’s needles. However, from the ~3 ms the stick stays in the right-hand input zone, there is about an 18% chance that the controller will poll for inputs just during that time, shooting the laser towards the right instead of the left, and making almost every fifth directional airborne neutral-B go off into the wrong way.

In Smash 4, a similar problem exists, where empty pivots (smash turns, they call them "perfect pivots" for some reason) allow turning around again immediately, even with tilt turn levels (about >=0.2875 in Melee input levels).

Here is some footage showing the snapback effect:

1. Video of flick inputs on a new (Smash edition) controller at 420 fps, showing the over-extension:

2. Oscilloscope curve of these flick inputs on the same controller (x-axis = time, y-axis = x-input with low = left and high = right), blue lines are the dead zone thresholds (edited in):

The snapback over-extension to the right (shown as upwards curve) exceeds dead zone levels for about 5 ms, so this controller has about a 31% chance of flick inputs causing the wrong orientation.
For reference, the 0.2875 levels are reached at about 1.32-1.48 Volt (left) and 1.88-2.04 Volt (right). There is quite a bit of noise in the signal levels, so these aren't the most accurate numbers, unfortunately. However, when testing controllers for snapback both in-game and with the oscilloscope, so far controllers that maxed out at 2.00 and 2.04 V did not reach snapback levels beyond the dead zone in-game, so I expect that the outer thresholds of the dead zone are the "real" thresholds. In other words, only parts of the yellow snapback curve that exceed the outer ends of the blue lines are capable to be read as directional inputs.

4. Oscilloscope curve of the same controller with the stick knob removed, proving that it's its mass that causes the mechanical over-extension:

With older GCC editions, this problem exists too. Here is a used purple controller with a type 1 stickbox:

Same controller with the knob removed, to show that also with type 1 stickboxes, the cause of the problem is the mass of the stickbox:

Spoiler: (Show)

Old black, used, with type 2:

Possible solutions
There are three different ways to eliminate the snapback problem.

1. Favorable potentiometer oddity / "good PODE" (Potentiometer Oddity Degradation Effect)
On some controllers, so far mostly JP White 2008 edition, with use the potentiometers can deteriorate in a strange way that makes the stick readings less linear, causing them to be read in slower steps.
From an engineer perspective, this is rather unwanted, but from a Melee perspective, this is very helpful.
On controllers that have this oddity, control stick input levels will not react fully to sudden changes. So while the stick knob still mechanically over-extends, the potentiometer readings will not follow suite, perhaps from improper trace contact.

Example flick input oscilloscope curve:

Another effect that this oddity has is that with smash turns, the "buffer time" during which the glitched oscilloscope skips input changes corresponds with the time usually spent in the tilt turn range, allowing one to skip that entirely, with no or drastically reduced risk of accidental tilt turns being read.

The exact change within the potentiometer that causes this oddity to appear is likely to be wear on the resistive carbon track that causes uneven electricity flow. The exact effect this degradation has on a controller's input response varies quite a lot, from very helpful to unusable drifting and is known to be very volatile: snapping off the potentiometer from the stickbox usually removes the effect, and sometimes even just opening the controller and cleaning it reduced smash turn success rate significantly, suggesting that PODE was removed by doing so.

Here is the same controller as in the last picture with flick inputs towards the right, in the upper half without the stickbox touched, and in the lower half with the stickbox taken out and put back in:

Until it is known how this oddity can be caused on purpose consistently, it cannot be the go-to method to deal with snapback.

2. Silicone grease lubrication of the stickbox (section kept for archival purposes, I strongly discourage to do this!)

Spoiler: Tutorial / description archived as reference, please read the important update below first!

By smearing a large amount of thick (NLGI class 2) silicone grease, the inner friction of the grease viscosity will provide enough deceleration to the control stick to prevent it from reaching over-extension levels.
Here are oscilloscope screenshots of a Smash edition controller that has been lubricated with silicone grease:

Before lubrication, type 3 white (T3W):

Spoiler

Before lubrication, type 3 black (T3B):

Spoiler

After first lubrication (T3B):

Spoiler

After second lubrication:

Spoiler

After third lubrication (very large amounts of grease used):

Spoiler

End result:

To apply the silicone grease, use a small toothpick or Q-tip with the cotton removed to pick up the grease, push the stickbox shaft in one cardinal direction and smear the grease into the hole on the opposite side opened by that.

The more silicone grease you use, the less overshooting there is, but the louder the sound of the grease will be when moved from neutral position, and the slower the stick will return to neutral, which will probably feel weird / unusual to you at first. Since the return travel from full extension to neutral position still takes less than one frame (16 ms) at most, it's still fast to not cause any real in-game problems, though.
For reference, a new unlubed type 3 stickbox takes about 8 ms to return from full extension to neutral position. At this point, however, it will only start the snapback vibration. Counting from the initial letting go, it takes 23 ms until the control stick is completely at rest.

Here is my tutorial on lubricating the stickbox until the snapback is gone entirely (quite long, 22 minutes – better take the time to watch all of it), includes additional explanations on the physical process, lubricating potentiometers to eliminate the squeaky noise that some of them have, and inserting wooden spacers to prevent the lubed stickboxes from having too little friction against the stick knob:

Which silicone grease to use:
I've been using this one here by Liqui-Moly. Other brands are fine too, as long as it's silicone grease and NLGI class 2.

Credits:
I had lubricated my stickboxes with rather thick silicone grease for years and hadn't had any snapback problems, but I failed to make the connection that there was a causal relationship between the greasing and the absence of snapback. Credit for that observation goes to @Gentlefox who mentioned it to me about two months ago.


Important update (three weeks after the video above)
Applying too much grease can cause drifting problems. I had used my main controller with moderate grease for two months before starting to sell this mod and write this post, so I thought that there were no bad side-effects, however eventually for at least two controllers I sold and for at least two people who did this mod on their own, bad drifting issues occurred. These are usually very rare, so unfortunately they definitely come from the lubrication. I recommend not doing this greasing mod, unless you want to take the risk of your control stick drifting heavily. Even then, you should only apply it sparingly, to decrease the chance of drifting happening, and combine that with a small capacitor.
If you have potentiometers with drifting issues, the best way to deal with that is to replace them with new ones (resolder). Here is a video on how you can alleviate drifting on a greased controller without replacing the potentiometers (might need to be repeated after certain amounts of time): https://www.youtube.com/watch?v=fMUXu9aoib0

3. 100nF - 1µF capacitor between pins 2 and 3 of the stickbox (updated 2020-11)
This mod retains the original sound and physical feeling of the stick, with the loud rattling vibration sound after flick inputs and the absence of the mushy silicone grease sound.
Snapback is filtered by a capacitor that attaches to the signal and ground pins of the potentiometer. Small changes in signal voltage get 'eaten up' by the capacitor.

The nF value (capacitance) of the capacitor is crucial – too low and there will still be snapback present, just at a lower rate; too high and there will be a delay on all stick inputs in the respective axis that's higher than necessary.

There is no specific value that's ideal for every controller. Here are some broad range recommendations that usually work:

• 470-680 nF: typical range needed on new or lightly used controllers that still have a stiff, snappy stick
• 100-330 nF: controllers that have been used for 9 months or more and only have snapback occasionally when not using a capacitor usually need capacitance in this range to fully eliminate snapback beyond the dead zone.
• 33-100 nF: controllers with beneficial PODE (good dashback in vanilla) that still have snapback rarely (1 in 40 times or less) are best fitted with capacitance in this range, which eliminates just enough snapback to not make it exceed the dead zone.
• 700 nF to 1 uF: controllers with either a very loose stickbox spring (low push resistance) or a taller / heavier stick knob (Nunchuk) and relatively new potenitometers usually need capacitance in this range, as those properties increase their mechanical snapback that we need to filter out electrically.
• > 1 uF: very rarely needed, with combinations like loose stickbox spring (type 3 black from C-stick / Smalley L1-L2 / type 3 white used heavily for several years) and a heavy stick knob (Nunchuk or similar). Using a metal stick knob will usually also necessitate more than 1 uF (I advise against using them in the first place for this reason).

Using overly high capacitance will make empty pivots less consistent, since very short smash inputs are sometimes "eaten" by the capacitor. Ideally, use an oscilloscope or SmashScope to see the signal curves as above and adjust capacitance until the flick input return path looks similar to the upper screenshot from before.
Here's an example for what return paths look like when the capacitance is too high for the current potentiometer performance:

Here is the same configuration with a pivot input:

The pink line shows the pivot / dash threshold, so in the sample above the effective smash turn range time (STRT) is extended from the manually input ~20 ms to ~27 ms, decreasing pivot success rate from ~80% to ~65%.
More details and explanations: https://sites.google.com/view/kadanosnapback/home


Also, controllers with this capacitor mod will not calibrate correctly on console (in emulation, they calibrate just fine). You'll need to either reset by holding X+Y+Start every time after plugging the controller in to prevent drifting, or use a no-recalibration circuit that delays the connection of the capacitance to the signal pin until after the controller has been calibrated. (Several controller modders make / sell / install no-recalibration modules, including myself.)

Installation types

1. SMD capacitor: direct soldering installation that can only be undone / changed with a soldering iron. Not recommended unless you always have a soldering iron at hand, as you can't adjust quickly to account for changes in potentiometer performance.
2. Modular jumper wire installation: capacitance is adjusted by sliding radial ceramic capacitors into a pair of female jumper wires soldered to the potentiometer terminals. Viewed from the back side, one of them goes to ground (rightmost pin) while the other one goes to signal (middle pin). As long as you use ceramic rather than electrolytic capacitors, polarity doesn't matter. In fact, it's even safe to solder to signal and 3.3V, snapback will be eliminated just as well.
3. Switch module installation: capacitances are combined additively by flipping switches. More effort to build / install, but much more hassle-free to adjust to ideal level. Also, you have much more fine control over the exact capacitance since the additive nature of the individual capacitors gives you very fine granular control. For example, with jumper wires you have to settle for 470 or 680 nF with no options in-between, while with a switch module that has 47, 100, 220 and 330 nF you can go for 430, 477, 550, 597, 650 or 697 nF. This allows better fine-tuning towards having just barely sufficient capacitance to eliminate snapback.

Again, more details can be found here: https://sites.google.com/view/kadanosnapback/home

 

[ssknight7]

sick

i'll donate $500 if you can find a way to reliably make controllers smash turn without adding electrical components :D

 

[Kadano]

sick

i'll donate $500 if you can find a way to reliably make controllers smash turn without adding electrical components :D

500$ is a hard number, but "reliably" is not.
We'd need to define it in how many milliseconds are allowed for the control stick to be moved from the dead zone to the gate's rim to qualify as a proper input, and in how many consecutive smash turns must be achieved in order to be called reliable.

From my experience, I'd say 6 ms and 40 consecutive inputs are realistic (and currently probably only met by the few controllers that have a strongly pronounced potentiometer phenomenon).

 

[Logos]

can you make another GIF showing the results after applying the silicone?

 

[Pauer]

can you make another GIF showing the results after applying the silicone?

This video shows that, though not in slow motion.

 

[Kadano]

can you make another GIF showing the results after applying the silicone?

I recorded one now, I had to wait for the sun to rise again (50 Hz flickering in electrical lighting (yes, even LED lightbulbs) makes high-speed video unwatchable). It's also embedded in the main post now.

 

[ssknight7]

500$ is a hard number, but "reliably" is not.
We'd need to define it in how many milliseconds are allowed for the control stick to be moved from the dead zone to the gate's rim to qualify as a proper input, and in how many consecutive smash turns must be achieved in order to be called reliable.

From my experience, I'd say 6 ms and 40 consecutive inputs are realistic (and currently probably only met by the few controllers that have a strongly pronounced potentiometer phenomenon).

what would you estimate the normal ms range is for controllers that do not have the phenomenon?

I don't have hard numbers I only have my experience but I trust you know what i want to accomplish and that the 6 / 40 figures will suffice.

also when I say you can't add electrical components I mean you can't add any new components like capacitors etc.

it's fine if you were to find a differently rated potentiometer and substituted the new one for the old one just no new components

lastly the solution must make empty pivots unreasonably difficult so if I'm able pivot grab or shortened dash dance grab fox at 21% with no di 10 times in a row on a normal controller I should be able to do it 7 times in a row on a modded controller

also I must be able to mod the controller and verify myself with your instructions.

legit offer and if you can do this I'll gladly pay 500. would help the community at large and would save me so much money on controllers

 

[Logos]

I recorded one now, I had to wait for the sun to rise again (50 Hz flickering in electrical lighting (yes, even LED lightbulbs) makes high-speed video unwatchable). It's also embedded in the main post now.

Wow, that is a immense difference. I may have to look into doing this so I stop getting turn around lasers.