Jump to content

Anyone Tried Saitek X-65F With Mwo


21 replies to this topic

#1 Midnight Camel

    Member

  • PipPip
  • Civil Servant
  • Civil Servant
  • 41 posts

Posted 06 October 2013 - 06:13 AM

Does anyone use the X-65 for MWO or has anyone at least tried it? It seems like the force sensing feature might make this stick more usable than other joysticks in MWO since it is supposed to be more precise. Any insight is appreciated.

#2 Loc Nar

    Member

  • PipPipPipPipPipPipPipPip
  • 1,132 posts

Posted 06 October 2013 - 09:22 AM

Quote

It seems like the force sensing feature might make this stick more usable than other joysticks in MWO since it is supposed to be more precise.


Not viable at all to use for reticule aim, and highly undesirable even to use in your left hand for movement (which is all a regular joystick is good for in this game). Force sensing is literally the exact opposite direction you want your controls to operate for MWO. The X-65 is also a terrible example of a force sensing stick even for flight applications too. The concept of using force sensing instead of mechanical position sensing does have more room for precision by human operates, but Saitek missed the boat when they made this stick and they skipped reading the rest of the memo.

Yes F-16's and most modern fly by wire jet aircraft technically use FSS, but it is a hybrid system where the stick actually does mechanically deflect a slight amount, and that little bit of tactile feedback makes all the difference in the world of FSS working or not and you will not find a single example of a static stick in even one airplane. There are only 2 proper FSS sticks out there, one is a conversion for the TM Cougar and costs about $600 in addition to the base cost of the stick and the other is for the Warthog and costs $800 on top of the base cost of the stick, making the totals of them $1-1.2k total The stick Seitek makes is a toy by comparison, and ignores the lessons DoD learned about FSS stystems when this was being pioneered in the 70's.

While those exotic sticks would be great to use for *some flight sims, they are still both boat anchors/liabilities in MWO (even worse than the unmodified versions of the Cougar and Warthog), since reticule aim in MWO is based around pure positional manipulation, called zero-order control. Unfortunately any joystick you buy is engineered around vector manipulation, which is called first-order control, and these two different concepts for aiming breed mutually exclusive mechanical arrangements to aid in their respective tasks.

For this reason, there is no suitable first-order control joystick that is viable for MWO, or any other zero-order aiming tasks, but again FSS moves you away from the direction you need to mutate towards for a joystick to work better here since it takes away the only property working in your advatage -tactile feedback from mechanical displacement.

In short, reticule aim is really hard with a normal spring-loaded isotonic joystick, even the mighty Warthog, but nearly impossible with an isometric one like an X-65f. Even spending $1.2k on a proper FSS stick gets you no closer to being viable, and in fact less so than even a regular stick due to the nature of zero-order control.

If you ever have any doubts or wonders about using any new controller for MWO, just set it up as a mouse emulator and try browsing with it for a full day in whatever browser you use. That is the exact same thing you will be up against in a match, since it's the same exact task. If the terms zero-order control or first-order control are unfamiliar, I recommend at least skimming the article I wrote on the subject, called Controls Demystified(?)

As far as working solutions, there are some but they require some fabrication work because no one currently manufactures s joystick suitable for use in a zero-order control application. I refuse to use kbm myself, so I have constructed my own stick just for MWO and it works quite well. As such, it is very different than a regular joystick, and all the mechanical properties that make it excel at MWO/zero-order aiming tasks (no spring centering or detentes, large range of motion in pitch/yaw, highly damped friction jooints, etc) work just as much against it the moment it is pressed into service as a first-order controller and is just as much a fish out of water in any flight sim as a regular joystick is in MWO. The X-65f comes in somewhere even lower on the food chain than even a $5 Logitech Attack3 in this game...

*FSS sticks only work good for modern jets and space sims, but WWII or WWII aircraft need the longest throw mechanical displacement sticks you can find.

#3 Midnight Camel

    Member

  • PipPip
  • Civil Servant
  • Civil Servant
  • 41 posts

Posted 06 October 2013 - 11:24 AM

I've read your other posts and understand your argument regarding input types but I'm actually just interested in getting feedback from someone who's tried the X-65 with MWO.

#4 XphR

    Member

  • PipPipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 3,513 posts
  • LocationTVM-Iceless Fold Space Observatory Entertaining cats...

Posted 06 October 2013 - 12:24 PM

View PostBelberry Johns, on 06 October 2013 - 11:24 AM, said:

I've read your other posts and understand your argument regarding input types but I'm actually just interested in getting feedback from someone who's tried the X-65 with MWO.


If that feedback never comes, his is the most reliable fall back informative you will receive.

#5 Alex Warden

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,659 posts
  • Location...straying in the Inner Sphere...

Posted 06 October 2013 - 01:20 PM

i watched people with expensive systems, also x65 and what not... listen to LocNar... they ALL are at best slow, clumsy and inaccurate compared to a mouse. yes it is sad, i´d love to use a stick for immersion purposes myself, but it´s simply not worth the bad performance

Edited by Alex Warden, 06 October 2013 - 01:21 PM.


#6 Moriquendi86

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 97 posts
  • LocationWarsaw

Posted 06 October 2013 - 01:47 PM

View PostLoc Nar, on 06 October 2013 - 09:22 AM, said:


Not viable at all to use for reticule aim, and highly undesirable even to use in your left hand for movement (which is all a regular joystick is good for in this game). Force sensing is literally the exact opposite direction you want your controls to operate for MWO. The X-65 is also a terrible example of a force sensing stick even for flight applications too. The concept of using force sensing instead of mechanical position sensing does have more room for precision by human operates, but Saitek missed the boat when they made this stick and they skipped reading the rest of the memo.

Yes F-16's and most modern fly by wire jet aircraft technically use FSS, but it is a hybrid system where the stick actually does mechanically deflect a slight amount, and that little bit of tactile feedback makes all the difference in the world of FSS working or not and you will not find a single example of a static stick in even one airplane. There are only 2 proper FSS sticks out there, one is a conversion for the TM Cougar and costs about $600 in addition to the base cost of the stick and the other is for the Warthog and costs $800 on top of the base cost of the stick, making the totals of them $1-1.2k total The stick Seitek makes is a toy by comparison, and ignores the lessons DoD learned about FSS stystems when this was being pioneered in the 70's.

While those exotic sticks would be great to use for *some flight sims, they are still both boat anchors/liabilities in MWO (even worse than the unmodified versions of the Cougar and Warthog), since reticule aim in MWO is based around pure positional manipulation, called zero-order control. Unfortunately any joystick you buy is engineered around vector manipulation, which is called first-order control, and these two different concepts for aiming breed mutually exclusive mechanical arrangements to aid in their respective tasks.

For this reason, there is no suitable first-order control joystick that is viable for MWO, or any other zero-order aiming tasks, but again FSS moves you away from the direction you need to mutate towards for a joystick to work better here since it takes away the only property working in your advatage -tactile feedback from mechanical displacement.

In short, reticule aim is really hard with a normal spring-loaded isotonic joystick, even the mighty Warthog, but nearly impossible with an isometric one like an X-65f. Even spending $1.2k on a proper FSS stick gets you no closer to being viable, and in fact less so than even a regular stick due to the nature of zero-order control.

If you ever have any doubts or wonders about using any new controller for MWO, just set it up as a mouse emulator and try browsing with it for a full day in whatever browser you use. That is the exact same thing you will be up against in a match, since it's the same exact task. If the terms zero-order control or first-order control are unfamiliar, I recommend at least skimming the article I wrote on the subject, called Controls Demystified(?)

As far as working solutions, there are some but they require some fabrication work because no one currently manufactures s joystick suitable for use in a zero-order control application. I refuse to use kbm myself, so I have constructed my own stick just for MWO and it works quite well. As such, it is very different than a regular joystick, and all the mechanical properties that make it excel at MWO/zero-order aiming tasks (no spring centering or detentes, large range of motion in pitch/yaw, highly damped friction jooints, etc) work just as much against it the moment it is pressed into service as a first-order controller and is just as much a fish out of water in any flight sim as a regular joystick is in MWO. The X-65f comes in somewhere even lower on the food chain than even a $5 Logitech Attack3 in this game...

*FSS sticks only work good for modern jets and space sims, but WWII or WWII aircraft need the longest throw mechanical displacement sticks you can find.


Wow thats really insane knowledge you have, I need to go trough other article. Thanks for sharing this.

#7 XphR

    Member

  • PipPipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 3,513 posts
  • LocationTVM-Iceless Fold Space Observatory Entertaining cats...

Posted 06 October 2013 - 09:23 PM

View PostMoriquendi86, on 06 October 2013 - 01:47 PM, said:


Wow thats really insane knowledge you have, I need to go trough other article. Thanks for sharing this.

Its worth the read.

#8 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 09 October 2013 - 09:40 AM

I disagree. If the stick does not move, that would mean that it is the only stick in existence that is suited to emulating a mouse.

By not moving, it is effectively a relative controller.

The more expensive Warthog and Cougar equivalents would be less suited because they move.

Not saying that it would make stick aiming any good, just that it is more suitable for how MWO currently processes aiming.

Once MWO supports absolute stick inputs for aiming though, it would be significantly worse.

#9 Foust

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 394 posts
  • LocationKentucky

Posted 09 October 2013 - 03:35 PM

You know, that is a excellent point.

Does the software for the stick allow for mouse emulation and axis scaling?

#10 Loc Nar

    Member

  • PipPipPipPipPipPipPipPip
  • 1,132 posts

Posted 10 October 2013 - 09:18 AM

Quote

I disagree. If the stick does not move, that would mean that it is the only stick in existence that is suited to emulating a mouse.


Quote

You know, that is a excellent point.


Lol... I'm pretty sure you guys are putting 2 and 2 together and coming up with 22. An isometric joystick is not closer to a mouse, precisely because it does not move. Them both being relative devices is besides the point. If you played MWO or browsed with a mouse running in absolute mode would be nearly identical to one reporting relative other than using fixed coordinates, and would still be zero-order.

A critical distinction in common shared by mice and regular joysticks in fact is that they are displacement based controls that *do move, regardless of whether they are relative or absolute. Displacement is a necessary ingredient for zero-order control, and it turns out to be desirable trait for repeatable precision in first-order as well. An isometric joystick is by definition incapable of displacement, making it mechanically incapable of replicating anything that resembles a mouse does, and is strictly limited to vector commands.

This ironic/funny, because the x65 is the one and only known example of a joystick that is utterly incapable of emulating anything a mouse does and the premise of it being closer is simply put, not correct.

#11 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 10 October 2013 - 09:40 AM

Sorry Loc, I have a lot of respect for the work you have done, but you are just plain wrong.

The reason a normal stick has trouble emulating a mouse:
If you are at 50% deflection left, to move to 50% deflection right you need to move the stick 50% of the whole travel distance.
THIS is the reason that absolute input methods have trouble emulating relative input methods.

This is why you want your absolute mapping for sticks - because with a regular relative mapping, if you are twisted 50% left and you want to twist right, you need to move the stick back past center to the right, which creates an input lag. With a pressure stick, you do not have to do this - the millisecond you release pressure to the left, you stop twisting left.

With a stick that does not move, to move from 50% left to 50% right is a travel distance of zero. You simply need to change the pressure you are exerting.
It is the same as a mouse - with a mouse, you move it left by exerting pressure to the left, and the mouse moves. As soon as you stop exerting pressure, it stops and input is at zero. With a stick that does not move, this is exactly the same.

Again, another reason why your "First Order, Zero Order" classification of input is not ideal for games. The terms "Relative" and "Absolute" describe what is going on much better.

Think of a n i p p l e (Sorry, forum thinks that is a dirty word...) mouse controller you find on laptops. It is exactly like one of these sticks.

Here you go: An isometric joystick used as a mouse, as fitted to IBM / Lenovo thinkpads:

Edited by evilC, 10 October 2013 - 09:44 AM.


#12 Hellzero

    Member

  • PipPipPipPipPip
  • The Phoenix
  • The Phoenix
  • 118 posts

Posted 10 October 2013 - 09:50 AM

As a player who played through MW 2/3/4 with a HOTAS, and after I tried the X-65F on MWO, I have to agree with Loc Nar 100%.
- It just doesn't work in this setting, the very least. Torso twisting & turning is simply absurd on a Heavier Assault, impossible on a light/fast/agile mech.
***

- When you try to turn the mech left, you somehow almost automatically twist the arms at the same time, because of the reading sensors taking your input differently. After I gave it more rounds, I was able to distinguish the way of controlling the mech, but it reacted by turning/twisting even to "trigger presses", as I tried to maneuver in the middle of a firefight.
- If I needed to start dancing with the opponent's mech in a brawl, the controlling became near-impossible/handicapped badly.

It turned in to a nightmare and it wrecked the nerves in my hand badly after an extended period of time, as I tried to force myself in to getting the hang of it(it demands more from your nerves of your arms, than muscles).

The problem is the complete lack of feedback from the joystick, due to differentiating arm/torso/leg movement.
- Pedals could have helped, but just my hint: Get anything except something that relies on sensors. It might work better on a flight/space sim, but not in this world of 100-ton ballerinas.
***

- This is just my personal opinion of the matter, after personal experience with the system in MWO. Even after I got the hang of it, the effort was more painful than fully-immersive. The traditional joystick would be better, in my opinion.


Edit:
As last words, try thinking about pushing as hard as you can from your hand, to pull an Atlas to the left, not being sure if it's being turned "100%", and you push it as hard as you can, and you realize it's still "86/92%", regardless of the settings in the joystick. It means you will be sweating/cursing of the maneuverability of your mech, because you can't be sure you are actually using it to the maximum potential of the Mech's maneuverability, at any time.

- Adding up the factor that your arms are turned at almost the same feedback, as is your leg turn; so you're pushing the system almost with your two hands to get the shot aligned(you need to experience it, to get a full picture of its frustrating effect after the long run).

- This is my biggest issue with the system, especially in MWO, where close-combat can become intense. It's more do-able in fire-support role, but still I'd take a normal joystick set-up over X-65F, any day.

Edited by Hellzero, 10 October 2013 - 09:59 AM.


#13 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 10 October 2013 - 10:02 AM

I would imagine that in the scenario you described, the stick was set to way too sensitive - as indicated by the fact that you were triggering a turn by pressing a button.

I can fully understand what you are saying though, but there is simply no way that any stick with a "throw" distance can compete as a relative controller with one that has no throw. The only way would be a system like loc nar has of mapping an absolute input to a relative one - ie when at 50% deflection left, as soon as you move the stick to 49% left, you have generated 1% right movement.

Think about it. Imagine a vertical line, with a mouse cursor on it. Move the mouse 1cm left, then stop moving it any more left - the mouse is 1cm to the left of the line. Move the mouse 1cm right and it is on the line. Move another 1cm right and it is right of the line. No input lag.

Now repeat with a regular stick.
Stick 1cm left, stop moving - crosshair is still moving left! Stick 1cm right (ie move back to center), crosshair is still moving left until you finish this maneuver! Stick 1cm right - nope, crosshair is still nowhere near the line. Stick 1cm more right, you may have moved near the line, but you are still not even centered, let alone started moving the crosshair right yet. Massive input lag!

Now repeat with the X65.
Stick 1cm left (or equivalent pressure), stop moving - stick stops near instantly. Stick 1cm right - crosshair will be roughly on the line. Stick 1cm more right, will be to the right of the line. Little to no input lag.

Edited by evilC, 10 October 2013 - 10:04 AM.


#14 Hellzero

    Member

  • PipPipPipPipPip
  • The Phoenix
  • The Phoenix
  • 118 posts

Posted 10 October 2013 - 10:13 AM

View PostevilC, on 10 October 2013 - 10:02 AM, said:

I would imagine that in the scenario you described, the stick was set to way too sensitive - as indicated by the fact that you were triggering a turn by pressing a button.

I have to agree with your post, and you are right in this. My concern is more that the main issue would be of the leg behavior/upper torso behavior, in the same stick. This was not a problem in the traditional joystick, but... I hope this scenario will be easier to understand:

You tilt the joystick vertically right to turn the upper torso right, and you "turn the joystick horizontally left to turn the legs left", at the same time battling with not moving your target-reticulation up/down at the same time of doing this, to keep the aim on target who you might be running away from. So you're turning left now, while trying to mow your aim/torso/arms right, to the target, without any form of feedback. to where/how much you are doing all this, and in what direction.

- I... was thinking of buying that X-65F, instead of CH Joystick HOTAS kit, but... after this scenario, that was the last deal.
- So 'til something like what Loc Nar is telling about, being "more available to everyone, for a "reasonable price"", I'll hold on to that CH HOTAS kit.

But thank you for your feedback, nevertheless; the X-65F comes with many progressive sensitivity settings for it, and almost the lightest was what I found the most "bearable to use in heavier mechs". In mediums I scaled it up on purpose.

Edited by Hellzero, 10 October 2013 - 10:15 AM.


#15 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 10 October 2013 - 10:50 AM

Oh well if you had torso twist and leg movement on one stick, that is a recipe for disaster. You cannot even accurately use two axes on one stick for throttle and leg turning (You get "bleedthrough") let alone four axes on one stick.

#16 Foust

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 394 posts
  • LocationKentucky

Posted 10 October 2013 - 03:03 PM

A question I have is how the x-65f, in a mouse emulation mode, treats the release of pressure on on a movement?

Would it keep the cursor in the location where pressure was released, or would it return the cursor to center?

#17 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 10 October 2013 - 05:44 PM

Even in "stick mode", it should hit center immediately.

There is no "center" for a mouse, that is the point of relative input. "Center" for a mouse is "Not moving".
That is why I am saying that for this stick, there is essentially no difference.

Note that I am NOT saying that it will be ideal for MWO - just better than a regular stick at imitating relative movement.
It is NOT CAPABLE of being as good as a regular stick with a mod such as Loc Nar's that maps an "absolute" stick to a relative range calculated to mimic an absolute range, which is what his mod is.

#18 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 10 October 2013 - 06:28 PM

View PostevilC, on 10 October 2013 - 05:44 PM, said:

It is NOT CAPABLE of being as good as a regular stick with a mod such as Loc Nar's that maps an "absolute" stick to a relative range calculated to mimic an absolute range, which is what his mod is.

Hmm, on thinking about it, this is not 100% correct. Being a stick, it is entirely as capable of doing it, but because you have little haptic feedback (Only pressure, not position in 3D space), it would probably not be as usable.

#19 Loc Nar

    Member

  • PipPipPipPipPipPipPipPip
  • 1,132 posts

Posted 11 October 2013 - 09:22 PM

Quote

Sorry Loc, I have a lot of respect for the work you have done, but you are just plain wrong.


Evil, I respect your projects and efforts as well too so I'm glad we have established each others mutual good will in these endeavors. I believe my position to be on solid ground here so feel the need to provide further explanation since it's been challenged. Sorry for the walls of text but I don't know how to do it in less words. Ironically I used to use absolute and relative to explain controls too (it's all I knew), which got me through casual conversations easy enough but occasionally an engineer or someone else that knew there stuff would step in and defy my ability to use those words alone to fully explain things short of making it up. Armed with new and better search terms each time while looking for adequate answers to these questions led me down a path of discovery, and forced me to revise my semi-hollow explanations and understanding and relearn some things I thought I knew.

Attempts at describing controls by using absolute or relative are inadequate as they only describe single parts of a process, device, or programming, and even after defining which part of the chain it applies to might not even describe the achieved control scheme and might still require further defining of what type of relative (relative to what?) and where it's applied (relative to where?) etc and it eventually becomes a list of variables that tie together into specific coherent systems, and those systems have been given names. Control-order is the naming convention used here, and it describes a resulting control scheme which may be comprised of many different combinations of relative or absolute components, so those words are only good for descriptors for the path taken to achieve a control scheme, and numerous paths can lead to the same resulting scheme.

Quote

The reason a normal stick has trouble emulating a mouse:
If you are at 50% deflection left, to move to 50% deflection right you need to move the stick 50% of the whole travel distance.
THIS is the reason that absolute input methods have trouble emulating relative input methods


I'm not certain how you come to this conclusion, but absolute devices do not have trouble emulating a mouse, at least not for *finite relative applications like MWO. I say finite because they are incapable of emulating infinite ones like Hawken where turning left/right is coupled to reticule aim. In that case you can't use a device running absolute inputs because you need to be able to move 360+ and by definition an absolute device has a limited range of motion. Infinite travel is one of the key defining criteria of relative input, so anything based on it is not viable for my stick to emulate... However with MWO, the reticule aim takes place in a fixed window with defined edges, so a stick emulating mouse running on absolute has no problem at all outside of varying degrees of mechanical impediments it may or may not have, and all mechanical issues have mechanical solutions, though some may not be worth the effort.

The only artifact that makes any tangible difference of MWO as to whether relative mouse inputs are being used instead of the alternative -a proper native analog absolute option, is potentially losing synchronization of the center if you go beyond the edge of what is mapped on screen. This is the same thing that happens if you go to far with your mouse too, but with a mouse you just pick it up and set it back down to fix it. With my stick I have to manually center it then push 'c' or use the china hat to jog it back. If MWO had absolute analog inputs, I would rarely if ever need to 'c' center regardless of my gain settings (you can still cork it out with opposing leg inputs though). The difference between absolute mouse emulation and proper analog absolute is much smaller than the difference between hardware that can actually utilize these inputs vs over the counter options though, and the answer would still be the mod/build-a-better-stick (gimbal) route. Am I missing something here?


Quote

This is why you want your absolute mapping for sticks - because with a regular relative mapping, if you are twisted 50% left and you want to twist right, you need to move the stick back past center to the right, which creates an input lag. With a pressure stick, you do not have to do this - the millisecond you release pressure to the left, you stop twisting left.


Hehe, no. First off, with a mouse, trackball, my stick, generally any other zero-order controller, inputs can be applied faster than even the lightest/twitchiest mech can fully react to them in-game since the weakest link in this chain is not the input device but rather the virtual servos of the mech. Once your mech is moving that fast however your eyes and the ability to process objects on the screen while it's whipping around at a blur becomes the weak link beyond that (well, mine do) and becomes even less relevant. The real reason absolute mapping is desirable is because that is the way to translate stick's position directly into target position, achieving a control scheme as close to mouse zero-order as possible. The other option, relative inputs, would translate stick position into velocity/direction and precludes the possibility of positional tasks as the resulting control scheme and is no longer zero-order control. This... this is why you want absolute mapping.

Speed ...that's pretty much only a [concern] of people using velocity commands and a normal joystick of sorts since one is then forced to choose between speed vs accuracy when determining gain settings (usually achieving neither), but is near meaningless for positional based inputs, at least for MWO where things only move at the speed of mech! :blink: If input speed mattered though, the x65 would win but there's little cause to celebrate this victory of a few milliseconds that can't be spent on anything. It's not all roses in the garden though! My stick and any that use a similar approach impose a different limitation than speed vs accuracy as a byproduct of using absolute positioning; and that is range-of-motion vs accuracy. Also, because it's an absolute device with a finite arc and due to technical details of relative reporting nature of mouse emulation this is where undesirable edge behavior comes into play as mentioned above. Not enough gain and my mech doesn't move far enough. Too much gain and it's hard to control and may begin losing center. It's a built in limitation of using any absolute positioning device vs a relative one for zero-order, but for me it's much preferable and an acceptable limitation.

While in confession mode, there is one mechanical issue with my stick that makes it more challenging than a mouse and it's one that most sicks suffer to varying degrees (Warthog grips have very strong trigger springs and greatly magnify the problem), and it falls more into ergonomics, but has mechanical solutions that can solve it. The location of the pitch axis pivot makes the stick prone to unintentional inputs when squeezing the trigger and to a lesser degree one of the hats. Ideally the pitch pivot point would be mid-stick (currently ~50mm below the palm rest), to allow the wrist to rotate neutrally throughout the range of pitch. The means the stick would either need either an actual pivot point extended up to there (bulky/awkward erector set), or would need to ride on appropriate radius arced track bearing (doable). I have been thinking long and hard about making a new gimbal to address this shortcoming, which would also be partially mitigated if I ever incorporated proper viscous dampers as opposed to greased plastic rubs, and in the meantime will likely put a weaker spring in the trigger... Yaw feels wonderful as is actually and is very easy/natural to fluidly manipulate, but with pitch req my entire arm needs to move forward/backward more than I would like because it affects accuracy and I'm tired of spamming it from stiff trigger events. With an arced bearing my arm would essentially stay put and only my wrist would move, and this 2.0 version of the gimbal would be an even further departure from the norm, moving it a step closer towards ideal.

Quote

With a stick that does not move, to move from 50% left to 50% right is a travel distance of zero. You simply need to change the pressure you are exerting.


It is the same as a mouse - with a mouse, you move it left by exerting pressure to the left, and the mouse moves. As soon as you stop exerting pressure, it stops and input is at zero. With a stick that does not move, this is exactly the same.


It is very dissimilar to a mouse, which physical displacement proportional to the amount needed (dependent on dpi/gain setting/mech/etc), which is then in turn converted into positional change. Yes it takes pressure to move a mouse, but it takes pressure to move any control. That is a silly comparison and you are on the losing end of a semantic argument by suggesting the pressure your hand exerts to achieve displacement is the same as an input that is actually measuring pressure. A mouse actually measures that displacement, but a pressure sensor measures pressure. These are not 'exactly the same'.

Quote

Again, another reason why your "First Order, Zero Order" classification of input is not ideal for games. The terms "Relative" and "Absolute" describe what is going on much better.


You're getting into an increasingly difficult position to defend. Understanding control order and identifying what input type a task is based in is vital for determining what controls will work and which ones won't without resorting to luck and trial and error, and reduces the need to rely on intuition and spurious arguments. It doesn't matter whether it's games, browsing, or flying the space shuttle, and it's not 'my' classification, this is the classification system used in the discipline of engineering called human factors. Those guys making the joysticks and relevant parts of games have training in this area, and this is what they call it and how they frame it to talk shop.

This is from the Human Factors in Simple and Complex Systems by Robert W. Proctor, Trisha Van Zandt. (university textbook) On page 406, start reading from Control system order onward for about a page... (won't let me copypasta it)

http://books.google....ometric&f=false

That short and concise summary pretty much outlines what I hammer at a lot and may seem like my own personal pet theory, but comes from established information in this field of engineering. If controls were a pyramid, control-order is the brick at the top. Words like absolute and relative are bricks underneath that support it and as such can be stacked in many ways to do so without changing the top brick.

Quote

Think of a n i p p l e (Sorry, forum thinks that is a dirty word...) mouse controller you find on laptops. It is exactly like one of these sticks.


Heh... true, they are technically like the x65, at least in how the sensor is processed and that they're static isometrics; so my statement of it being the only/lonely example was a little loosely worded I suppose. What was meant was it's the only example of a full sized stick/HOTAS that is static and pressure based. There are other pressure based sticks (Cougar and Warthog FSS mods that deflect +/- 6mm or so) and they are reported to work fantastically in the right situations, spoiling lucky simheads that can afford the luxury and use them in appropriate vehicles. My issue is with static controls to being used for precision vehicular management; not with pressure sensors themselves, and not with people's nippple mice or other casual-use pointing devices. Citing their existence as evidence of their value in a competitive game setting would be curious however, although it's true the x-65 is not alone in the universe. Being compared to one is not exactly flattering for a joystick intended for flight or other precision tasks in a competitive environment.

Quote

A question I have is how the x-65f, in a mouse emulation mode, treats the release of pressure on on a movement?

Would it keep the cursor in the location where pressure was released, or would it return the cursor to center?


Now that is an interesting notion. Ignoring the fact that it would be extremely difficult to control (unless you spend most of your time with your reticule in the center of the screen!), this I think would technically qualify it as zero-order control because cursor displacement is directly tied to sensor pressure in that scenario, but not being displacement based makes this an interesting grey area that may force me to refine my zero-order definition accordingly. Note- this wouldn't make it viable for MWO by any means since it's the opposite of mechanically optimized for this task, but an interesting aside I'm curious to learn more about as there has to be other applications that would benefit from this, possibly immensely. Of course this might involve serious coding that is not in place, but I see no technical reasons that would prevent it from working. This nullifies one of the points I've used against the x65 earlier (that is is only capable of velocity commands) so have to retract that point since this would definitely be a positional command.

Interestingly this further highlights the importance of calling control schemes by their proper names - by their control order. Saying the input device and application are both relative input based does not describe the achieved control scheme and would be equally applicable to vector based controls (first-order) as it would positional based controls (zero-order) until it is specifically laid out for each step of the process of the device, the software, and the application by using a lot of words to describe each of their attributes and contributions that define them from each other. Alternatively it can just be identified by it's order of control.

#20 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 12 October 2013 - 08:28 AM

Well it's gonna take me a while to fully digest that wall-o-text, but I have given it a first pass. It seems that both of us have some valid points.

Regarding using an X-65 in MWO - I did say that it would not be ideal, but it would be better than a regular stick when used without mouse emulation. There would be no coding needed at all to make it work to the best of it's abilities, because it lacks the recenter time of a regular stick, so when MWO interprets the stick's input as a relative shift on an absolute scale, it just works.

Obviously if you were to try to map it to an absolute axis, it would be rubbish. To aim FPS-style with a stick, you need an unsprung input (very difficult holding crosshair in same place and places strains on arms) which the X-65 is exactly not.

Quote

I'm not certain how you come to this conclusion, but absolute devices do not have trouble emulating a mouse

A mouse can switch from generating a left output to a right output almost instantly. The best an absolute device can do is go from "left" to "less left" in the same time. When using it for eg torso twist, this manifests itself as input lag, because the physical distance in 3D space that your hand has to travel to change from "twisting left" to "twisting right" is massively more than with a relative input device. Because the throw of the X-65 is basically zero, it is effectively a relative input device - as long as you interpret it's absolute output as a relative input (Which MWO does), it will behave much more naturally.

Edited by evilC, 12 October 2013 - 08:28 AM.






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users