New Product: Ultimarc UltraStik 360 Analog/Digital Mappable Joystick

[Tiger-Heli]

Also, bear in mind this joystick can be connected to a standard Ipac and thus be used with programs or operating systems that don't support analogue or even USB.

I haven't studied that element of it in detail.  Can it be hooked up to both the I-PAC and USB together and select which way to output on a command-line basis?

Otherwise, if I am principally planning to have it output to an I-PAC, I don't see why I would buy this over an $10 Super or Comp.

Although it would be interesting to test this stick connected to an I-PAC and a Happ Super connected to an I-PAC and see how they compare.

That would be another case where the mapping software could come in useful, b/c the I-PAC would have no other way to know what it means if the stick is "mostly up but a little bit right".

[Grasshopper]

I don't know enough about consoles to know which ones would support any USB analog stick, but even if it did, I think it is asking a lot to expect people to plug the stick into their PC, enter a particular mapping into it, and then plug it into their console, if it would remember it (which I think it will).

Actually I think this would work quite well. But the joystick would have to have the capability of storing several different maps simultaneously. The user would also need to have a means of selecting the correct map for the game being played perhaps by pressing a switch. Maybe that facility could be built into a future firmware upgrade (Andy are you reading this? ).

[u_rebelscum]

The following quote is edited: I put in the bold, and removed Randy's original bold and italics to saw my point.

...
That being said, I have some real concerns as to whether any external mapping system, outside of simple dead-zone control, will have an effect on a true analog stick where, theoretically, actuation is equal in all directions of movement.  Chopping up an already excellent range of actuation into coarse blocks and then forcing the controller to guess the intentions of the player is, well, puzzling at best.  Especially when that information is already available with full accuracy.

To explain a little further, there are very simple maps that, theoretically, are the best maps available for any analog joystick that offers equal actuation of the sensors in every direction, as is usually their goal.  Coincidentally, these maps are already being used in any game where its programmer has even the slightest clue.  They are are as follows:

The problem is mame is not well written to translate analog to digital*.  Mame looks at each axis separately; the value of the other axis is not looked at during the analog 2 digital translation.  This makes Mame's look more like the following:

Depending on "a2d_deadzone" and other settings, it could look more like the first or the second (I left in the right's lines for closer comparison), or imagine a larger deadzone instead of the shown smaller.  Notice the lines are exactly horizontal or vertical.  As mentioned above, mame looks at each axis individually, resulting the such lines.  Notice how a smaller a2d_deadzone increases the diagonals and narrows the cardinals.  Also check out how bad mame's analog to 4-way convertion is (although to 4-way diagonal is actually pretty good with small a2d_deadzone values).

Which is where UltraStik360 comes in.  It is doing a better job than mame, period.  However...

Not that I don't agree a more flexible mapping than a 9x9 on the UltraStik would be wrong. 

Setting the angles, increasing or decreasing the deadzone size, changing the shape of the deadzone, or doing crazy things like making curvy lines or difference angles between left/upleft than for right/upright would be nice.  Are they needed?  Do they make that much of a difference than a 9x9 grid?  Are they possible to be assigned from the computer on hardware fast enough to do the calculations while cheap enough for retail?

In talking to Andy a little while ago, he mentioned that using angles instead of a grid would be much more computation intensive than possible for the hardware used.
Andy is that still true?

And Randy, TigerH, everyone else, is there some other way than angles to get setable diagonal/cardinal/deadzoneareas on a joystick?  I can think of times (tron) when an angle from the center won't be as good as, say, 60 degree triangles with the point at edge of the circular deadzone between the cardinal directions.

*(FWIW, the best analog conversion in mame IMO is the current 720.  It uses arctan to calculate the angle.  And as much as Aaron says I helped with that, my code makes up about 2% of what he wrote and put in mame. )


To sum up:
While well written PC games don't need the analog to digital mapping in UltraStik360, mame needs it badly. 
The mapping also "puts the controls in the users' hands" instead of having to use whatever the well (or poorly) written game mapping has. 
There are finer ways to remap than with a grid, but they aren't easy to do in practice, especially while being changeable & fast on joystick hardware.
Mame could use an analog to digital go over (but needs to be neater than I code, and not conflict with analog to digital "button" convertions to be accepted).

-edit- reworded second to last summary point to make more sense.

[Minwah]

If you ask me this stick should be superb and worth every penny.  My only concern is whether the feel of the stick is good enough to play all the 8-way games (fighters in particular) sufficiently well - we shall soon know about that.

For people not too bothered about 4-way games the mapping will be sufficient.  For those 4-way purists the 4/8 way restrictor should do the trick, just like the J-stik (what I've been using for ages).

[alank2]

I haven't studied that element of it in detail.  Can it be hooked up to both the I-PAC and USB together and select which way to output on a command-line basis?

It will always report to USB, but you can configure MAME to listen or not listen to USB by enabling or disabling the joystick option.  Depending on what map you have loaded it may or may not report keystrokes to an IPAC.

This is probably how I will use it if I get one.  I will have it hooked up to both the IPAC and USB.  On games where an analog stick makes sense I plan to enable the joystick option in MAME and download an analog map to it.  With an analog map loaded it won't report any keystrokes to the IPAC.  Then, for non analog games, I will download the appropriate map and disable the joystick option in MAME so the game only looks for keyboard input from the IPAC [from the Ultrastik].

I wonder how the throw will be with and without the circular restrictor plate installed as well as the return spring strength.

Thanks,

Alan

[Tiger-Heli]

Great points, Urebel !!!

The problem is mame is not well written to translate analog to digital*.  Mame looks at each axis separately; the value of the other axis is not looked at during the analog 2 digital translation.  This makes Mame's look more like the following:

Depending on "a2d_deadzone" and other settings, it could look more like the first or the second (I left in the right's lines for closer comparison), or imagine a larger deadzone instead of the shown smaller.

So if I am looking at this chart correctly, MAME is using a (very crude) 8-way map for the A-D conversion, even in 4-way games, correct?

While well written PC games don't need the analog to digital mapping in UltraStik360, mame needs it badly.

Okay, I did not realize this.

 
The mapping also "puts the controls in the users' hands" instead of having to use whatever the well (or poorly) written game mapping has. 

Granted, and that can be useful.

There are finer ways to remap than with a grid, but they aren't easy to do in practice, especially while being changeable & fast on joystick hardware.

Actually, my point was that there was no point downgrading to a 9x9 grid if MAME is already using the data from a 256x256 grid.  Expecting a user to input 65536 entries into the map would be silly, so this is a good compromise.

Mame could use an analog to digital go over (but needs to be neater than I code, and not conflict with analog to digital "button" convertions to be accepted).

So it would appear.

[AndyWarne]

Some great discussion here!
I don't actually have a great deal to add apart from a few minor points:
Firstly, just to confirm that when maps are downloaded (via USB) they do have an effect on the 4 digital directions sent to an I-PAC if one is connected. This is where mapping really comes into it's own. Any map which contains all analog cells will appear to the I-PAC as all centered, hence nothing will be sent, but other squares will be sent as mapped.
Randy might well be correct in that in 8-way mode there is not a great deal of value in mapping when connected only via USB. But that is missing the point because the ability to change from 4 to 8 way is the essence. It may well be that many users never create any custom maps at all, just flip from the supplied 4-way to the supplied 8-way.
I have no plans to enable a mechanical switch for changing maps, as this is better done programmatically. It can already be done in Mamewah for example on a "per game" basis.
On the subject of restrictors I have a sneaking suspicion that I might make the circular restrictor a standard fitment. I am waiting feedback really, on whether the standard travel is too long for non-analog games. The problem of course is anything to do with joysticks is very subjective, which is why a design intention was to make it as configurable as possible.
I chose a cell-type map rather than angular for simplicity and less real-time processing being required on the stick itself (eg large look-up tables).
Interestingly, the number of cells in the matrix and also the spacing between cells are programatically adjustable, although the current software does not support this in the GUI. I might consider adding this support if there is interest.
Another point of interest: using a restrictor plate on the stick does not compromise the analog resolution. The internals of the stick use a resolution of 16384 X 16384 so reducing this to, say, half the travel and changing the scale factor (which is done by the mapping software) still maintains a huge resolution advantage over what Windows sees.
Andy

[RandyT]

Randy might well be correct in that in 8-way mode there is not a great deal of value in mapping when connected only via USB. But that is missing the point because the ability to change from 4 to 8 way is the essence. It may well be that many users never create any custom maps at all, just flip from the supplied 4-way to the supplied 8-way.

If this is is what I was perceived to have stated, then I offer my apologies.  I must not have made my point very well.  It was more along the lines of "With an analog joystick that has an optimal 8-way configuration and range of actuation that so closely resembles that of a standard 8-way (which cannot benefit from digital restriction attempts, e.g. 8-way to 4-way), how is it that the addition of maps to the same physical mechanics can make a difference in that conversion?  Especially if the gaming software already divides the 256x256 map properly.

Regardless, Urebel pointed out some rather interesting (and slightly disturbing) limitations to the MAME code which provides a partial answer to that question.

Based on what he has written, mapping must be provided with an analog stick in order to be properly used with MAME, as analog sticks are very poorly handled by MAMEs 8 and 4-way conversion code.   So mapping will certainly correct for that particular issue. The question I still have is related to the rest of the point I had attempted to make, and that is; now that the translation is more along the lines of what is optimal (the analog maps I drew), how does that specifically make it a better 4-way than a switch based 8-way using a 4-way actuator (which provides an actuation map very similar to the optimal 4-way map,  yet historically do not work very well?)

Ultimately, this is the aspect I am most interested in understanding and hearing about from the early adopters.

RandyT

[Tiger-Heli]

Randy might well be correct in that in 8-way mode there is not a great deal of value in mapping when connected only via USB.

I believe that Randy's point was there with a standard PC game, there was not a great deal of value in mapping in either 8-way or 4-way mode when connected via USB.  But that was before Robin clarified how MAME does (or rather doesn't) translate analog to digital.

But that is missing the point because the ability to change from 4 to 8 way is the essence. It may well be that many users never create any custom maps at all, just flip from the supplied 4-way to the supplied 8-way.

Agreed.

I have no plans to enable a mechanical switch for changing maps, as this is better done programmatically. It can already be done in Mamewah for example on a "per game" basis.

Agreed, except for console games, not Windows OS, unless you are working on these solutions.

On the subject of restrictors I have a sneaking suspicion that I might make the circular restrictor a standard fitment. I am waiting feedback really, on whether the standard travel is too long for non-analog games. The problem of course is anything to do with joysticks is very subjective, which is why a design intention was to make it as configurable as possible.

I would be interested to hear feedback on the throw and spring return rate of these.

I chose a cell-type map rather than angular for simplicity and less real-time processing being required on the stick itself (eg large look-up tables).

Makes sense.

Interestingly, the number of cells in the matrix and also the spacing between cells are programatically adjustable, although the current software does not support this in the GUI. I might consider adding this support if there is interest.

I think Kremmit had a very good point that an oddxodd map works better for 8-way mappings and an evenxeven map would work better for 4-way mappings.

Another point of interest: using a restrictor plate on the stick does not compromise the analog resolution. The internals of the stick use a resolution of 16384 X 16384 so reducing this to, say, half the travel and changing the scale factor (which is done by the mapping software) still maintains a huge resolution advantage over what Windows sees.

You could have left a 16384x16384 matrix in the GUI and then people could REALLY create their own custom maps!!! 

On a serious note, that implies it might well be possible to have a long throw version of the stick for analog games and a restricted throw version for digital games, which is excellent news also. 

[mark shaker]

Andy:

16384 X 16384 Resolution?

What technology allows you to capture that resolution?

It doesn't sound like pots: They are so jumpy that a resolution that high would be meaning less.

It doesn't sound like optical: Going from a "traditional" 7X7 grid to 16384 X 16384 is beyond belief.

Is it a Hall-Effect stick?

   - Mark

[Tiger-Heli]

I understand, just a little - no comprende, it's a riddle . . .  Seriously, I followed about half of Randy's post, but I'll waddle through it anyway!  Randy, please clarify if I missed anything.

Randy might well be correct in that in 8-way mode there is not a great deal of value in mapping when connected only via USB. But that is missing the point because the ability to change from 4 to 8 way is the essence. It may well be that many users never create any custom maps at all, just flip from the supplied 4-way to the supplied 8-way.

If this is is what I was perceived to have stated, then I offer my apologies.  I must not have made my point very well.  It was more along the lines of "With an analog joystick that has an optimal 8-way configuration and range of actuation that so closely resembles that of a standard 8-way (which cannot benefit from digital restriction attempts, e.g. 8-way to 4-way), how is it that the addition of maps to the same physical mechanics can make a difference in that conversion?  Especially if the gaming software already divides the 256x256 map properly.

Which if the gaming software is MAME we now now that it does not do.

Regardless, Urebel pointed out some rather interesting (and slightly disturbing) limitations to the MAME code which provides a partial answer to that question. Based on what he has written, mapping must be provided with an analog stick in order to be properly used with MAME, as analog sticks are very poorly handled by MAMEs 8 and 4-way conversion code.   So mapping will certainly correct for that particular issue.

That's the way I understand it.

The question I still have is related to the rest of the point I had attempted to make, and that is; now that the translation is more along the lines of what is optimal (the analog maps I drew), how does that specifically make it a better 4-way than a switch based 8-way using a 4-way actuator (which provides an actuation map very similar to the optimal 4-way map,  yet historically do not work very well?)

Okay, you are talking about something like a Happ Super with the 4-way actuator, and or the Perfect 360 with the 4-way actuator, copared to the Ultra-Stik using a 4-way map, correct?

Most of what I have heard was that these do work fairly well, except for the circular motion and the PITA of removing the E-clip to switch from 8-way to 4-way.

To answer your question, (from a guess, at least) - I can tell you how it might work better:

If you minimize the diagonals on Andy's matrix, you end up with 9 dead zone points and 9 diagonal points which would be set to Sticky mode (or 9 out of 63 active points) or 15% of the active positions.

We know that say a Happ Super with a 4-way actuator DOES actually allow the diagonals to register - unless you tweak the leaf actuators, etc., even though it shouldn't.  So if the Happ stick allows the diagonal to register more than 15% of the time, Andy's stick will outperform it.  If the Happ stick allows the diagonals to register less than 15% of the time, it will outperform Andy's, but Andy's could do better by setting up a higher resolution map.

At least that's how I see it.

Ultimately, this is the aspect I am most interested in understanding and hearing about from the early adopters.

I'm most interested in the physical feel (throw and return) of the sticks.

[trcroyle]

Is it a Hall-Effect stick?

Mine just showed up and they do, indeed, seem to be Hall Effect. If you look at the photo on the first page you can see the flat magnet at the bottom of the shaft, and if you squint and use your imagination that black blob behind the usb port is where the sensor is. Or, here:



It definately has some throw, though I have a couple of sticks with leaf switches that come close. Here it is mounted through a 1 1/8" hole:



Just played a quick game of Ms. Pac Man. Nice and smooth. Silent. Took a a couple deaths to get used to the feel. I definately think I'll get some of the stronger springs since I have a tendancy to let go of the stick in some games, the center is a just a tiny bit "bouncy" with the stock spring. I like it well enough to get two more.

[Timoe]

that balltop looks akmost metallic?  and from the photo it looks large.  Ithought the balltop would be the exact same as the J-stick.

Am I just not seeing these photos right or something?

[trcroyle]

that balltop looks akmost metallic?  and from the photo it looks large.  Ithought the balltop would be the exact same as the J-stick.

Am I just not seeing these photos right or something?

No, hard plastic. Maybe 1.5" across? I'd have to dig out my caliper to be sure. Not too much bigger than that hole. Normal sized. It doesn'ts eem big in the hand.

[Kremmit]

U_Rebel:  Thanks!  I was not aware of that limitation in MAME.  The diagrams really helped.  Definately proves custom mapping is worth having on  these sticks, for things other than just Tron.

Is it a Hall-Effect stick?

...and what's it use, hall-effect sensors?

Heh, great minds think alike!

Interestingly, the number of cells in the matrix and also the spacing between cells are programatically adjustable, although the current software does not support this in the GUI. I might consider adding this support if there is interest.

The only solution I can see is to either a) use an odd # of rows for some grid mappings, and an even number for others; or b) use a much higher resolution grid...

I think Kremmit had a very good point that an oddxodd map works better for 8-way mappings and an evenxeven map would work better for 4-way mappings.

There's two votes, right there!  Just allowing an 10X10 map for the 4-way modes is enough to eliminate the dead/sticky zones, but adding higher resolution will also allow the "jaggedness" of the dividing line to be smoothed out as well.  Either fix is probably good enough, if one is more difficult than the other.

Another point of interest: using a restrictor plate on the stick does not compromise the analog resolution. The internals of the stick use a resolution of 16384 X 16384 so reducing this to, say, half the travel and changing the scale factor (which is done by the mapping software) still maintains a huge resolution advantage over what Windows sees.

If that's true, then:

On the subject of restrictors I have a sneaking suspicion that I might make the circular restrictor a standard fitment.

Is starting to sound like a good idea, pending user testing.

Sounds like these are going to be very nice!

[Tiger-Heli]

I have one thing that is bothering me about the Ultra-Stick.  I don’t understand the advantage of the I-PAC mode.  I understand the stick can be simultaneously connected to the USB and I-PAC and you can specify which mode to use, I just don’t see the point . . .  Here are my conclusions and then I will get to the reasons I decided this:

USB Mode:  This may be the most versatile stick yet.  I’ll wait for the head-to-head to make that pronouncement, but it definitely is looking good at this point.

I-PAC Mode:  You lose the analog capability, you gain one less level of computation (more on that below), and (the only real advantage I can see) - you gain the ability to use the stick in PS/2 mode if you lack USB capability.  (And you gain the ability to use the stick with games that only support keyboard and not analog stick inputs).

Now the logic - let’s look at the type of games you could play:

Analog games - (Star Wars, Sinistar).  USB mode works fine.  In I-PAC mode, the stick will either be centered or hard over, making the games unplayable, AFAICT.

Digital games (PacMan, 1942, Donkey Kong) - Robin has already shown that MAME does a poor job of resolving this games with an analog input, and I have discussed how MAME sometimes guesses wrong in the 4-way games.  Now, in USB mode, the stick’s firmware is going to resolve almost any position into one of the 4-cardinal modes (4-way mode) or one of the 8 cardinal modes (8-way mode) and pass this to MAME as an analog signal which MAME has to resolve.  Robin has shown that MAME does not resolve these modes well, but if in 4-way mode, the firmware is going to tell MAME the stick is either centered, full up, full right, full left, or full down.  Similarly, in 8-way mode, the stick will be either centered, full N, full NE, full E, full SE, full S, full SW, full W, or full NW.  No matter how bad MAME’s mappings are, MAME can resolve this.  In I-PAC mode, the sticks firmware is going to make the same calculation, but then it is going to pass MAME a direct keypress value through the I-PAC.  This eliminates the clunky A-D calculations in MAME but the end result is the same, and I can’t see the MAME calculations affecting gameplay to any significant extent . . .

Am I missing anything???

[Tiger-Heli]

I also came up with Revision 2, 3, and 4 to the stick last night.  These are all based on varying the amount of travel of the stick.

As has been said before, the problem with a do-all solution like this is analog games want a lot of stick travel and digital games do not.

First let’s look at some of the more common sticks and where they fall on the stick travel graph (from what others have said):

T-Stick, Mag-Stick, Omni-Stick, Prodigy, Suzo 500, MCA joysticks:  VERY short throw.  Excellent for 4-way games (Pac-Man, etc.) Fair to good for 8-way games if you get used to the feel.  I can’t imagine using one for analog games as the difference in movement between centered, a little bit right, medium right, almost full right, and full right would be almost imperceptible.

J-Stick, Sanwa, Ultra Stick () Reunion Stick, Wico Leaf Stick:  I am ASSuMing the Ultra-Stick has a throw in this range, since it seems to be similar to a Sanwa in many ways.  These sticks have a short-to-moderate throw.  4-way games are okay on them, but 8-ways should really shine.  Also popular for Japanese fighters.  This amount of travel might be acceptable for some analog games, but it will be at the short end of tolerable.

Happ Super, Happ Comp, Happ P360:  Moderate-to-long throw.  The travel on these sticks is less than ideal for 4-way games, although many BYOAC’ers like them (probably b/c they are some of the most common USA sticks).  They are good and pretty much standard for 8-way games and fighters.  They should also give decent play on analog games, although possibly still a little short overall.

GGG, Happ, Williams 49-way:  Long throw.  This stick actually has an extremely long throw, so much so that Markrvp and others have recommended mounted the stick so that the mounting hole actually slightly physically restricts it to reduce the amount of travel available.  I would think it would be bad for most digital games and especially for 4-way games, but most reviewers and actual users do not rate it this way.  It is really designed for analog games and should do well with them.

True Analog stick (Space Harriers (?), Space Ace (?), Afterburner, someone help me here):  VERY long throw.  Hard to imagine playing 4-way games on these.  Even 8-ways could be tough.  True analog games will shine.

REVISION 2:  Look at the list above.  Some of the purists will insist on a long throw for analog games and a short throw for digital games (and a 4-way restrictor).  But many users will be able to find a compromise that will work well for them for all games.  Personally, I think I would shoot for something between a Sanwa and a Happ Super.  But this target will likely be different for each user.  So Revision 2 (possibly increases the un-restricted throw) and introduces a fixed restrictor that can be set to the amount of travel that the user desires.

REVISION 3:  Now we go after the purists.  The resistor is modified to have two adjustable stop points, a short throw digital stop point, and a long throw analog stop point.  Revision 3a introduces a method where the position of the restrictor can be changed from beneath the stick.  Revision 3b introduces a method where the position of the restrictor can be changed from above the panel, similar to the Prodigy or the T-Stick or Mag-Stick.

REVISION 4:  This is the ultimate.  We build on the concept of revision 3, but now we introduce a worm gear servo or solenoid (preferably powered directly off the USB bus) which allows us to change the restrictor position via software command.  So, using a utility like SirPoonga’s Set49Way - we can just start an analog or digital game, and the software will automatically configure the joystick correctly for us and off we go . . .

[destructor]

Wow, restrictor plate is more than required here for 4/8 way games. Throw is more than large without restrictors. Otherwise for analog mode restrictors aren't required.

Offtopic:

T-Stick, Mag-Stick, Omni-Stick, Prodigy, Suzo 500, MCA joysticks:  VERY short throw.  Excellent for 4-way games (Pac-Man, etc.)

T-Stik for 4 way games? Are you sure? What with precision?

[trcroyle]

Wow, restrictor plate is more than required here for 4/8 way games. Throw is more than large without restrictors. Otherwise for analog mode restrictors aren't required.

The perspective on the photo may be a little misleading. The top of the stick travels maybe half an inch in any direction. It's not short by any means, but it also didn't feel excessive to me. I had more issues with releasing the stick and getting a little bounce (which I think the heavier springs would fix) than I did the throw.

[SirPoonga]

So, using a utility like SirPoonga

[Minwah]

The perspective on the photo may be a little misleading. The top of the stick travels maybe half an inch in any direction. It's not short by any means, but it also didn't feel excessive to me. I had more issues with releasing the stick and getting a little bounce (which I think the heavier springs would fix) than I did the throw.

Looks OK to me...don't forget in digital games the direction will be activated long before the extent of the sticks travel.

Edit: I was backing up your statement, not disagreeing btw

[SirPoonga]

ndy, got a question about the command line interface and havinf a frontend set this up.

The only thing a frontend can do is set the UGC file, right?
So one has to make these files first.  Is there a batch way of making all the nessessary files based on defaults instead of going through each type (2way, 4way, 8way, etc) and creating a UGC file?

Edit: I just realized the answer to my question is going ot be no.  The person has to configure as needed.  Nevermind

[AndyWarne]

Tiger-Heli:
About the I-PAC mode, the reason was simply versatility. When the stick is first plugged in, it always sends un-mapped analog data via USB and mapped directional data to the I-PAC, if connected.

As an aside, it is not possible to return to this mode once a map has been downloaded via USB, because once downloaded, mapped data is sent to both USB and the I-PAC, that's something to be added as a choice in a later software version.

Anyway I digress. The way it is designed, you can use it if you wish as an alternative to a P360. Just drops in. This opens up other application areas. Don't forget that there are other uses for joysticks than MAME. The cost/benefit of adding this capability works favorably as the cost is very low. Incidentally that I/O connector would have to be fitted on the PCB regardless since that is how the chip is originally programmed.

I have already had one user mention that he will always use the stick in the "as shipped" setup, with analog via USB and directions via I-PAC (which he already has) because he is not particularly bothered about setting up maps for all his games. He will never need to download any maps at all, as analog games will use analog via USB and digital will use directional via the I-PAC.

Your wish list makes a lot of sense. Some considerable challenges there though!

[Grasshopper]

I have no plans to enable a mechanical switch for changing maps, as this is better done programmatically. It can already be done in Mamewah for example on a "per game" basis.
Andy

I'm very sorry to hear this. I can understand your reluctance to make significant changes to your product so soon after launch but I hope that in time you'll reconsider.

It's kind of ironic that RandyT's GPWiz49 interface originally had the opposite problem. You could change maps by pressing buttons but not though software. Several people here (including myself) argued that switching through software would be a far better solution and Randy eventually implemented this.

However the the point is, the two approaches are not mutually exclusive. This is one situation where you can have your cake and eat it.

Clearly switching maps through software is preferable for most users, myself included. But as things currently stand, that option is only available if you're running a later version of Windows on a USB capable PC. What about people using older PCs, or people who want to use MSDOS, Linux, Apple Macs, or even consoles (with converters)?

OK admittedly the last map used is retained in flash memory but all this means for non-Windows, non-USB users is that they are able to use the UltraStik as a very expensive 4 or 8 way non-switchable digital joystick.

If you're only going to offer map switching through Windows then it might have been simpler and cheaper to have designed a very simple analogue joystick with no sophisticated on-board electronics, and then just written a Windows driver to handle the 4 or 8 way digital joystick emulation.

The ability to connect the UltraStik to an Ipac (or other encoder?) is potentially an excellent idea as it means this joystick can be fully utilised (apart obviously from pure analogue mode) by people who either don't use Windows or don't have a USB port. However, without the ability to switch maps through button presses (or some other non-software means), you're effectively spoiling the ship for a ha

[SlikStik Christian]

For the love of god, you people are just never happy. The man releases a revolutionary new first of its kind analog joystick that covers just about every base known to man when it comes to joysticks and all it gets is tons of criticism and I wish had

[Grasshopper]

Dude, you need to chill out! You've got a very strange attitude towards running a business.

We're not "criticising" we're just offering suggestions for future improvement. That doesn't in any way mean we're not impressed with what Andy has come up with so far. The UltraStik looks very good indeed on paper (and no one here has said otherwise) but no product is perfect and there will always be room for improvement.

Most businesses welcome feedback, both good and bad, from their customers. It's called market research. For every potential customer who makes the effort to express his opinion on a product there are dozens who simply decide not to buy the product if it doesn't fit their needs.

[SlikStik Christian]

Feedback? I think feedback it great as well as needed to some respect. I wouldn

[rugby1]

Can you get anywhere without typing c:/

[thorus]

I want to get in on the icon war to you and a  to you.

 

Back on topic I think that grasshopper is right to be saying what he is saying and that Andy can read it decide if its feasible and then do it or not.  I wouldn't get agressive with grasshopper maybe point out, kindly, that he dreaming but I only saw grasshopper dreaming out loud there. 

Oh BTW the joystick looks awsome can't wait to see reviews of it.

my $.02

[Tiger-Heli]

I will consciously choose not to respond to Christian's comments, other than to point out that while DOS may be close to dead, Linux is not, and it would be nice to offer a Linux solution for the software other than running it through WINE (which I am not sure you can start and exit from) but probably.

Is that required? - absolutely not - many of Andy's competitors don't have linux programming software and they still make sales.  OTOH, I am sure there are cases where the I-PAC gathers sales that it might otherwise lose due to Linux support and adding it to the stick would be a nice complement.

Off the  on to more important things:

About the I-PAC mode, the reason was simply versatility. When the stick is first plugged in, it always sends un-mapped analog data via USB and mapped directional data to the I-PAC, if connected.

Mapped directional data using which map (more below)?

As an aside, it is not possible to return to this mode once a map has been downloaded via USB, because once downloaded, mapped data is sent to both USB and the I-PAC, that's something to be added as a choice in a later software version.

I didn't follow this either - are you just saying there is no way to get back to having it send pure analog via USB and mapped data to the I-PAC once I download a map, but I could still get back to pure analog (with maybe the I-PAC being confused until I download a new map.

Surely you are not saying that in USB mode, once I download a map I am stuck in mapped mode and can't get back to pure analog?  (That would be a pretty serious drawback if you could play Star Wars and then PacMan, but then you had to unplug the stick to get back to Star Wars???)

Anyway I digress. The way it is designed, you can use it if you wish as an alternative to a P360. Just drops in. This opens up other application areas. Don't forget that there are other uses for joysticks than MAME. The cost/benefit of adding this capability works favorably as the cost is very low. Incidentally that I/O connector would have to be fitted on the PCB regardless since that is how the chip is originally programmed.

This part I also understand.  I mentioned apps that can't use an analog joystick.  Not sure if you mean as a replacement for an 8-way only (or 4-way only) P360 - more on that below.

I have already had one user mention that he will always use the stick in the "as shipped" setup, with analog via USB and directions via I-PAC (which he already has) because he is not particularly bothered about setting up maps for all his games. He will never need to download any maps at all, as analog games will use analog via USB and digital will use directional via the I-PAC.

This is what I was getting at above.  What map is loaded by default?  I am ASSuMEing it will be 8-way.  If said user only plays 8-way and analog games, then this should be fine.  Otherwise I see big problems with the 4-way games - well not big but the same as using an 8-way digital stick to play them, no benefit.

The stick or the I-PAC can't know that you loaded up Pac-Man so it should switch to the 4-way map, can it?  Pretty good trick if it does!  (And if it is still in 8-way mode, the stick is going to send two inputs to MAME whenever you press a diagonal, and your game will mess up.)

OTOH, I do realize and wish to point out that it should be pretty simple to use something like a modified version of set49mode to set the correct map and download it from launch from your front-end.

Your wish list makes a lot of sense. Some considerable challenges there though!

I never said it would be easy.  (Wouldn't be much fun if it was!!!)

[sirwoogie]

Feedback? I think feedback it great as well as needed to some respect. I wouldn

[Chris]

[I'd like to interject for a moment. I don't think you're reading the constructive criticism as such. Personally from an engineering standpoint, hearing people question design considerations as well as making wishlist is a big bonus.

If I may take the unpopular side for a moment... I am a big believer in constructive criticism.  I live for it on my jukebox.  But is his product announcement thread really the place for it?  When the thread first went up, my first thought is "I've got to get me some of these!"  Now after reading three pages of criticism and suggestions, it seems to be more prudent for me to sit back and wait for v2.0.

Just my two cents.

--Chris

[Kremmit]

If I may take the unpopular side for a moment... I am a big believer in constructive criticism.  I live for it on my jukebox.  But is his product announcement thread really the place for it?  When the thread first went up, my first thought is "I've got to get me some of these!"  Now after reading three pages of criticism and suggestions, it seems to be more prudent for me to sit back and wait for v2.0.

Actually, if you go back in time, you'll see that the discussion about potential problems and/or upgrades started in a different thread, but were MOVED INTO this thread by a moderator who, I'm sure, meant well. 

As for sitting back and waiting for 2.0- the suggestions that seem likely to be implemented all center on software upgrades, which ought to be available via download.  Unless you think Tiger-Heli's pie-in-the-sky motor-driven restrictor plate has a snowball's chance in hell- I doubt even TH thinks that one's likely.  I don't see much point in waiting.  Honestly, the only reason I didn't order on day one is that I'm short on cash.  Soon as payday hits, I'm ordering.

Christian-  This is a board full of the most inventive, curious, do-it-yourself, upgrade-happy, hardware-modding, hyper-picky hardware-geeks in the entire retrogaming hobby.  Furthermore, of the people on this board, some of the most knowledgeable, intelligent members we've got are the ones driving this thread.  What do you expect, nothing but "Dood, kewl!" and "Awesome, those r leet!"?  Try Myspace if that's your cup of tea.  Interesting new products are going to get some real discussion around here.  Sure, some of the suggestions will be over-the-top, too expensive, impractical, or just plain impossible, but there may be some good stuff in there, too.  In the end, it'll be up to Andy to decide which, IF ANY of our suggestions he takes.  I've already stated, in more than one post in this thread, that I'll buy even if he changes NOTHING.  I'm sure plenty of others will, too.  The point of it all is we're HELPING Andy produce the best product he possibly can.  He seems interested to hear what we've got to say, too. 

[Xiaou2]

 This is the hard part about innovation..  is that you try to keep it a secret from your competitors.. yet, if you introduce the idea here.. you may in fact receive great feedback that will make it even better..

 All too often, a hardware guy wuill make something based on what they feel
is what the majority desire..  but thats not always the case.  Which is why its a good idea to get great feedback from multiple perspectives and prospective
customers.

 I personally think its a cool tech toy..  probably excellent for Sinistar (maybe the best reason to get it).   As well as other non trigger analog games.

 However, I personally can not see how it would be used in any other way...
as the throw is just too much for a typical digital joystick.   And if the restrictor system is a pain to change.. having to dig arround under-panel... then its not what Id be interested in for digital games.

 If there were a way to easily turn a top dial that stuck up from  the control panel,  in which it limited 8 way and 4 way very easily.. then sure, It may be very much a consideration for total replacement solution.

 But, that also assumes its feel and reaction will be Very simular in nature
to the real arcades digital controls.   

 One thing I think should be under consideration, is a rubber spider "X"  simular to that of sinistar.   This would make a perfect solution to a popular game that  is nearly impossible  to control with standard means.   As the  'X" gives the player much more control by multiplcative resistence... rahter than a single resistence  that  once is overcome..  will be  too loose and too easy to slide from slow to hyper fast speeds.

 I have no clue it there is room for such a thing.. if would require a mod, a redesign..ect.   But, its something Im sure that a LOT of people would
want.    Btw- the multiplcative resistence also works on other analog games excelently.. as it gives much  finer control.   Only on  certain games where you need to constantly go swiftly from extreme left to extreme right (for example), might it be a little too slow.  Other than that,  it works great.  Ive tried it via a home-built 'sini-stick'  - simular to andys, but using pots. (and cut up bike tires as the rubber X)

 An add on system as desribed.. which supported motor or electromagnetic  restrictor changes sounds pretty sweet.   I think people wouldnt mind paying an extreme amount for a joystick that does it all..  but.. it really has  to preform as good as them all too...  else it may not do so well in  the hands of true classic purists.   People that actually care very much about
games performance.. and spend and do whatever it takes to get the authentic feel and proper control that they grew up playing with...

 Anyone else agree about the  Sinistar X?  Make your voices heard...


 
 

[u_rebelscum]

Address a bunch of posts, and a first look at analog mode:

The problem is mame is not well written to translate analog to digital*.  Mame looks at each axis separately; the value of the other axis is not looked at during the analog 2 digital translation.  This makes Mame's look more like the following...

So if I am looking at this chart correctly, MAME is using a (very crude) 8-way map for the A-D conversion, even in 4-way games, correct?

You might say that.

However, I'd say mame has NO analog to X-way stick "map" at all. 
It only has the one analog-2-digital-switch "map" for all digital inputs.  IOW, mame converts all the analog (that are set to digital inputs) equally (using the a2d_deadzone setting) and at the same time.  Only then does it ask "is it button, 4-way, 8-way?", as if all were digital inputs in the first place.

I-PAC Mode:  ... you gain one less level of computation (more on that below)....

Digital games ... in USB mode, the stick

[rugby1]

It is not unreasonable to ask that a CLI be included in the next version.   

The UltraStik can be mapped from the CLI already.  From the start AFAIK.  (See posts by SirPoonga and around that area.)

So either it's not unreasonable to ask that it remains in, or it's unreasonable to ask that it's (re)added.

After some further research, you are 100% correct about the CLI already included (http://www.ultimarc.com/ultrastik_prog.html)   

[AndyWarne]

I decided on a 9X9 grid because it seemed like a reasonable size, enough squares to define any possible map which most users would need. Although the hardware is capable of dealing with a larger number of squares I can't see any advantage in this and would mean a longer map-creation process.

I do think that almost all of the comments here are constructive and not overly negative but the kind of healthy discussion that I would have expected. Couple of exceptions, one of them I know the persons first language is not English and that accounts for the reason his comments appear very negative when he does not mean them to be (as he confirmed to me in an email). 

Changing maps via switches: Well I could be persuaded, I had maybe overlooked the Linux issue. There are of course already connection points for switches and the firmware is changeable by the user using a download. I already have plans for alternative firmwares with differing functions.

Yes there is a command-line option.

[AndyWarne]

Surely you are not saying that in USB mode, once I download a map I am stuck in mapped mode and can't get back to pure analog?  (That would be a pretty serious drawback if you could play Star Wars and then PacMan, but then you had to unplug the stick to get back to Star Wars???)

N0. You can revert to analog via USB by downloading an all-analog map. But once this is done, the directional (I-PAC) interface will also be receiving the same map, ie no directions will be sent as it's an analog map, until you again download a digital map.
The option to "force analog via USB" will be added in the next software version. The stick firmware already supports this function.

[syph007]

I would be fine with only software switching, but if there were way to manually switch... what about something like having a 'set mode' button.  Hide it under the cp, or somewhere out of the way. 

To change a mode you hold down the button and move the joystick in 1 of 8 directions to select a mode.  Of course that would assume the stick could hold 8 'maps'.  Thats the least intrustive way I can think of to do a manual switch.  I wouldn't want to have to add more than one button to fasciliate that feature. 

Again, I'm happy with the software switching, but I can see why others would love both methods.

[sirwoogie]

Actually, if you go back in time, you'll see that the discussion about potential problems and/or upgrades started in a different thread, but were MOVED INTO this thread by a moderator who, I'm sure, meant well.

Yeah, that was me. It was better to keep the discussion in one as to not have a ping-pong effect.


Andy,

Another wishlist item: open source of the API and the ability to 3rd party flash the firmware. Have you considered that possibility for something in the future?