How do you set arcade accurate analog settings ??

[reebboy]

I was playing around with my trackball and spinner settings today and realized that through Mame you can set the speeds as fast or slow as you want. I was setting the speed in Centipede but cannot remember how fast the shooter moved around in the actual arcade game. I guess my question is how do I set the speeds to match as close to the arcade settings as possible. I dont know if my track ball and spinner settings are too fast or slow.

[pocketbikez]

the only way would be to do a side by side comparison with a real arcade machine.

and one setting isnt always perfect for every type of controller. trackballs are different sizes and different brands spin the encoders at different speeds. there are many types of spinners all with different sized encoders. various gas pedals turn the potentiometer at different degrees.

when i adjust analog settings i almost always set key/joyspeed to 1% to eliminate any backspin then i adjust sensitivity until the games feels best.

[Martoon]

Actually, I've thought that it would be nice to have this kind of thing documented somewhere, if measurements could be taken by people with access to actual arcade machines.  For example:

Centipede: Shooter traveling from left edge of playing field to right edge maps to top surface of trackball moving a linear horizontal distance of 8.37".

Tempest: One full revolution of shooter around circumfrence of level 1 playing field maps to one full revolution of spinner.

Arkanoid: Paddle traveling form left edge of playing field to right edge maps to 0.46 revolutions of spinner.

etc.

NOTE:  THESE ARE NOT ACTUAL MEASURED VALUES!  These are random values pulled from my hindquarters as an example of how actual measurements could be documented.

Maybe this could be part of the controls.dat project?

[Minwah]

Trouble is, since it varies on most peoples hardware it is difficult to measure or reproduce exactly.

Since most of the controls.dat stuff is obtained from images/documents rather than actual machines, I think it would be extremely tricky (impossible) to accurately obtain this info for each machine.

It would be nice if people with 'real' machines could document this info somhow tho....

[Martoon]

Trouble is, since it varies on most peoples hardware it is difficult to measure or reproduce exactly.
Since most of the controls.dat stuff is obtained from images/documents rather than actual machines, I think it would be extremely tricky (impossible) to accurately obtain this info for each machine.

It would be nice if people with 'real' machines could document this info somhow tho....

Which is why it would need to be done by people with access to the original hardware.  Although, I think some of the information could also be derived with sufficient documentation.  For example, given the diameter of the trackball roller in Centipede and the slotcount of the optical disk, with sufficient information about how the ROM translates input signals into motion, you could probably calculate the relation between trackball linear motion and onscreen motion.  But definitely easier if you just have a machine you can measure from.

I certainly wouldn't want it to hold up the other things being documented in the controls.dat project, but if/when this type of info is collected somewhere, the comments field in controls.dat seems like a logical place for it to be copied.

Anyone interested in starting a repository for this kind of measurement info, and getting people to contribute?  I have access to virtually no original machines where I live, so I'm afraid I can't really contribute anything myself.

[NoOne=NBA=]

I just adjust them until they feel right to me.
It's all about playability to me.
If a game had too long of a throw in the arcade, or the trackball was not as sensitive as I think it SHOULD have been, I don't see any reason not to "improve" them.

On that note, I also rearrange controls the way I think they should have been in the arcade.
If the controls were bad in the arcade, I don't see any reason not to improve upon them, as well.
A good example of this is Tempest--which SHOULD have had the spinner on the left, like all the other spinner games at the time.

[reebboy]

I figured it would be next to impossible to get accurate speeds without performing some serious mathamatical calculations. If anyone has access to the actual arcade games maybe they can post  what the speeds or sensitivity should be around. I.E. Does the the shooter zip accross the screen with barley any touch of the ball or do you have to spin it several times just to get it to move 1/2 way accross the screen. If I can get the settings close, that will be good enough.

I can't remember but did Millipede and Missile Command require alot of spinning to get the cursor/shooter to move or am I thinking of something else.

[NoOne=NBA=]

That is really a subjective question, and depends on the condition of the trackball in question, and your playing style.

In general, no they didn't require alot of spinning--if you could move the trackball quickly.
If you get on a machine with a poorly maintained trackball, then it will require more spinning to get the cursor where you want it.
On Missle Command in particular, you could "spin" the trackball, and have it continue after your hand left the ball--getting you to the other side.

I set both so that a quick spin in a given direction, with very little "hands off" time, will get you all the way from corner to corner.
That way I don't "run out of hand" on them.

[pocketbikez]

i did a side by side comparison of my mame driving cab and my original spyhunter standup machine. i got the controls to mimic the standup exactly but this info would only be useful to someone that is using a spyhunter wheel, a happcontrols flat pedal, and interfaced with a dual strike joystick hack (with the dual strike software range of motion adjusted the same). ill post it just in case.

P1 pedal 1 key/joyspeed 1%
P1 pedal 1 sensitivity 35%

Paddle key/joyspeed 46%
Paddle sensitivity 36%

i also did it with my original APB standup. the gaspedal setting would be accurate for someone using a happ flatpedal/ dual strike hack, and a happcontrols 360 degree steering wheel interfaced with a mouse hack.

P1 pedal key/joyspeed 1%
P1 pedal sensitivity 130%

Dial key/joyspeed 1%
Dial sensitivity 21%

* i should add that this is with mame version .83

[telengard]

i did a side by side comparison of my mame driving cab and my original spyhunter standup machine. i got the controls to mimic the standup exactly but this info would only be useful to someone that is using a spyhunter wheel, a happcontrols flat pedal, and interfaced with a dual strike joystick hack (with the dual strike software range of motion adjusted the same). ill post it just in case.

P1 pedal 1 key/joyspeed 1%
P1 pedal 1 sensitivity 35%

Paddle key/joyspeed 46%
Paddle sensitivity 36%

i also did it with my original APB standup. the gaspedal setting would be accurate for someone using a happ flatpedal/ dual strike hack, and a happcontrols 360 degree steering wheel interfaced with a mouse hack.

P1 pedal key/joyspeed 1%
P1 pedal sensitivity 130%

Dial key/joyspeed 1%
Dial sensitivity 21%

* i should add that this is with mame version .83

I like the idea of a web page to store all of these hardware configs w/ associated analog settings.  I would find them useful if for nothing else than getting me hopefully close and then tweaking a little.  I have a friend who is a Marble Madness maniac and we spent quite a while tweaking the settings ( which I think I may have since lost ) so it'd be nice to store them somewhere.

[reebboy]

I'm curious if I use someone elses sensitivity/speed settings on my system, are my controls going to play close but not exact to that other persons settings or is it going to be way off.

It would be interesting to see someone with 2 mame systems, with different controls and configurations but same analog settings, as to how much of a difference the same settings would be.

Just a thought.

[Larry Smith]

Depending on your controls, it could be close or it could be way off. The right way to document this stuff is to map gameplay elements to control info on a real machine. This has been done for tempest (on board 1 on trip around the circle eqals either 2 or 3 turns of the spinner i forget which. do a search for a preveios post). Unfortunately I don't know of any other mappings and am not sure of good measurments to take for other games. Martoon's ideas for centipede and arkanoid seem reasonable does anyone have any ideas for other analg games, especially driving games like pole position. If no one else is willing to do it I guess I will be the one to make a web page documenting this stuff, but I will need help with ideas about how to measure as well as the measurements themselves. And let me repeat this will be about measuring real arcade machines not what sensitivity and peed setting you like to use in mame.

[Sylentwulf]

Question I've been meaning to ask is whats the difference between sensitivity and joy/key speed?

[Minwah]

Question I've been meaning to ask is whats the difference between sensitivity and joy/key speed?

Sensitivity applies when using a real analog device, joy/key speed is when you have a digital input mapped to an in-game analog one.

[telengard]

Question I've been meaning to ask is whats the difference between sensitivity and joy/key speed?

Sensitivity applies when using a real analog device, joy/key speed is when you have a digital input mapped to an in-game analog one.

Does this mean that if I am using my spinner that the speed setting doesn't come into play at all?

[Minwah]

Does this mean that if I am using my spinner that the speed setting doesn't come into play at all?

Yes, afaik.

[alank2]

I have information on Tempest that some people with a real Tempest were kind enough to post for me:

There is NO acceleration component to Tempest.  If you turn the spinner 180 deg fast or slow, you move the same amount in the game.  You may have to change your windows mouse settings to eliminate acceleration.

If you turn the spinner 180 degrees, your shooter will move 3 full spots in the first level.

I have  an Oscar Controls Vortex and I found that a sensitivity of 40% worked perfectly to get me 3 spots @ 180 degrees.

Each persons control might be different.  What are their windows mouse settings?  What size is the trackball?  How fast does your spinner move the pointer?  This is why it is best to list the measurements from an actual arcade game, so each person can adjust their controls to the original.

I don't know any games except Tempest as far as measurements, but I've been thinking of posting and finding someone with a real Arkanoid to ask next...

Good Luck,

Alan

[pocketbikez]

ive noticed on my older computers that key/joyspeed has more effect. when key/joyspeed is set high i get backspin with a spinner or trackball, a good example is in the game Kick. when i adjust key/joyspeed lower the backspin is eliminated.

on oscarcontrols website this key/joyspeed setting is discussed:
http://www.oscarcontrols.com/support_spinner.htm

also, on my spyhunter settings it seems key/joyspeed effects the initial movement of the race car while the sensitivity setting affects the overall turning speed.

[NoOne=NBA=]

URebelScum posted awhile back that, although the key/joy speed SHOULDN'T have any effect on true analog controls, it DOES for some reason.

[alank2]

I think key/joy speed worked differently in different versions of MAME.  The very latest version seems to have replaced it with "digital speed".  I know that in recent versions, the key/joy speed did NOT have any effect because I tried setting it to its lowest and highest settings and it only affected the joystick control, not the analog control.  I have heard that in older versions of MAME it did play a role in analog as well though.

Thanks,

Alan

[Jakobud]

Is the analog setup information stored (and editable) in the new CFG file format now?

[telengard]

If anyone is interested I spent some time (and some tokens) taking measurements on the analog machines.  They are not 100% accurate or scientific in any way but they will be enough for me to hopefully get the settings on my cab to match:

Warlords
120-130 deg one traversal end to end of castle

Super Breakout
120-130 deg one traversal across the screen

Turbo
2 1/2 turns to go across the screen

Sprint
1/2 turn is 90 deg

Crystal Castles
1st screen BB moves length of line above his head from center of
trackball moved to one side

Centipede
1 trackball move 1 end to other (stock tb)
2 1/2 mushroom lengths

Missle Command
1 full end to end visible trackball movement goes from left to the
right of the little bump after the missle base

Super Sprint
360 deg turn of wheel moves about 170 deg

Arkanoid
120-130 deg one traversal across the screen

Tempest
1 full turn moves 5 spaces on first screen

I tried a bunch of times to measure pole position, but it is kinda wierd how the steering works and I had very little luck.

I know that some of my descriptions sound kinda wacky but I'm happy to just get in the general area to start.  I also know that someone who owns these machines could probably do a better job measuring (hint hint).  If anyone has better ideas for taking measurements I can try them next time I go to funspot.

[telengard]

If anyone is interested I spent some time (and some tokens) taking measurements on the analog machines.  They are not 100% accurate or scientific in any way but they will be enough for me to hopefully get the settings on my cab to match:

Warlords
120-130 deg one traversal end to end of castle

Super Breakout
120-130 deg one traversal across the screen

Turbo
2 1/2 turns to go across the screen

Sprint
1/2 turn is 90 deg

Crystal Castles
1st screen BB moves length of line above his head from center of
trackball moved to one side

Centipede
1 trackball move 1 end to other (stock tb)
2 1/2 mushroom lengths

Missle Command
1 full end to end visible trackball movement goes from left to the
right of the little bump after the missle base

Super Sprint
360 deg turn of wheel moves about 170 deg

Arkanoid
120-130 deg one traversal across the screen

Tempest
1 full turn moves 5 spaces on first screen

I tried a bunch of times to measure pole position, but it is kinda wierd how the steering works and I had very little luck.

I know that some of my descriptions sound kinda wacky but I'm happy to just get in the general area to start.  I also know that someone who owns these machines could probably do a better job measuring (hint hint).  If anyone has better ideas for taking measurements I can try them next time I go to funspot.

Just another observation while I'm at it.   I played a lot of Crystal Castles while there, and found I could get a pretty decent score (around 80k or so).  At home I can't even get over half of that.  I noticed that when I move up or down the bear tends to go off in a different direction when reaching an intersection.  I can't think of a better way to describe this.  It definitely didn't feel like this in the arcade.  It felt very precise.  I wonder if the trackball circuit has some sort of debouncing circuit or something.  Or maybe the trackball was oriented slightly different.

And a question.  If people do end up matching analog settings in MAME for real hardware measurements (for certain hardware, i.e. settings for model 3, settings for vortex, etc) does the Windows setting for the mouse come into play?  I have acceleration turned off but how much does the mouse speed come into play?

[telengard]

Just another observation while I'm at it.   I played a lot of Crystal Castles while there, and found I could get a pretty decent score (around 80k or so).  At home I can't even get over half of that.  I noticed that when I move up or down the bear tends to go off in a different direction when reaching an intersection.  I can't think of a better way to describe this.  It definitely didn't feel like this in the arcade.  It felt very precise.  I wonder if the trackball circuit has some sort of debouncing circuit or something.  Or maybe the trackball was oriented slightly different.

And a question.  If people do end up matching analog settings in MAME for real hardware measurements (for certain hardware, i.e. settings for model 3, settings for vortex, etc) does the Windows setting for the mouse come into play?  I have acceleration turned off but how much does the mouse speed come into play?

I tried lubing the bearings in the trackball and although it is a lot smoother now it still exhibits the same issue.  Any ideas?

[reebboy]

Thanks for the measurements. They'll definitly get me in the ball park for setting my controls.

I have a question on Centipede, are you saying that one 360 degree movement of the track ball only moves the shooter 2 1/2 mushroom lengths?? That seems like alot of spinning just to get it to move a short distance. Thinking about it now, I believe I do remember back in the day having to spin the hell out of the ball just to get it to move.

[Carsten Carlos]

I have a question on Centipede, are you saying that one 360 degree movement of the track ball only moves the shooter 2 1/2 mushroom lengths??

I guess telengard meant to turn the visible area of the ball from one side to the other! (Placing your finger on the left edge of the ball and turn to the right edge)

[telengard]

I have a question on Centipede, are you saying that one 360 degree movement of the track ball only moves the shooter 2 1/2 mushroom lengths??

I guess telengard meant to turn the visible area of the ball from one side to the other! (Placing your finger on the left edge of the ball and turn to the right edge)

Exactly.  For me to test a full revolution I would have had to mark the trackball in some way to know it's gone around and that might have been frowned upon.  

I'm finding that the Analog Sensitivity isn't enough to duplicate the feel of some games.  As I mentioned I am able to get the trackball movement amount to match in Crystal Castles, but there is something missing.  I've been looking at the schematics and MAME code to see if there is something missing.  All I've seen so far is a chip called LETA.  It seems to handle the trackball input.  My guess is that it does things a little differently than something like the optipac.  I could be totally wrong though.  One thing is for sure, the trackball movement is not easily duplicated for that game.

[Matt Berry]

Has anybody looked into this any further. I'm trying to get my settings close with centipede. I'm using a happ 3" trackball interfaced to a opti pac. Any help would be greatly appreciated.

[telengard]

Has anybody looked into this any further. I'm trying to get my settings close with centipede. I'm using a happ 3" trackball interfaced to a opti pac. Any help would be greatly appreciated.

Hi Matt,

I wish I could say yes.  The closest I've come is the measurements I took that you can see in this thread.  I also have a Happs 3" and cannot seem to get it exact.  Keep us posted if you find anything.

~telengard

[Pixelhugger]

What a fantastic thread. Thanks a million for those Tempest measurements Telengard and Alan. I was almost at the point of knocking the door of some stangers house after seeing an old Tempest machine in his open garage. THANK YOU THANK YOU!!! I been waiting for this info!  

[Kremmit]

What a fantastic thread. Thanks a million for those Tempest measurements Telengard and Alan. I was almost at the point of knocking the door of some stangers house after seeing an old Tempest machine in his open garage. THANK YOU THANK YOU!!! I been waiting for this info!  

You should knock anyway- he may have "had that thing in there taking up space for years", but never plays it...

Quote from: telengard on September 12, 2004, 10:48:17 PM

Or maybe the trackball was oriented slightly different.

I've got a Crystal Castles cab here- it's been (badly) converted, but the original control panel is still on it- with joysticks poking out of the button holes.  If anybody's can find a pic of the original Atari trackball, I can match up the mounting hole pattern with my control panel and see if they mabye did mount it at an angle- which sounds unlikely, but just might make sense, since the playfield directions all ran to diagonals.