Build Your Own Arcade Controls Forum
Main => Main Forum => Topic started by: Martoon on July 22, 2004, 02:16:31 pm
-
I'm sure this is a common enough issue that's it's addressed in a FAQ or other threads, but the keywords I'm trying don't seem to be turning anything up.
I'm a newbie to building and interfacing arcade controls (I'm a software engineer, not a hardware guy). I got a used Betson Imperial 3" trackball, and hacked an old PS/2 mouse. The trackball moves the mouse pointer as expected.
When I use it in MAME (I'm running Mame32), the way the games respond to the trackball doesn't feel quite right. In Centipede, if I spin the ball quickly, the ship slows down and jitters, like it's missing samples from the trackball. So then I tried Centipede with a regular USB optical mouse, and noticed that it does the same thing if I move the mouse quickly. The thing is, I don't remember actual Centipede machines behaving this way. It seems like you could give the ball a hard lateral swipe, and your ship would glide across the screen fairly fast.
Or does an actual Centipede machine work like this, and I'm just not remembering correctly? If the original machine didn't have this problem, is there a better way to interface a trackball to MAME to more accurately reproduce the game's behaviour?
In the hardware settings for the PS/2 mouse, under the Advanced Settings tab, the sample rate was set to 60 reports/second. I pushed it up to the max of 100, and it seemed to make no difference. Also, I'm running Win2k. Would I get better mouse response under a different version of Windows?
Any help appreciated.
-
This is called backspin. You are spinning the trackball faster thatn it can read the samples. There are tons of threads on backspin.
The question is where is the backspin occurring? If the backspin doesn't happen in Windows, turn down the sensitivity in the game. Changing mouse drivers could help. Hacking to USB instead of PS/2 might help, as USP has twice the sample rate of PS/2.
You should at least be able to get PS/2 up to 200 samples per second. Try the Microsoft IntelliMouse drivers.
--Chris
-
Ahh, backspin. Now that I have a keyword to search for, I'm finding plenty o' discussion. Thanks for that.
Just from watching the mouse pointer in Windows, and watching things in-game, I'm thinking that most of the backspin is occuring in the game. With the new info and understanding I have now, I'm guessing an actual Centipede machine would in fact exhibit backspin, if you whacked the ball hard enough. After tweaking the sensitivity settings for each game, I guess they do feel about right. It's just been a long time since I had the priveledge of playing some of these games in a real arcade. It's good to be back. :D
I installed the IntelliMouse drivers, but the sample rate for the PS/2 mouse still has a maximum setting of 100 sam/sec., but that seems to be good enough (since I don't see any visible backspin artifacts when moving the mouse pointer on the desktop, but sometimes it can be hard to tell from the mouse pointer).
Now to get this thing mounted in a panel....
-
Try Logitech PS/2 mouse drivers. I have seen 200 samples per second as the max on some of those.
Maybe it's just the mouse you hacked too. They aren't all created equal, so another brand may fare better.
FWIW,
RandyT
-
Try Logitech PS/2 mouse drivers. I have seen 200 samples per second as the max on some of those.
Thanks, I tried the Logitech driver, and it does, in fact, let you set the sample rate up to 200. It didn't seem to perform any better at that sample rate, though, and somehow the mouse pointer motion "felt" less smooth with the Logitech driver (at both 100 and 200 samp/sec). However, that's all very subjective, and it's probably just my imagination. In any case, I think the backspin is happening in the games, not the mouse driver, so I'm just going with default Windows PS/2 driver.
-
I'm not 100% sure (I have only played a real Centipede once), but I thought I read somewhere that the real machines had backspin problems at times...
-
Yes, centipede did experience backspin problems when really whipping that trackball around.