Hi guys.
Stopped in to say I've upgraded my cab from Vista to Win 7. In doing so, I looked into what it takes to get a higher USB mouse polling rate then the stock 125 HZ.
This is needed so games like Tempest and Centipede don't overflow and start moving the opposite direction when you move the mouse fast and exceed +127/-128 (8bit) of movement per 8ms.
IIRC newer OptiPac solved this by using 16bit data and GroovyGameGear hard limits the data to +127/-128. I still like the faster poll rate which should give less lag.
In a previous post, that I figured was best not to resurrect from the dead, I posted info on how to use windows in test mode and use an unsigned USB driver.
http://forum.arcadecontrols.com/index.php/topic,52061.msg510015.html#msg510015There is now a signed driver, meaning you can run it normally without test mode. (hidusbf-7-8-81-10)
http://www.overclock.net/forum/375-mice/1597441-digitally-signed-sweetlow-1000hz-mouse-driver-5.html#post25075429I tested this on Win 7 SP1 64 bit with a USB 2 port. The thread goes on to mention how to get it working with USB 3, but I did not look into it. Something about a 2K poll rate driver that works with USB 3 (best) or else you need to set your USB 3 port to forced 2.0 mode in your BOIS, which seems pointless.
Run Setup.
Click on Install Service.
Then I needed to select Process Parent and then click Restart. Turn off Process Parent and Restart and then turn on Process Parent and then click Restart again.
I also needed to select Filter on the device and parent. Depending on your mouse you may not need to. My Microsoft mouse worked right away, but the OptiPAc needed the extra work.
No need to go past 500Hz.
For those testing Keyboard Lag, you might want to turn off Mice Only and do the same setup for the keyboard and see what happens.
As far as Centipede goes, it has been a while since I looked at the MAME code, and I don't have time at the moment. But I remember the Centipede hardware polled at 240Hz because it only used 4bit data and needed the faster rate so the trackball did not overflow.
IIRC MAME reads the Computer hardware at the poll rate, adds it up until the driver's (not screen) vertical refresh rate. Can't remember how it handles hardware that asks for data quicker then the vertical refresh rate. But I saw a video of real gameplay and yeah, there is definitely an issue, that is not solved by the faster poll rate.
D.
P.S Use mouse rate checker to check the mouse poll rate.
Quick Google search gave me this:
https://eu.cybergamer.com/files/2260/