Mouse Drift measure tool

SvenBent

2[H]4U
Joined
Sep 13, 2008
Messages
3,323
I' just finished a very internal alpha version of a little utility to measure mouse drift.
Basically it will test of you mouse suffers from mouse drift which is the movement of the cursor even when you are not moving the mouse. This can typically be caused buy dirt/hair on the optical sensor, or incompatibility between sensor and mouse surface.

If there is any interest for the utility i will wrap it up to a more finished and user-friendly states and release it here on this forum.

With a little tweaked code it will also measure the refresh rate of your mouse.
 
So, for example it would allow me to narrow down spontaneous wake-ups on PCs with wake-on-usb enabled to the mouse 'drifting'? Or sending any other spurious signals?
 
Yes it has two testing modes
quick/Passive which means you just leave it there and it will check mouse movemnts until you hit enter ( might change that) and it will tell you if you mouse have moved. aka it have disturbed sleep mode
Long/exhausitive : requries you to move you mouse up/down and after that left/right. this is to better catch drift that occurs during motion.

So far i haven't had any drift besides if I have provoked it myself by tapping the table or flicking at the mousecord. im going to try with differetn surface just to see how usefull it is to do mouse drift measuring
 
I'm onboard then! I'd gladly give it a go if you like. I have some mice that I have to test with a Linux kernel because udevmonitor and the like allow me to see moments when the mouse 'has problems' :D

The refresh rate will be 125 hz on any default system, yes? Unless I have the 'hacked' inf file installed for a given device?

I used to use dseo and the like fo force 500hz on my MX518 so I can test your tool on XP and on 7.
 
As far as 'generating' drift, I just tried a blue laser pointer (the type that burns skin and sets stuff on fire) and a simple flash at the sensor area provoked the cursor to move around.
 
Yes it has two testing modes
quick/Passive which means you just leave it there and it will check mouse movemnts until you hit enter ( might change that) and it will tell you if you mouse have moved. aka it have disturbed sleep mode
Long/exhausitive : requries you to move you mouse up/down and after that left/right. this is to better catch drift that occurs during motion.

So far i haven't had any drift besides if I have provoked it myself by tapping the table or flicking at the mousecord. im going to try with differetn surface just to see how usefull it is to do mouse drift measuring

A while ago, I bought a Logitech Marathon M705 mouse but something was terribly wrong; I couldn't seem to click anything accurately. I confirmed that with this mouse accuracy game which made it very obvious the mouse was drifting off target while in motion. The M705 is a wireless mouse, and apparently its signal was getting interference. Into the trash it went. I tell you this in case you want to obtain a mouse that might have drift out of the box, though whether you have the necessary interference. You can also consider the idea that wireless mice in general might be vulnerable to interference that causes drift.

That mouse accuracy game is probably good enough for testing accuracy in motion but I haven't seen anything that tests for drift while stationary.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
You can also consider the idea that wireless mice in general might be vulnerable to interference that causes drift.

Personally I have seen this happen once - the mouse was stationary and there were no vibrations but it would move in a straight line in a particular (repeatable) direction when a desk lamp was lit.

OP do you use WM_MOUSEMOVE? or something lower level? You don't have to say, just curious really.

To add from the top of my head, a glass surface should provide a great testing ground.
 
Sorry for the late feedback been working on my Project Mercury tool.

Mouse drifter is comming off the mouserate meausre is working fine. but the sensing mousedrift while moving is pretty hard and i need a good precise way to do this.
passive mode aka mouse drifting when the mouse is not beeing moved works pefecyl so far

and yes I'm monitoring WM_mousemove. not directly though yet but pretty much the same effect.

michalrz
Ty for the info i'll try repeat that test
and you you mouse will be running with whatever is its default polling rate. im using logitech g500 and it has a 500hz polling rate as default with no drivers installed


evilsofa
I might look into picking on up from ebay.
oterwise i get alot of used mice in through my work as i work in Electronic refurbing company

give me a week and i should release first mouse drift version
 
Mouse drifter is comming off the mouserate meausre is working fine. but the sensing mousedrift while moving is pretty hard and i need a good precise way to do this.

No worries, it's definitely a niche product and definitely thinking outside the box. Projects should be fun :)

I'm guessing you should ask the user to run the cursor across a horizontal line, vertical and 45 degrees. On each mousemove, calculate the 'right' cursor position and then actually check it via getcursorpos. One horizontal sweep will give you like a thousand data points. Calculate the error and present it in the form of a XY distribution diagram for each axis. That would already be plenty, no?
 
My old Microsoft optical mouse from 2001 did this. Haven't had the issue since.
 
Hey! I was looking for a tool like the one discussed here and didn't found any. So I made a simple online drift test. It's just checks the mouse while stationary, but it is better than nothing I guess.
I hope it would be helpful for others like me stumbling across this topic in search.
 
Back
Top