Potential testing methodology to measure continuous input lag and misbehavior of mouse movement

[Brainlet]

Put a mousepad over the bottom half of a laid down monitor with the center exposed. Match DPI + sensitivity for a 1:1 physical mouse and on-screen cursor movement in a 2d environment.
Place the cursor vertically centered on-screen and place the mouse directly under the cursor on the mousepad.
Use a construction similar to this https://youtu.be/a2kITylfxXQ?t=20 to move the mouse horizontally.

[Chief Blur Buster]

This is good for reviewers -- the upcoming TestUFO Mouse Test would be compatible with such a rig.

That said, TestUFO Mouse Test is also designed to be users, and I've got a mechanism that sort of filters hand-movement error, to measure mouse stutter volatility. There are multiple layers of stutter onion (other than hand stutter, stiction effects, potentially pollrate vs refreshrate, pollrate vs framerate, framerate vs refreshrate, etc) that has to be somewhat unpeeled, but rawinput + deltas + timestamps tells a lot of story in statistical analysis + visualizations. What's amazing is that same-human mousemoves of different mouse settings, show quite fascinating differences.

[Conan]

This is good for reviewers -- the upcoming TestUFO Mouse Test would be compatible with such a rig.

That said, TestUFO Mouse Test is also designed to be users, and I've got a mechanism that sort of filters hand-movement error, to measure mouse stutter volatility. There are multiple layers of stutter onion (other than hand stutter, stiction effects, potentially pollrate vs refreshrate, pollrate vs framerate, framerate vs refreshrate, etc) that has to be somewhat unpeeled, but rawinput + deltas + timestamps tells a lot of story in statistical analysis + visualizations. What's amazing is that same-human mousemoves of different mouse settings, show quite fascinating differences.

I had to go back in time to search for this post since I had no time of replying when I saw it.
When is TestUFO Mouse Test gonna have a green light? You gonna do some post about it? Like in-depth point of view?
I hope most mice reviewers, preferably from gaming department will utilize it in their reviews.

[Chief Blur Buster]

I had to go back in time to search for this post since I had no time of replying when I saw it.
When is TestUFO Mouse Test gonna have a green light? You gonna do some post about it? Like in-depth point of view?
I hope most mice reviewers, preferably from gaming department will utilize it in their reviews.

TestUFO mouse test is launching probably this month (or next).

Also, I am now in the need of a mouse testing rig for 1000Hz-vs-8000Hz tests, so I am researching building a DIY rig.

I am evaluating purchasing an app-controlled motorized camera slider (~$300) and modifying it to move a mouse underneath. With an iPhone/iPad, I would be able to cause repeatable consistent-speed mouse movement for my stroboscopic mouse-arrow tests.



Right now, I'm getting dramatically different stroboscopics from 1000Hz pollrate vs 8000Hz pollrate, so it has perked my interest in this.

[Conan]

As I don't understand the complexity of this specific subject, I can fathom due to a similar knowledge depth in a different field of expertise and there for I find this to be extremely exciting.

I do wish you keep us informed about the progress Chief.
Thank you

[Chief Blur Buster]

Just like audio-frequencies, odd beat-frequencies and harmonics can occur digitally between two frequencies (poll rate and refresh rate).

For example, here's 1000Hz versus 8000Hz at 360Hz.
Camera: Apple iPad Mini 5, DSLRCamera app, 1/8sec (approx), 50 ISO, manual focussed to screen as I circle 8000Hz mouse on 360Hz PG259QN

Sneak preview tweet:
https://twitter.com/BlurBusters/status/ ... 4119605248

1000 Hz Mouse on 360 Hz Monitor


8000 Hz Mouse on 360 Hz Monitor


As the DISPLAY REFRESH RATE gets closer to the MOUSE POLL RATE, the jittering between two frequencies begins to get worse and worse.

So as we continue the refresh rate race to retina refresh rates, bumping the poll ceiling higher is important, making those poll granularities finer (0.125us instead of 1ms)

This is just a sneak preview. Many more of these tests will be part of my upcoming mouse article (awaiting more testing equipment).