Originally Posted by badben25
Button down is when your gun in an FPS starts shooting. I think that alone is enough to question why it has to be a full click like you ask?
Not sure if you're still under misunderstanding, or if I'm not understanding your query.
1) That's exactly what I have been saying in my previous posts. Because of the linking to techpowerup and techpowerup using the same term "click latency" (which I'd call "min click durations", as I've explained) I had assumed you're measuring the same here.
2) I never asked for a full click. I did the opposite and pointed out that techpowerup's "click latency" is irrelevant for competitive FPS where input latency matters most.
3) My query was quite simple, I've repeated it and you're still not understanding it.
Let's make this as simple as possible:
Q1) How do you define "Click Latency" as used in #1 in this thread
Q2) How did you measure the values in the table in #1?
Originally Posted by cdcd
qsxcv's program records button press events, not button press and button release. Debouncing happens for the button press event, and that's what's measured. This applies to any figures gathered with qsxcv's program, i.e. both the ones in this thread here and on TPU.
The l.exe and r.exe programs measure click duration (mouse down to mouse up event) of the left and right button respectively. a.exe indeed measures difference between left and right mouse down events.
Debouncing dictates min click duration. For input latency (physically pressing the button to the OS receiving the button down event) it is irrelevant.
Which is why your second sentence makes no sense to me and I don't understand your last sentence. Debouncing doesn't happen for the button press event. When the firmware reads a button press it should immediately send this as an event and start debouncing
. As such, debouncing doesn't have any effect on input latency.
But as I've already explained above, the firmware could be written to "cheat" with debounce time by delaying sending that initial press event. The l/r.exe programs would then display a shorter click duration even though physically debounce time wasn't worted (that's the cheating part). This would come at the cost of increased input latency.