Overclock.net › Forums › Components › Mice › Finalmouse 2015
New Posts  All Forums:Forum Nav:

Finalmouse 2015 - Page 412

post #4111 of 4669
Our current debounce logic registers the click immediately it's the first thing that happens.

And I believe polling is taken care of. I think we just wanted to know the actual poll counts to add to the debounce before another click can be registered. I guess he wanted to save time trial and erroring?

And on a side note as you all know tournament pro is available in North America .

Edit: how would 5ms added on the top make it more robust? Could you clarify this for me as I'm curious and will pass along the info.
Edited by FinalmouseJude - 1/28/16 at 4:43pm
post #4112 of 4669
cannot understand your second sentence tongue.gif
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
post #4113 of 4669
Nvm you had answered the question in your third bullet point.

15-25 counts. Sounds right smile.gif
post #4114 of 4669
Quote:
Originally Posted by FinalmouseJude View Post

Nvm you had answered the question in your third bullet point.

15-25 counts. Sounds right smile.gif

When will Europe get stock?
post #4115 of 4669
Just don't add the 5ms, it is a compromise.. Previous method used in FM15 was good, keep using it..

In csgo fastest fire-rate is with Negev 1000 rpm, meaning 60 ms between every bullet pressing key for 25 ms should be enough for shooting 1 bullet at a time..

If you added 5 ms delay you would get 20 ms duration for 1-tapping 5 ms different in duration wouldn't make as big of a difference as it does in click delay for fast flicks..

However I guess csgo has problems allowing next bullet to fire if button release is closer to the time of next bullet than first.

Meaning if the first bullet leaves at the time of 0 ms and at the firerate of 1000rpm next bullet would leave at 60 ms.. Now if you press button it registers at 0 ms when first bullet leaves and the button release registers at 25ms it still should only shoot 1 bullet since 25 ms is closer to 0 ms than 60 ms.. But if you get bugged for over 5 ms because of some kind of fluctuation or anything and it takes over 30 ms for the button release, then the second bullet would leave as well..

Fluctuation could be caused because of server tick-rate for example, on 128 tic servers one tick last about 7,8 ms.. If you have bad luck and you release the button just before the next tick, it would add over 7 ms to the delay before the server registers your button release making it 32 ms and causing second bullet to leave even if you try to shoot 1 bullet at a time..

Well, this all is only speculation I haven't tested this at all and my knowledge about client-server behavior and ways how the game handles firing mechanism is limited.. So take this with grain of salt and you can do research according to those things I mentioned.

Purely from my experience I can say that I has been more difficult to 1 tap with some of those Logitech mice using this kind of denouncing even though looking at pure numbers it shouldn't be a problem.. That's why I started wondering those other reasons for that behavior I mentioned above..

tl;dr = Debouncing in FM15 was fine, keep using it, don't add any delay..
post #4116 of 4669
Don't add any delay. If you really have to add filtering only add 1ms of filtering. Any more /should be/ unnecessary for mouse buttons. Just do the "don't release for X milliseconds" thing. I would actually use 7ms since that's slightly lower than the fastest click-unclick I'm humanly capable of and it's below a strict 1/128th of a second, but it might be pushing it without doing any filtering. 10-15ms would be fine.
post #4117 of 4669
it is beneficial in the long run to add a delay to the releases. doing so is more robust than the method i proposed in the past (ignore changes for ~20ms after a click or release)

i think this is how logitech does it too
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
post #4118 of 4669
Yeah I meant to the signal as a whole. I brought up release delay in the last three sentences.

Definitely important for you to bring up the difference between "ignore everything for X milliseconds after clicking" and "don't unclick until you have X straight milliseconds of being released", though. The former doesn't do anything to prevent momentary reclicking when you release.

Actually now I remember TF2 has sticky bombs where they fire when you release the button. 20ms delay on release would be really bad for that. How about this:

1) 1ms filter
2) Xms blockage after it detects a click
3) Xms blockage after it detects an unclick

Just doing 2 doesn't prevent glitchy clicking when you release, but adding 3 to the system does. It's not as robust, you're right, but also doing filtering should help prevent problems and not add as much latency to releases as blocking on releases does.
Edited by wareya - 1/28/16 at 6:03pm
post #4119 of 4669
Quote:
Originally Posted by qsxcv View Post

it is beneficial in the long run to add a delay to the releases. doing so is more robust than the method i proposed in the past (ignore changes for ~20ms after a click or release)

i think this is how logitech does it too
Why is it beneficial to have that extra delay on releases? I definitely don't know much about debouncing but couldn't this problem be solved with different hardware instead of with software? Sorry if those are stupid questions. Like I said I don't know much about this topic.
post #4120 of 4669
You can mess with different hardware designs, but it's a whole new world when you do that. Definitely not something you would do outside of RnD.

With the current hardware, maybe you could dynamically switch between post-release blockage and release delay depending on how long the mouse button has been pressed and what the sensor is doing. Remember when someone had to put something over their sensor when doing mouse button smash tests because it ramped up the filter?

(Don't increase click down delay when someone lifts up the mouse, people who are sniping sometimes lift their mouse to make sure they don't move their crosshair when they click.)
Edited by wareya - 1/28/16 at 6:20pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Mice
Overclock.net › Forums › Components › Mice › Finalmouse 2015