XIM Community

1000hz on pc vs console?  (Read 2502 times)

Offline antithesis

  • MVP
  • *
  • Posts: 9875
    • View Profile
    • Mod Squad
Re: 1000hz on pc vs console?
« Reply #15 on: 09:47 PM - 12/07/19 »
Thanks for the correction, that does sound like XIM APEX input polling is 1ms and Response Rate is output polling. But I'm not sure where you're getting the XIM-induced lag.

Leaving the mouse at 1000Hz means XIM APEX will detect a button press within 1ms of the input being passed from the mouse to XIM APEX. At 125Hz RR, the button press signal that is output from XIM APEX to the console will be 1ms old at most and will be passed in the next available 8ms tick to the console. So...where's the lag?

Conversely, setting the mouse to 125Hz and XIM APEX to 125Hz is still an 8ms frame, but there may be lag from when the mouse button is pressed and that input signal is passed to XIM, meaning it could be up to 7ms old rather than 1ms. That is mouse-induced lag, not XIM-induced and is the opposite of what you've described.

OBsIV is simply saying that the input signal XIM APEX passes through to the console is at most 1ms old, not up to 7ms, regardless of the Response Rate. So XIM isn't inducing lag at all.

Your "real" option will pass input from the mouse to XIM APEX up to 7ms later than the button press and then pass in the next Response Rate tick to the console, inducing lag.

What this implies is there is zero benefit of reducing the mouse input polling below 1000Hz - XIM APEX will drop input down to the defined Response Rate and output that to the console (which should reduce jitter etc).

Does that make sense?!?
« Last Edit: 04:01 AM - 12/08/19 by antithesis »

Official Australian distributor for XIM APEX, Titan One & Titan Two at Mod Squad
XIM APEX demos on Twitch and YouTube. Follow Twitter for live stream alerts

Offline Bb4life1991

  • MVP
  • *
  • Posts: 716
    • View Profile
  • Gamertag: YT Boogeyman
Re: 1000hz on pc vs console?
« Reply #16 on: 11:27 PM - 12/07/19 »
Thanks for the correction, that does sound like XIM APEX input polling is 1ms and Response Rate is output polling. But I'm not sure where you're getting the XIM-induced lag from.

Leaving the mouse at 1000Hz means XIM APEX will detect a button press within 1ms of the input being passed from the mouse to XIM APEX. At 125Hz RR, the button press signal that is output from XIM APEX to the console will be 1ms old at most and will be passed in the next available 8ms tick to the console. So...where's the lag?

Conversely, setting the mouse to 125Hz and XIM APEX to 125Hz is still an 8ms frame, but there may be lag from when the mouse button is pressed and that input signal is passed to XIM, meaning it could be up to 7ms old rather than 1ms. That is mouse-induced lag, not XIM-induced and is the opposite of what you've described.

OBsIV is simply saying that the input signal XIM APEX passes through to the console is at most 1ms old, not up to 7ms, regardless of the Response Rate. So XIM isn't inducing lag at all.

Your "real" option will pass input from the mouse to XIM APEX up to 7ms later than the button press and then pass in the next Response Rate tick to the console, inducing lag.

What this implies is there is zero benefit of reducing the mouse input polling below 1000Hz - XIM APEX will drop input down to the defined Response Rate and output that to the console (which should reduce jitter etc).

Does that make sense?!?

Im barely keeping up on this lol im slow but where does sync play into all this? Example why does it feel super floaty weird if you do say 125hz xim then default? To me low hertz doesnt feel right unless you use off sync if you mix like 125hz xim then sync common or default or anything other then off it feels super odd an bad? But at 1000hz xim then common or sync etc etc feel ok an normal. Wish the devs would explain what is deallt happening with sync an does it have anything to do with input or output

Offline antithesis

  • MVP
  • *
  • Posts: 9875
    • View Profile
    • Mod Squad
Re: 1000hz on pc vs console?
« Reply #17 on: 04:04 AM - 12/08/19 »
Because it's supposed to be one or the other. Sync @ Default, Common & Slow is intended for 1000Hz mouse input.

500, 250 & 125Hz are covered by Sync Off.

Official Australian distributor for XIM APEX, Titan One & Titan Two at Mod Squad
XIM APEX demos on Twitch and YouTube. Follow Twitter for live stream alerts

Offline Bb4life1991

  • MVP
  • *
  • Posts: 716
    • View Profile
  • Gamertag: YT Boogeyman
Re: 1000hz on pc vs console?
« Reply #18 on: 04:07 AM - 12/08/19 »
Because it's supposed to be one or the other. Sync @ Default, Common & Slow is intended for 1000Hz mouse input.

500, 250 & 125Hz are covered by Sync Off.

ahh ok that makes sense

Offline Mb1966

  • Member
  • *
  • Posts: 299
    • View Profile
  • PSN ID: mbooze1966
Re: 1000hz on pc vs console?
« Reply #19 on: 02:09 PM - 12/08/19 »
Thanks for the correction, that does sound like XIM APEX input polling is 1ms and Response Rate is output polling. But I'm not sure where you're getting the XIM-induced lag.

Leaving the mouse at 1000Hz means XIM APEX will detect a button press within 1ms of the input being passed from the mouse to XIM APEX. At 125Hz RR, the button press signal that is output from XIM APEX to the console will be 1ms old at most and will be passed in the next available 8ms tick to the console. So...where's the lag?

Conversely, setting the mouse to 125Hz and XIM APEX to 125Hz is still an 8ms frame, but there may be lag from when the mouse button is pressed and that input signal is passed to XIM, meaning it could be up to 7ms old rather than 1ms. That is mouse-induced lag, not XIM-induced and is the opposite of what you've described.

OBsIV is simply saying that the input signal XIM APEX passes through to the console is at most 1ms old, not up to 7ms, regardless of the Response Rate. So XIM isn't inducing lag at all.

Your "real" option will pass input from the mouse to XIM APEX up to 7ms later than the button press and then pass in the next Response Rate tick to the console, inducing lag.

What this implies is there is zero benefit of reducing the mouse input polling below 1000Hz - XIM APEX will drop input down to the defined Response Rate and output that to the console (which should reduce jitter etc).

Does that make sense?!?

Makes perfect sense.
But would we rather have all of our input values passed through to the console or 1 out of 4?
Even with Xim polling at 1000hz. On both ends, how consistently can any console processor keep up? With all the other processes being handled in games there are bound to be several occasional mistimed polls. In this situation button presses would be “perceived” to us as “lagging” when in actuality it’s Xim throwing away half of our input values and then trying to force the console to accept the input it does have at 1000hz consistently.
Button presses or (input) would be “perceived” as more fluid or consistent if it had time to fill the space between polls with consecutive input events. Instead of for instance one poll returning random input then one or two consecutively nothing. Etc. and so on.
would you agree?
I admit low latency is important but imo “game feel” is even more so...
You put up a great argument antithesis!!
We’re both probably right here. But with you being the most knowledgeable I stand down. Keep schooling us bro there’s a lot to learn here. 
« Last Edit: 02:17 PM - 12/08/19 by Mb1966 »

Offline antithesis

  • MVP
  • *
  • Posts: 9875
    • View Profile
    • Mod Squad
Re: 1000hz on pc vs console?
« Reply #20 on: 03:05 PM - 12/08/19 »
The thing is, 7/8 values at 125Hz aren’t dropped, they’re overwritten with the most current value. So it’s not laggy to pass the most current input value to the console as only the most current value is passed through. The console doesn't know any different because it's getting input at the Response Rate.

Note also that a button press won’t register on a controller if it’s less than about 20ms, so we’re really talking about the flow of input from tick to tick rather than spikes of input. 20ms crosses several ticks at 125Hz, so that button press is not going to be lost. By using the most recent input values, we’re going to feel more connected to the game, at least in terms of button responsiveness.

For context - CronusMAX and Titan One have processing loops of 10ms because J2Kbr (the inventor of both devices) used a PS3 and that’s the polling of a DS3. Similarly, OBsIV used an Xbox 360 which has a 125Hz polling rate, so all prior generations of XIM before APEX ran at 125Hz to sync with controller input. Both of them measured the USB polling rates of XB1 and PS4 and both are capable to 1000Hz USB input, hence the option for us to dial APEX and Titan Two up to 1000Hz.

PS4 handles 1000Hz far better than XB1. PS3 polls USB at 1000Hz, so it’s likely that PS4 does the same but ignores 3 of 4 input values when using a native DS4 at 250Hz. XIM APEX is forcing that back to 1000Hz with no notable ill-effects.

XB1 seems less capable of handling 1000Hz from XIM APEX (for some games and input devices) which is why there's an Audio Compatibility Mode option in XB1 Smart Translators. However, that mode dials XIM APEX all the way back to 125Hz, which kinda defeats its purpose. You can achieve similar results by setting Response Rate to 500Hz.

Long story short - we're already getting the lowest possible latency, which is great for buttons, but not necessarily great for mouse input at 1000Hz on XB1. PS4 is better able to handle the variable stick value input at 1000Hz, but ticks may be dropped on XB1 because of the OS's inability to operate at a consistent 1000Hz, resulting in jitter. 500Hz smooths that out and reduces load on XB1 resources.
« Last Edit: 03:27 PM - 12/08/19 by antithesis »

Official Australian distributor for XIM APEX, Titan One & Titan Two at Mod Squad
XIM APEX demos on Twitch and YouTube. Follow Twitter for live stream alerts

Offline Bb4life1991

  • MVP
  • *
  • Posts: 716
    • View Profile
  • Gamertag: YT Boogeyman
Re: 1000hz on pc vs console?
« Reply #21 on: 11:32 AM - 12/09/19 »
Because it's supposed to be one or the other. Sync @ Default, Common & Slow is intended for 1000Hz mouse input.

500, 250 & 125Hz are covered by Sync Off.

so if i wanted to use say 500hz due to idk if my mouse is consistent at 1000hz common sync would prob feel a lil floaty or weird???

Offline antithesis

  • MVP
  • *
  • Posts: 9875
    • View Profile
    • Mod Squad
Re: 1000hz on pc vs console?
« Reply #22 on: 03:26 PM - 12/09/19 »
500Hz should be fine, but I wouldn't be mixing 125Hz with any Sync mode other than Off.

I currently use 1000Hz mouse polling, 500Hz Response Rate and Sync Default on my configs.

Official Australian distributor for XIM APEX, Titan One & Titan Two at Mod Squad
XIM APEX demos on Twitch and YouTube. Follow Twitter for live stream alerts

Offline Bb4life1991

  • MVP
  • *
  • Posts: 716
    • View Profile
  • Gamertag: YT Boogeyman
Re: 1000hz on pc vs console?
« Reply #23 on: 03:43 PM - 12/09/19 »
500Hz should be fine, but I wouldn't be mixing 125Hz with any Sync mode other than Off.

I currently use 1000Hz mouse polling, 500Hz Response Rate and Sync Default on my configs.

Ahh ok i see. Why dont you like 1000hz? I just cant tell if with 1000hz if im consistently doing it or if its affecting my long range aiming. I can def feel a difference from 500 an 1000hz though as far as responsive goes

Offline antithesis

  • MVP
  • *
  • Posts: 9875
    • View Profile
    • Mod Squad
Re: 1000hz on pc vs console?
« Reply #24 on: 04:13 PM - 12/09/19 »
I run with 500Hz RR because I typically use Sync Off in games and 1000Hz runs a higher risk of jitter. But I've gone with Sync > Default lately, so there's really no reason to run 500Hz.

Official Australian distributor for XIM APEX, Titan One & Titan Two at Mod Squad
XIM APEX demos on Twitch and YouTube. Follow Twitter for live stream alerts


Admin