Author Topic: Theoretical question about input from two usb sources  (Read 2424 times)

0 Members and 1 Guest are viewing this topic.

Offline IOVERCALLHISTIOCYTES

  • Thread Starter
  • Posts: 1958
Theoretical question about input from two usb sources
« on: Fri, 11 October 2019, 18:25:21 »
Lets imagine a split keyboard where each half does not talk to each other, but instead has its own usb connection to the computer. What's the expected lag from the different sources?

When I numpad with one device, and use nav cluster with another, I'm reasonably quick and don't notice any problems.

Offline Venaros

  • Posts: 84
  • Location: Ontario
Re: Theoretical question about input from two usb sources
« Reply #1 on: Mon, 14 October 2019, 17:25:38 »
Don't think there is going to be any lag unless your computer is a literal potato.

Offline Steezus

  • Keeper of Facts
  • * Elevated Elder
  • Posts: 2497
Re: Theoretical question about input from two usb sources
« Reply #2 on: Mon, 14 October 2019, 17:56:32 »
I am currently using a TKL board along with a separate Numpad, each with its own connection. After a simple test of typing on both simultaneously I did not experience any input lag, every keypress was registered as well.
TGR-Jane CE | TGR-Tris CE | Lyn Montage | LZ PhysiX | Exclusive e8.5

Offline algernon

  • Posts: 311
  • A tiny mouse, a hacker.
    • Diaries of a Madman
Re: Theoretical question about input from two usb sources
« Reply #3 on: Tue, 15 October 2019, 02:04:13 »
A split board where the halves don't talk to each other... lag (of which there won't be any) will be the least of your worries. Modifiers - at least on some operating systems - are per-device, so pressing Shift or Control on one half will not affect the other half on those. That's kind of a pain. If you ever want to use some of the cool stuff QMK & co have like OneShots, such a split board will not work. If you ever want synced RGB animations or such, likewise.

Having a separate numpad is *much* different than having a split that doesn't talk to the other half.

Offline yui

  • Posts: 1082
  • Location: 127.0.0.1 (in azerty)
Re: Theoretical question about input from two usb sources
« Reply #4 on: Tue, 15 October 2019, 03:08:54 »
The USB protocol should be plenty fast enough to support more than 2 keyboard while staying well into the real time realm for humans(no perceivable latency).
A normal keyboard polls at 125Hz , from some quick googling (might be wrong) the maximum number of hid keyboards on a USB1 connection would be (5000000/*5Mb/s*/ / 128 /*bits per transaction*/ ) / 125 /*polling rate (transaction per seconds)*/ = 312.5 keyboards per USB host (use C syntax for comments).
But if i recall correctly you can only cascade up to 4 levels of hub (4 ports each) so it would be 256 keyboards.
And USB2 (the one on most every oldish PC) is 100Mb/s and USB3 is > 1GB/s (to many versions, too lazy to search it is way too high anyhow).

So the quick version is no you should never get latency from using 2 keyboards (halves or not) on USB (in fact most multimedia and gaming keyboards act as 2 or more keyboards on the usb).
vi vi vi - the roman number of the beast (Plan9 fortune)

Offline IOVERCALLHISTIOCYTES

  • Thread Starter
  • Posts: 1958
Re: Theoretical question about input from two usb sources
« Reply #5 on: Tue, 15 October 2019, 22:53:32 »
A split board where the halves don't talk to each other... lag (of which there won't be any) will be the least of your worries. Modifiers - at least on some operating systems - are per-device, so pressing Shift or Control on one half will not affect the other half on those. That's kind of a pain. If you ever want to use some of the cool stuff QMK & co have like OneShots, such a split board will not work. If you ever want synced RGB animations or such, likewise.

Having a separate numpad is *much* different than having a split that doesn't talk to the other half.

This was the next problem I was gonna get to. Saw that Karabiner worked for OS X but would need so,etching windows. Currently use a split board, a numpad, and a macropad. Had hoped to get ...roughly a skidata worth of keys across a split board which would be my most common macros (typically they’re sentences or paragraphs).

Offline mizzoperator

  • Posts: 66
  • Location: Hot Topic
  • "This can't be good for me, but I feel great!"
    • Milk Time: The Mizz Domain
Re: Theoretical question about input from two usb sources
« Reply #6 on: Thu, 17 October 2019, 08:18:48 »
That's an awful, awful idea (having a split keyboard with two separate USB cords), but I wouldn't think there would be any input lag. A separate numpad is completely different to this concept you have generated.
Linears are for linear people. No offense if you use linears.
I prefer tactile switches, I'm reluctantly using the AULA SI-859 and my pronouns are she/her.