New Posts  All Forums:

Posts by TranquilTempest

Variable refresh needs hardware that did not exist before g-sync. It doesn't have to be proprietary, but it does have to be designed around variable refresh. There is a ton of misinformation floating around on this, but anyone that tells you it's "just a firmware update" is lying or misinformed. That includes eDP based displays.I'm willing to admit I'm wrong if someone publishes firmware that turns a pre g-sync product into an adaptive sync or g-sync display.
When I hear "gaming keyboard" I think low latency and NKRO. When a marketing person hears "gaming keyboard" they think MORE LIGHTS AND BUTTONS!
Take a look at this: http://www.ganssle.com/debouncing.htm
I imagine not quite as awkward as the 2KRO.Yeah, latency is cumulative, so it doesn't really matter if some other part of the system is slower than the keyboard, you still benefit from a lower input latency. The way I see it, it's a lot cheaper to cut down the input latency of the keyboard by 10ms than making the same improvement somewhere else, like the display. Granted, decreasing latency in the graphics chain decreases apparent mouse and network latency as well.Bottom...
The 1000hz polling rate is just what the WASD has set in the firmware, so you don't need special drivers for that, and curiously, ~70% of keypresses were in the 9ms to 10ms range. I can't really say anything useful about laptop keyboards because I don't have one, and the people I could borrow a laptop from probably wouldn't appreciate me taking it apart. There are a few options: use a normal keyboard MCU and hook it on a standard interface(USB, PS/2, etc.), use one MCU...
http://forums.blurbusters.com/viewtopic.php?f=10&t=1836&start=10
I'm more excited for QD LED matrix displays.
Here's the video: https://vine.co/v/OjqeYWWpVWK
Dunno, probably 5~10 bucks/chip instead of 1~2, not that it's necessary if you're willing to put some more time designing the PCB and sampling scheme. Lets say you want up to 144 keys, with NKRO and hardware debouncing, using straightforward matrix design. You'd probably spend ~3bucks on diodes to get NKRO, a couple bucks on caps resistors and schmitt triggers to get the hardware debouncing, a couple bucks on a microcontroller, and a big wad of cash on switches. You pick...
yes, the interrupt will trigger many times in that 5ms, but after the very first one you can tell the computer the button has been pressed, and you either poll 5ms later to tell if the button has then been released, or you measure how long a bounce can take, and if the switch is off for at least that amount of time, you know the switch isn't just bouncing, it's actually been released. OTOH, If you use an instantaneous poll to sample the keypress, it's somewhat random when...
New Posts  All Forums: