For real though, the more I try other keyboards, the more I like this Model01.

<3

corsiar k95 key scan rate imagebin.ca/v/3zi1JDVKRoDa

250hz

that site is incredibly slow/unreliable from work... it still didn't load any image :/

Was there an intentional change a while back that increased the iteration rate of the LED animations; in particular, LEDRainbowWaveEffectt (though I think all the ones that work like that one got faster)?

I am trying to write a doc today, so I'm pondering things like that instead of working on the doc, though I did notice it a monrth or three ago. :)

algernon: yeah it's not working for me either

PugMajere: we drastically sped up the main loop, which may have affected LED animations

PugMajere: there were a few that did not use timers, but relied on the speed of the main loop - those got faster. Rainbow may be one of these

ah, that's good to know.

yeah, just looked at the code, Rainbow does animations based on the amount of loop cycles, not based on timers

the faster the loop, the faster the animation

yeah; do you know offhand how many milis the main loop is? (approx); I can work it out experimentally, but it'd be easier to get close without guessing. :)

depends a lot on what plugins you use, and how many things are going on with the keyboard.

with a recent change, the main loop is considerably faster when there are no keys pressed

so it can even vary during runtime, and as such, should not be relied on for timing

there's the CycleTimeReport plugin (github.com/keyboardio/Kaleidoscope-CycleTimeReport) that can help you figure out the length of your loops

hrm. I didn't record my timings in the PRs, only on IRC :/

(and too lazy to search back right now)

PugMajere: with my own sketch, cycle times are around 3.8ms when idle, 4.02 with 10 keys held. the factory firmware is probably a little faster.

heh, yeah, I was just trying to get an idea of how many milis (approx) a cycle was, so I could multiply and slow it down.