Closed mattheww closed 9 months ago
The above figures are for display::nonblocking
.
display::blocking
defaults to a 2ms refresh per row, giving 6ms whole-display refresh for the micro:bit v1 and 10ms whole-display refresh for the micro:bit v2.
I think the pragmatic thing is to double the frequency for the micro:bit v2, rather than calculating a whole new GREYSCALE_TIMINGS
table and teaching tiny-led-matrix
to use it.
The current display code is showing each (internal) row of LEDs for 6ms.
For the micro:bit v1, where there are three rows of LEDs internally, that means the whole display refreshes every 18ms.
This is the same timing that
microbit-dal
uses for the micro:bit v1.But the micro:bit v2 has five rows internally, which means the whole display refreshes every 30ms. That looks visibly flickery to me.
codal-microbit-v2
uses 3⅓ms per row. I think we should adopt those timings.