I'm aware of LIRC - linux infrared remote control project
Looking at the LIRC archive there seems to be specious accuracy in the figures, AFAIK 10 percent timing error-band is about the norm for domestic r/c (not complex ones) so they can use ceramic resonators rather than quartz. eg LIRC gap 46,229 uSec for their particular one tested, scope trace of order 45uS +/- 1 uSec below , I'm assuming is the complete block repeat timing.
Starting from a known quantity for a JVC r/c, I have sitting around, can anyone cross-correlate the LIRC data to oscillscope trace view. From LIRC archive for r/c name JVC_PQ21953F
LIRC bits 16 flags SPACE_ENC|NO_HEAD_REP|CONST_LENGTH eps 30 aeps 100
header 8488 4126 one 590 1508 zero 590 462 ptrail 590 gap 46229 min_repeat 2 toggle_bit 0
begin codes # setting A-VCR operate 0x000000000000C2D0
for the "operate" button only , for simplicity here
SCOPE
From scope simple timings/accuracy without using a frequency counter , but could do so if more accuracy is required carrier frequency 40KHz , 25uS period full code repeats after 45uS code pulses of order 0.5mS to 0.6mS duration ident code at 0,2,4,5,6,7,8 mS ie 101011111 (same for all button pushes) "operate" code starts at 10mS and code reads
1101011011111 as on the screen ie first on the left (varies on button pushed ) I could not find the meaning of eps , aeps, ptrail etc Whatever way of, reading/inverting/adding unseen zeros, the scope-read "operate" pattern, I cannot get hex code C2D0