I can shed 'some' light on what panasonic sent. You may already know most of this.


Word | code (hex) | comment

1 0000 learned/raw data
2 0070 carrier freq (36.68 khz) (cycle timing = 27.26us)
3 0000 first burst pair sequence length (zero in this case)
4 0032 second burst pair sequence length (50 burst pair)
including the lead-in and lead-out
5,6 0080 0040 8T 4T, the header for this device, 80h = 128d,
128/36682 = .003489 seconds or 3489us or about 8T. The
calculations for 4T are the same yeilding about 1745us.
7-70 determines the group number, each pair of words determines
a '1' or '0'. 0010 0030 pair is '1' and 0010 0010 pair is '0'.
That's the data 0 = <-T-> <-T->, 1 = <-T-> <---3T---> .
Every 4 bits makes a nybble and is read as such. 64 words
equal 32 bits. The 32 bits are read a nybble at a time.
The group number is 4004 0100 for power (on, off, toggle)
71-86 First byte of button code, 1st nybble of 1st byte always
repeated in 1st nybble of 2nd byte (don't know why)
87-102 Second byte of button code
103,104 lead-out, should be about 76,000us

That's about all I can tell you about what they sent.