I still have not scrutinized Shirriff's library. But it is getting closer... Funny is that the numbers start with a gap (are < 0), which is normallly considered nonsensical. And the final number should probably be a gap (negative) and is for some reason given as hexadecimal, while all the other numbers appears to be digital. Discarding the initial gap (it is of course meaningless anyhow, what should be meant by an IR signal starting with silence?), changing the last to any "large" negative number, it can be decoded by IrMaster (paste into the Generate & Analyze window, add a "+" in front of the first number (2400 or 2500, not the negative number, fix the last) press Decode). Or, unfortunatelly, the first two can be decoded, the last two do not decode. The first two decode as protocol = Sony15, device = 100, obc = 44 and protocol = Sony15, device = 100, obc = 49. The rest can be considered as "unclean learns". Please retry these.