I read recently on here how to extend the duration of an IR code by altering certain parts of the Hex.
This was the gist of it for those that didn't see it:
Example: Panasonic Viera Power ON code.

0000 0070 0000 0032 0081 0040 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 0010 0010 0010 0010 0010 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 0010 0010 0010 0030 0010 0AC3

The first four ‘words’ are what is known as the ‘Preamble’, The Fourth Hex ‘word’ is the one that dictates the ‘repeat’ or in other words the ‘burst pairs’, the words from the fifth onwards are the actual code and contain those 50 repeats, to extend this to 100 repeats adjust the fourth word to show Hex 64 (decimal 100) then copy and paste the fifth word onwards (the Red section) and add them onto the end of the existing code.

Hopefully this info is correct, So based on the above, is there a method to 'shorten' an IR code, again by altering it's Hex?
This is the code I'd like to shorten to the Shortest possible IR output.

0000 0073 0000 0015 0060 0020 0010 0010 0010 0010 0010 0020 0010 0020 0020 0010 0010 0010 0010 0010 0010 0010 0010 0010 0020 0020 0020 0010 0010 0010 0010 0020 0010 0010 0010 0010 0020 0020 0020 0010 0010 0020 0010 0010 0020 0B61

Sorry for the long post, but I just wanted to be clear in the first instance.

The 3rd quad of the preamble is the number of following burst pairs which are to be sent once, and the 4th quad is the number of burst pairs to be repeated. Both of your sample signals have zero "send once" burst pairs, but with different IR signals, you may need to take that into account.

Anyway, you can make the notation send fewer burst pairs, but presumably you'd like the IR signal to be correctly recognized.  Your second example decodes to RC-6-20, and has the minimum number of burst pairs possible with that IR protocol. The way I decided that was to use IRScope, import the Pronto Hex, look at the decode, and compare it to the IRP description of the signal.

However, you can use IRScope more easily to see if a given Pronto Hex can be shortened.  Take your first signal, modified to send 64h burst pairs.  If you import this into IRScope (specifying 1 repeat), it shows
Panasonic 128.0  62   + 1 copy

The "+ 1 copy"  tells you that this is a repeated signal.  If you now export this decoded signal using the "export using decoded signal" option, it will be shortened to the right length.

Conversely, the second example (the RC6-6-20 IR signal) has the notation 
"no repeat".  So it can't be shortened.  The only exceptions to this are IR signals that use different toggle values for variously the first, or repeated, or end signal of a group of the same IR signals.  These signals can frequently be shortened to just send the first signal of the group, and to skip the following ones.  XMP is usually like that.

many thanks for the reply.
I was hoping the RC6 code be shortened because the issue I'm having is that my Source product is responding as if the command were an extended code.
I think it's Hardware related, I'm going back Tuesday to try some other fixes.

Just so we're on the same page, the Pronto Hex you posted is not RC6.

RC6, when referred to in full,  would be RC6-0-16.  It has 16 bits of data, usually interpreted as 8 bits of Device and 8 bits of Function.

Your post had a signal which is RC6-6-20.  It has 20 bits of data, usually interpreted as 8 bits of Device and 8 bits of Function, and 4 bits of Subdevice.

So depending on what you mean by "extended", maybe that's what it is supposed to do.  Or maybe it is an incorrect learn.  Do you have any other signals from the same unit that we can decode to ee if the have the same IR protocol?.

BTW, RC6-6-20 is very uncommon.

Hi, thanks for the clarification, that particular code was generated using Makehex, the Original learnt code was this:

0000 0072 002A 0000 0061 0021 0010 0010 0010 0010 0010 0020 0010 0020 0020 0010 0010 0010 0010 0010 0010 0010 0010 0010 0020 0020 0020 0010 0010 0010 0010 0020 0010 0010 0010 0010 0020 0020 0020 0010 0010 0020 0010 0010 0020 1206 0061 0021 0010 0010 0010 0010 0010 0020 0010 0020 0020 0010 0010 0010 0010 0010 0010 0010 0010 0010 0020 0020 0020 0010 0010 0010 0010 0020 0010 0010 0010 0010 0020 0020 0020 0010 0010 0020 0010 0010 0020 1206

and another one of the same command

0000 0072 002A 0000 0060 0020 0010 0010 0010 0010 0010 0020 0010 0020 0020 0010 0010 0010 0010 0010 0010 0010 0010 0010 0020 0020 0020 0010 0010 0010 0010 0020 0010 0010 0010 0010 0020 0020 0020 0010 0010 0020 0010 0010 0020 119A 0060 0020 0010 0010 0010 0010 0010 0020 0010 0020 0020 0010 0010 0010 0010 0010 0010 0010 0010 0010 0020 0020 0020 0010 0010 0010 0010 0020 0010 0010 0010 0010 0020 0020 0020 0010 0010 0020 0010 0010 0020 0020

I used IRTool to decode this and then generated the 'clean' code, both codes work however.

I think those look like very clean learns.  For my own curosity, what kind of equipment is using this RC6-6-20 IR protocol? 

They are UK SKY HD codes, It's a Satellite PVR, quite popular over here.
Some Remotes handle the codes really well, others not so much it seems.