I too was suprised how low the duty cycle is, but I am sure the test circuit was not distorting the waveforms. I was capturing clean pulse with rise and fall times in the region of 200ns. I used a BPW41N high speed IR PiN diode as part of my test circuit.

I do have the entire bit sequence as raw sampled data, all 2Megs worth at 5MS/s. I am working on some software which will convert the raw sample data into the 115kbaud bit pattern and then to Pronto Hex format as I certainly do not have the patience to do it by hand.

Unfortunately, I no longer have access to a digital storage scope, as I am in the process of changing jobs and left my old company on Thursday, hence I can no longer borrow their £30k scope. Hopefully I will be able to borrow another scope soon.

If someone has a scope, a pronto, and some patience I would be grateful if they could see if the pronto can cleanly generate arbitrary bit patterns of the timings I previously discussed.

In the meantime, I will have a look at your MakeHex utility. Hopefully all this effort will be worth it in the end. At this stage I am just trying to determine if the Pronto is capable of generating the high speed bit patterns required, before I start trying to convert the 80Megs worth of sample data I have into Hex codes.