Most remotes read in the raw timing data and classify burst pairs. So even though the exact duration of the on/off pulses may vary by 10% or so, the durations are all set to a single value. In the case of PPENG, I suspect that it also recognizes the code as F12, and tries to "clean up" the timings. But the primary change that it made to the Pronto Hex that I posted was in the delay between repeated signals, and that shouldn't have kept the blinds from responding, at least briefly. So let's assume that we need something different. Let's try the exact F12 specification, as close as can be described by Pronto Hex: Open Device Code: 0.1 Function: 8 0000 006D 0000 0018 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0010 0530 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0010 0D30 Close Device Code: 0.1 Function: 16 0000 006D 0000 0018 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0530 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0D30 Stop Device Code: 0.1 Function: 4 0000 006D 0000 0018 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0530 0010 0030 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0030 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0D30