This is a mystery signaling standard that I'm trying to sort out. It looks more or less like RS-485 to me, but I don't recognize the long looking start bit, and the "idle" state of the line looks a little strange to me. Using an RS-485 to RS-232 converter, I can't seem to get it to translate to anything intelligible. Note the different offsets for CH1 and CH2. When in the "idle" state (at the beginning) the two lines are at essentially the same potential, and when transmitting, they differ by the amounts show.
I guess in particular, I'm more curious about the framing than I am about the physical layer itself, since my rs-485 converter seems to interpret it OK (even if the data appears to be nonesense to me)
Here's the character of the data over a longer time span, note the timebase (same vertical settings)
Here's the close-up trace, but with infinite persistence turned on. I'm triggering on the long start pulse, which seems to be present for all messages: