I combined all the traces and used the signal on line 0 to sync them all up to a common time signal (line 0 in retrospect wasn’t the best choice as it was not a back plane signal (more on that to follow) but at least it allowed me to align all the transitions and get everything roughly in order.
I then plotted all the traces together for inspection. The traces are pin 15 at the top down to pin 0 at the bottom giving a total of 16 traces.
Observation shows the top four lines to be the same signal albeit it displaced by 1/4 of its cycle; i.e. the signal has eight distinct steps with each of the four signals being advanced by two steps from the previous one. These are the common back plane signals that run through the LCD panel. (As previously noted the signals are also themselves multilevel with a baseline of zero volts and then three levels of 1/3, 2/3 and 3/3 of the bias voltage).
The other 12 signals are presumably segment lines. The actual drive to each segment of the panel is provided by one back plane and one segment line. The drive itself is the difference between these two signals and is a pure AC voltage with zero DC bias so as to not damage the crystal structure of the display, we will work out the voltages next.