Here's an improved version.
As well as some visual refinements it introduces an additional technique, using OSC cues with durations to fire Script cues continuously (with the actual frequency of script cue execution controlled by post waits on the script cues) to handle text formatting and inclusion of system variables like the time that OSC cues can't do themselves.
Although I haven't included it in the example, as it makes it trickier to use on different interfaces, in reality you would probably want your audience to listen to a full mix of the track instead of 3 stacked filtered versions. In this case you would probably want to send your filtered audio to a high numbered unused cue output e.g 64. If you didn't have an actual output 64 on your device (if you are using built in audio for instance) you can create a virtually silent one by assigning output 64 to output 1 in the edit patch routing matrix at a level of -59.5dB. This will mean you can have full audio output on the filtered cues Ch 64 without it being audible
The first OSC argument after /liveAverageLevel/ would then be 64 e.g;/
/cue/LEFT/liveTranslationY #/cue/LO/liveAverageLevel/64 -1300 0 #
I'll put it on the QLab Cookbook, with a full explanation of how it works, sometime in the next few weeks.
Mic