Alright, so this one is for my fellow TV-post fader jockeys out there. I've never really found a solid answer to this, not even from some of the networks we sell to (I wish we actually sold more to Discovery and NatGeo).
We're mixing to 5.0/5.1 for Dolby-E encoding as a delivery medium. For that specific final mix, we're using an internal spec of -24db LKFS/LUFS (infinite) with a maximum true peak of -9dB. You probably know that an E-encode includes an LtRt version of the full mix in channels 7 & 8 of the bitstream. The question is, should we be worrying about meeting the same spec for that LtRt like we do with the full surround mix? We've also taken to simply encoding a LoRo for a Mix-Minus when we're only providing stereo masters to a client network: complete main LtRt on 1&2, Mix-Minus LoRo on 3&4. So, I have concerns about the specs of these fold-downs beyond their inclusion in the Dolby-E bitstream.
In our current process, the -9dBTP goes out the window after LtRt encoding. It doesn't hit anywhere near zero (usually only going up to -8 or -7dB true peak), but it's still over. I haven't had time lately to run a full analysis of the LtRt to determine its "dialnorm," but I'm fairly confident that it's not sitting at -24. [We're going to need a new term for that in the future, I think.]
So, anyone out there got a definitive answer...idle speculation...?