Agreed on crual missing detailed engineering about rms calculation
As far as I have understood DC shares the hardware buffer space used for data transfer to proceed on the fly calculations, thus integration becomes highly depending of its size and latency expected
Note that if Totalmix and DC are set on screen together, the rms metering in both is wrong as notified in the help, and shall not be compared with external metering in your daw for instance
I presume that the integration processing is of some kind of secret recipe, but probably no attack or decay nor overlapping of data here, simply a flat windowing against time integration and large averaging to output rms and slow rms
RME to help?
However, I have checked the accuracy of the spectrum analyser compared to 1st class 1/3 octave bank filters, and the result is of both very good match, thus I trust the rms metering features to be relevant for the purpose as well:-)
HDSP9632, HDSP9652, ADI-8QS, FFUC, Quadmic II