Issue link: https://digital.copcomm.com/i/1500716
62 S U M M E R 2 0 2 3 I C A S Q U A R T E R L Y While we're defining terms, it's important to define what "anchored," "full program," and "integrated" mean, as there's some confusion here, too. I often hear people using "full program" and "integrated" interchangeably, which is incorrect. Integrated simply means "over the full length of the program," so everything has both integrated dialogue and integrated full program measurements. The important distinction is between "dialogue anchored" and "full program," which are also sometimes referred to as "dialogue-gated" and "relative-gated." Just wait. It gets more complicated. There is a wide misconception among content distributors (TV and cable networks mainly, but also the streamers) and QC houses that CALM compliance requires an integrated full program measurement. This, categorically is not what A/85 recommends as that type of measurement is specifically reserved for commercial advertising content (the "CA" in "CALM"). To quote A/85 with regards to commercials; "[M]easurement of the long-term integrated (or average) loudness of the content's full program mix over its duration should be reported as the Dialogue Level of the short-form content." This approach was specifically designed to reign in the volume of commercial mixes, which would use tricks to increase the apparent mix level in order to get the viewer's attention. As obnoxious as it is, it makes sense in as much as commercials are designed to do specifically that. But it got so bad that it garnered the attention of a congressperson, and they started drafting legislation with the guidance of the Advanced Television Systems Committee (ATSC) to regulate it. BS.1770-1 contained a dialogue detection system called "dialogue intelligence" (or dialogue gating) which measures the level of dialogue when detected, along with a simultaneous full program measurement that measures the whole mix. This was simple and clear and, at first, people seemed to understand it. However, commercial mixers realized they could game the algorithm by inserting a second or so of silence into the mix (which doesn't sound like much but if it's a 10-second spot, that's 10% of your mix). This could trick the measurement to read lower than the actual value of the portion that has audio present. To correct for this, a revision was made to 1770 that added a level activated-gate (also referred to as a "relative