Issue link: http://digital.copcomm.com/i/987065
C A S Q U A R T E R L Y S P R I N G 2 0 1 8 73 but it's only for short-form (commercials). For long-form content, you use the anchor element—and the anchor element is the dialogue. It is still a 1770-based measurement, but only measures the loudness of the dialogue/speech of the program." Scott Kramer from Netflix is in the unique position of being privy to almost immediate end-user feedback about the content they provide. Their platform allows consumers to submit their options and experiences at any time, directly to the service. Scott Kramer, Netflix: "As far as how we arrived at our current spec, we would seek to determine what the industry standard is and then conform to that. When we didn't feel bold enough to look at our needs and possibly forge a new path, we wanted to do something that was in line with what the larger industry was doing, and that's where the 1770-3 full program measurement came from. I've heard from a lot of mixers, and I know it myself from having been a mixer, but mixers generally prefer a dialogue- weighted measurement. So we're rethinking whether the full program measurement works for us, and we're finding that it doesn't: full program just doesn't work as well as a dialogue-weighted measurement. We've found that our main customer complaint, just like they are from every broadcaster, is around dialogue intelligibility. Therefore, the goal has to be consistent dialogue across the service. You've got to go from one show to another, including features, and have the dialogue be consistent. So that's why we're planning on moving to dialogue measurement and correcting and normalizing to dialogue. "I do understand the intention behind 1770-3 and the difficulty of working with commercial breaks but, in my talks with engineers at other studios and broadcasters, it doesn't work for long-form." Closing Thoughts: There is a definite thread to the responses we've gotten to our questions here. The 1770-3 (full program mix) loudness measurement does not work for long-form in general, and the reason may be because it was not fully tested using content with wider dynamics. Very dynamic content, such as premium episodic, does not measure appropriately with 1770-3, and this is the reason why dialogue/anchor-based loudness measurements are recommended. One of the main additions to 1770-3 is that a gate was added to remove the ability to "trick" the measurement by inserting small sections of silence in the spot, in order to skew down the LKFS measurement. Dolby's Scott Norcross recalls, "I remember back in 2013 when we updated A/85 to -3, it was for commercials explicitly." A/85 recommends that long-form content still be measured with a dialogue anchor to ensure consistency amongst all types of long-form content. To be absolutely clear, long-form programs are supposed to be mixed and measured with a dialogue anchor, not a full program measurement. 1770-1 works well because dialogue is the focus; 1770-2 and 1770-3 will work but are not necessary once dialogue is established as the anchor and focus of the measurement. (A/85 2013 has a footnote grandfathering -1 in this application.) Long-form content is not to be measured full program, or even the entire length of the show, and certainly not act-by-act. The end result of this misapplication is that we have content on any given television or cable channel, streaming service provider, or even a box set of DVDs or Blu-rays that can have up to a 10 dB swing in apparent loudness. That's between channels, between services, or even between episodes of the same show on a single service. Given that the stated goal was to unify and normalize the user experience, it seems quite apparent that missteps have been made. But it's not too late! If we can all agree on the best implementation of the standard, which from our discussions here it truly seems that we can, then we should be able to stop, re-calibrate, listen, and then move forward with a delivery spec across all platforms that allows the shows to be presented in a consistent, dynamic, sonically rewarding fashion that satisfies the creative intent of the filmmakers while simultaneously meeting the legal requirements of the CALM Act and the recommendations laid out in A/85. As crazy as we might be, that doesn't seem crazy to us—and we encourage the discussion, participation, and engagement of our collective sound community in order to make this a reality. Please feel free to contact us at jon@kingsoundworks .com and/or greg@kingsoundworks.com with any thoughts on this matter, alongside Scott Kramer (skramer@netflix.com) at Netflix, as we endeavor to arrive at a unified spec. Additionally, we're striving to put together an event with the ATSC, Dolby and as many content creators and providers as we can rally in order to discuss the current situation, enlighten stakeholders, and strive to achieve the intended outcome as intended. •