Issue link: http://digital.copcomm.com/i/1122868
68 S P R I N G 2 0 1 9 C A S Q U A R T E R L Y recorded in the same day, a naming convention for files can include episode name. How many of your programs approximately are requiring theatrical 5.1/2.0 mixes, as well as near-field 5.1/2.0 mixes? What is the percentage that are requiring Atmos deliver at this time? For original features and indie features, theatrical 5.1 is on the deliverables list. These are used only for theatrical DCP screenings. We prefer that crews work on the near-field mix first and adapt a theatrical mix later. We understand that this is not the traditional way, but in the vast majority of cases, the theatrical mix will be heard very little. We prefer that mixers work on the near field as the "hero" deliverable—simply because that's what the audience will hear. We request Atmos for many features and series. For other content categories (indie features, international originals, documentary, comedy, unscripted, licensed), we ask for Atmos when it makes sense for the creatives and there is budget to pay for it. What was your intention in using ITU-R BS.1770-1 versus ITU-R BS.1770-3 and to deviate from the broadcast standard of -24 (+/-2) LKFS with a loudness recommendation of -27 (+/-2)? 1770-2, -3, and -4 nearly always include full program-gating. We've found that measuring content based on the full program results in average dialogue levels which vary widely from title to title. Loud action sequences in one episode will effectively push the dialogue down, while another episode without action will have louder dialogue—all to hit the full program spec. We moved to ITU-R BS.1770-1 dialogue-gated because it does not normally include a program gate, which would add too many variables to readings. We chose -27 (+/-) 2 LU simply because that's where our content library clustered—we wanted to change what is being measured but not change how things are mixed. Because most dialogue sits at -27 for our content, it caused the least disruption for those delivering to us. We joined NBCU, HBO, and CBS in moving to dialogue-gated measurement. Disney Home Entertainment now also uses -27 LKFS +/-2 LU dialogue-gated. To learn more about the consequences of algorithms to measure audio program loudness on your mix, check out Jon Greasley CAS and Greg King CAS's article in the CAS Quarterly Spring 2018 edition entitled "Stay Calm and Carry on Mixing." We like very much that you specifically describe the desired fold-down process. However, your mention of a possible optional Low Frequency Effects (LFE) channel at -12 dB is unusual. Why this reduction? The ITU BS.775-1 fold-down spec drops LFE, and some engineers feel LFE content can create distortion in stereo mixes. On the other hand, most mixers I've come across keep LFE between -8 to -12. -12 was a compromise number given the concern about distortion. We may instead define a range in a future revision. Regarding Atmos deliverables, what's the purpose for requiring delivery of the recorder stems, Atmos DAMF file, and ADM? We removed the stem requirement with our v2.0 Atmos spec. The BWAV ADM is our primary streaming deliverable, and we request the DAMF for archival because it's the primary master file generated by the Dolby Renderer. Should folks use the https://help.prodicle. com/hc/en-us/categories/115000232072- Technical-Specifications or https:// partnerhelp.netflixstudios.com/hc/en-us/ categories/202282037-SPECIFICATIONS- GUIDES site for gathering info? Prodicle Help Center (help.prodicle.com) is the primary source of truth for Netflix documentation and specifications. Click the link for Production and Post Technical Specifications and Guidelines.