Issue link: http://digital.copcomm.com/i/1122868
C A S Q U A R T E R L Y S P R I N G 2 0 1 9 67 What do we need to know about the new higher quality audio feature? Following feedback from the Stranger Things creative team, we learned there was room for improvement on our sound quality. We put together an engineering team which worked for over a year to add adaptive streaming to our audio. A bitrate ladder allows quality to rise and fall as network bandwidth requires. Most TV devices are capable of receiving improved audio. Depending on your device and bandwidth capabilities, the bitrate you receive may vary: • ● 5.1: From 192 kbps (good) up to 640 kbps (great/transparent) • ● Dolby Atmos: From 448 kbps up to 768 kbps The goal of this project is all about protecting the creative intent of the mixes, and sending a studio- quality experience to our members. We expect these bitrates to evolve over time as we get more efficient with our encoding techniques. Improvements to AAC stereo streams (mobile, PC) will come later this year. Your Production Sound: Recommended Practices requests that "Script character names and microphones should be included in the metadata for each ISO mic." Are you requesting model numbers of mics, or something as general as boom or iso and character name? How necessary is this requirement? We publish both specifications and recommended practices. Freedom to work however crews want is very important to us, and we only establish rules (specs) when we have to. The Production Sound: Recommended Practices document is not required and is not enforced. It's very important to us that we offer helpful advice, but don't tell people how to work. Also, every Netflix production holds production and post kickoff calls to discuss workflow and find something that works best for the particular crew. Anything determined there would supersede our documentation if it differs. We prefer to have the make and model of mic listed, or just the model—everyone knows what an MKH 60 is, so Sennheiser is not necessary. Other production mixers, ADR mixers, dialogue editors, and re-recording mixers should find the extra information helpful, so they don't have to guess when trying to match dialogue. In this same document, you ask that "Each file will be named for its corresponding scene and take, as well as included in the metadata for each file." Depending on the model of audio recorder used, it's usually an either-or situation. For instance, Sound Devices actually titles the filename: SR002- 48B-011 for Scene 48B on sound roll 002. Meanwhile, Zaxcom will name the file with the folder name and the file number. For instance, SR002005 would be the fifth file in the SR002 folder. Then metadata extraction software applied later usually replaces the filenames with scene and take information from the metadata. Are you asking for filenames and metadata to be redundant? If so, have you found benefit in this method in assembling dailies later? It is not a specification, but a recommendation. The spirit of this request is to have scene and take clearly labeled for picture assistants and sound editors. There's consensus that scene and take in the metadata is essential for automatic tools in the post process to marry and assemble audio. We've found that redundancy in filenames and metadata doesn't hurt when the recorders are capable of it. There is a request for series projects to use different folders for each episode and to avoid duplicate file names. If scenes are recorded on the same date due to location availability, etc., do you still wish for separate folders to be used? Duplicate date folders with varying names may complicate assembly later in post. What is your recommendation in cases such as this? One folder per day is the overwhelming industry standard and seems to be most convenient for all involved. If scenes from different episodes are