Dolby ATMOS Mix Delivery Best Practice Resources?

[JKHansen]

I'm wondering if anyone can point me to some authoritative resources on best practices for the delivery of ATMOS / Binaural mixes to various distribution services. I'd also be interested in reading more technical information on ATMOS and the re-render process in general.

I have one specific sub-topic that is of a more pressing interest - binaural render/delivery/decoding. This relates to the desire to have an ATMOS mix played back with binaural encoding on the user end.

Specifically, when delivering an ATMOS mix to streaming services, how is that mix handled in delivery? Do they rely on the ATMOS decoding of the consumer's playback system when it sees a two channel playback setup or do the streaming services process the ATMOS file locally before delivery? If the former, presumably the playback system will fold down the mix into standard stereo and not perform any binaural encoding.

The context is a desire to deliver music and narrative podcasts in a format that would be played back with binaural encoding when reproduced on headphones. Does this require the re-render and delivery of a separate binaural file from the renderer and, if so, can I expect that the file will be labeled accordingly by the streaming services.

I'm a dedicated RTFM geek and am happy to consume anything from summaries to technical documentation so any authoritative resources would be greatly appreciated.

[its2loud]

The binaural output of the renderer is solely for your own monitoring purposes to setup the binaural settings in the rendere for how you want that mix generated once the full ADM file is delivered. The typical delivery of a full Atmos mix is to deliver the ADM Wav file to the vendor. The format it’s delivered to will use the binaural metadata to generate the binaural mix you specified. UNLESS you are delivering to Apple. They use their own algorithms to generate the binaural mix for Apple Music. Nothing you set in the binaural settings of the renderer will apply to Apple Music.

[JKHansen]

I understand the binaural monitoring option is obviously for monitoring purposes. However, the ATMOS renderer does have a re-render option to output a binaural encoded stereo file. Obviously this file will sound terrible on a standard stereo reproduction system so that doesn’t “solve” the issue.

This is why I’m trying to get a full understanding of the under-the-hood operation of how streaming services handle ATMOS encoded files during playback.


Sent from my iPhone using Tapatalk

[its2loud]

I suggest you look up some delivery specs for certain streaming services. This will give you a good idea of what they want for delivery. As I mentioned, most want the full ADM Wav file. It all just depends on the service.

[JKHansen]

Thanks for that. I have done that and understand that they simply want a full ADM file. I guess my curiosity boils down to how do the streaming services determine which mixdown format to deliver (standard stereo vs binaural) when seeing a two channel playback system?


Sent from my iPhone using Tapatalk

[its2loud]

Atmos in 2-channel delivery is binaural. If the user selects to listen in “Atmos” they get the Atmos binaural mix. If the user chooses stereo, they get the plain old standard stereo mix.

Maybe you could explain your concern about this or are you asking if the platforms play two different mixes when the user selects?

[reichman]

The context is a desire to deliver music and narrative podcasts in a format that would be played back with binaural encoding when reproduced on headphones. Does this require the re-render and delivery of a separate binaural file from the renderer and, if so, can I expect that the file will be labeled accordingly by the streaming services.

Some of the podcast services support Dolby Atmos natively (Wondery and others). Back in 2017 there was no mechanism for this, and we would release podcasts as a binaural re-render 2-channel WAV to regular podcast distribution. Clients loved the sound, but the method was not ideal. It sounded very cool on headphones and in the car, but not so great on Bluetooth loudspeakers or stereo systems. Best practices now are to partner with a distro that properly supports Atmos, and deliver ADMs. The consumer experience will then be more similar to the music DSPs where the mix adapts to the available loudspeakers automatically. Ping me if you have more questions about this. The Wondery team is very approachable:

https://www.dolby.com/experience/wondery/