User16599226418749942478 (Customer) asked a question.
What are some best practices in limiting when using the DAPS?
The moment I use an object it bypasses the routing within pro tools and is sent directly to the DAPS, correct? If that's accurate, then how do I limit that signal? Do I need a limiter on the actual track of the object itself? I haven't found a resource explaining that to me.
Additionally, when viewing my mix on the binaural meters, it plays significantly hotter. Why is that and how do I accommodate for that level bump?
Additionally, when viewing my mix on the binaural meters, it plays significantly hotter. Why is that and how do I accommodate for that level bump?
The routing for Object is set in the I/O set up output and bus page so it doesn't bypass routing in Pro Tools but will require limiting per object track. Groups can be used to apply similar parameters across tracks. VCA's can also be used to apply volume changes across tracks if not limiting. For Beds there are several mutlichannel limiters such as the Waves Spherix and Flux Audio Elixer that could be useful. There is also something from Hornet Audio called SAMP that creates a master bus for Atmos but I haven't tried it. There is a demo version. One an ADM BWF is created there is limiting that can be applied to the ADM BWF using the Dolby Atmos Album Assembler Tool as well.
Any renderer involves summing so levels and peaks can produce surprising results at time due to phase coherence and multiple objects panned to similar coordinates in time. This can be exaggerated when rendering to stereo. The binaural renderer does run hotter that the renderer for speakers/headphones. However, delivery specifications for streaming services should be based on the 5.1 re-render so this should be the focus on any gain and limiting operations.
Best,
Adam
If you are mixing to a target loudness level of -18LKFS for the Atmos measurement (5.1 re-render) than getting some clipping on the binaural monitoring output is not unusual. The only way to address any clipping on the binaural meters would be to further drop the overall level below -18LKFS or limit more. Since the binaural is not the deliverable to be encoded I wouldn't be overly concerned if you see some clipping on the binaural meters. If you feel this is impacting how you mix then you could drop the level for binaural monitoring and bring it back up to the target level prior to creating the ADM BWF. Final QC should be on speakers.
Hope this helps.
Best,
Adam
The MP4 with the Atmos mix delivered via the DDPlus JOC codec can be loaded onto IOS devices for QC with Apple's Spatial Audio over headphones/buds. For streaming services that use AC4-IMS codec that delivers the binaural render using the binaural metadata the best approach is to export a "BIN" re-render .wav file to audition on headphones.
For information of using the MP4 file on Apple devices please see this article: https://professionalsupport.dolby.com/s/article/How-can-I-QC-my-mix-on-Apple-devices
Thanks,
Adam
Sorry for the delay. The Binaural Re-Render is just a PCM .wav file. The MP4 only includes DDPlus JOC which doesn't include binaural metadata. There is no way to get AC4-IMS in the exported MP4 file exported from the Renderer to audition on Android devices. There are a few existing thread that go into this a bit further that might be worth a look though some information is out of date:
https://professionalsupport.dolby.com/s/question/0D54u00007k0lOeCAI/ac4ims-a-2channel-binaural-audio-signal-without-metadata-without-any-rendering-during-playback
https://professionalsupport.dolby.com/s/question/0D54u00007k0o2GCAQ/is-apple-spatial-audio-different-from-what-we-listen-to-in-the-dar
AC4-IMS is a two channel codec that is very close to the binaural output. It carries used for mobile applications only and carries additional metadata that is used when for Atmos virtualization for playback over phone/tablet speakers.
AC4-IMS can be used on Android and also on IOS with streaming services that use a player application with onboard decoding vs DDPlusJoc which is decoded at the OS level. Let me know if this helps.
Best,
Adam