Hi there, haven´t posted here for a while but good to be back
I am wrapping my head around the future of “immersive audio” at the moment. Having explored ambisonics and am working with that for most of my projects I am discovering dolby atmos and seeing the commercial world using atmos more and more. It is a bit complicated to wrap my head around a lot of new information, there are a lot of videos explaining bits and pieces of it, with this one being one of the better ones https://youtu.be/FcZWOCMVkWE?si=8j1iAWKAONk-GOh_
What I am specifically looking for is a way of playing back a composition on headphones binaurally, but with the soundfield tilting and shifting due to headposition. So headtracking data should alter the rendered binaural image in real time.
This could be done with ambisonics b format decoded to binaural, or also atmos decoded to binaural
Setting something like this up on my laptop could be fairly easy, but having the same on a smartphone with a media player? As I am understanding it dolby atmos becomes the predominant format for that (maybe because with ambisonics is patent free? ) so it might make sense to work in Atmos, or convert from b-format to atmos for certain projects.
Long text , short question: Anyone here working in ambisonics or atmos doing these kind of projects? what would be an easy way to lets say make an immersive album and play it back binaural via headphones from a smartphone?
Having said that the guy in the video after explaining really well and thoroughly the atmos technical parts goes on in a rant kind of way how to master in atmos which is pretty annoying to my ears. So recommended the first hour lets say , afterwards it goes off topic…
The problem with Atmos is also that it is a proprietary format and even the encoders/decoders are. So you cannot even listen to your exported audio unless you use some closed source app.
Here is also another good counter-argument:
At this point in time, there is no support for Atmos in Ardour.
If you want to stay with Free/Libre Software: Ambisonics is likely the way to go. It is big in gaming and you can do binaural downmixes with it.
These days the most convenient is the IEM plugin suite. Many GNU/Linux distros package it:
Thanks for the input and the suggestions, yes that is actually almost what I am doing at the moment: Working my stereo sound stuff in Ardour, and working my ambisonics stuff in reaper with the iem plugins and sometimes trying the sparta plugin suite, but that is not that stable on my system.
Thanks for the video , it is good to see how badly Atmos translate (at least till now) to the consumer. But I have to say the same applies for Ambisonics. There are no films delivered in Ambisonics, nor streaming platforms use it, etc etc. It just works well for me doing experimental sound projects in a dedicated space or render to stereo or binaural for the listener. What I think would be amazing is, and maybe it exists and I don’t know: A media player where I can just throw in a file encoded in Ambisonics b format and the player recognizes the actual speaker setup and decodes to it ( for example on a smartphone rendering to binaural when headphone are connected and decoding to stereo when some bluetooth speaker is connected) or you can just dial in quickly your loudspeaker configuration before playback.
This way you could deliver your music or film and you could be assured it would be listened back “the right way”.
From what I experienced till now working with ambisonics that it is impossible to have sounds travelling through the listener position, having the perception that a sound is coming closer. The sounds usually just moves on the spheres. And as it is also not object based , it is of limited use for gaming as you cannot “walk” through the sound field but you are fixed to one listening position. I don´t know if these issues are adressed in Atmos, but these considerations initially sparked my interest to look further and see if there is an alternative to ambisonics as we know it.
Hi Nettings, I m just going through your talk, very interesting! Also that you mention the Blauert fields, I use them sometimes (pushing 8kHz for example) to trick vertical perception and it works surprisingly well!
