• Profile picture of warusfel
  • Profile picture of T.


Groupe Public active 1 day ago

User group for Spat, panoramix, ToscA , and ADMix

Question about binaural coding.

Auteur 2 Utilisateurs souscrits |
Profile photo of HEM de Genève


Question for Mr T.

We recently decided to make some tests to compare different HAO decoding “methods” and “types” for . Just to know what sounds good, or not. For now it’s not finish. And it’s not the aim of this mail.

So we test with our HOA system but also in binaural. Until recently, if I understand well, we can’t transform a HOA “B-format” directly into a binaural stream. It need a “room” with speakers, fake or not. Because our ambisonic speaker system is not “perfect” we prefer to use a “soccer ball”. A virtual space with 64 speakers well positioned by the spat. Then we send the signal to monitoring bus and record it. This way to process means that the binaural stream depend of the method and the type of decoding of our 64 HOA buses. So we thought it was a good idea to include this part in our tests and the first results are, in a way, surprising. It changes a lot in function of the persons, but once again it’s not the point…

I also noticed (lately…) the presence of the new binaural and I ask my self : Do you finally make it possible to transform sound space directly into binaural ? And, because it looks to be the case, do you use a “soccer ball” with particular decoding methods and types, or it’s a new way to do without any room transition ?

Our tests won’t be useless because for now, it’s not possible to send a-format, b-format, eigenmike or any ambisonic stream into a binaural bus. Will it change soon or not ? Just to know if we continue our tests or if the futur updates will solve this problem.

Thanks a lot.

Décembre 4, 2018 à 11:25 #28458
Profile photo of T.

Hi Christophe,

Your understanding is correct: at the moment, there is no way to directly “transcode” a HOA stream to binaural. A decoder through virtual loudspeakers is required. As you noticed, one major drawback of this approach is that the resulting binaural mix strongly depends on the decoding stage (number of speakers, layout, and decoding method/type).
We definitely intend to replace that with a “direct transcoding” approach; however I cannot provide an agenda for that (hopefully next year?).
In the meanwhile you should continue your tests — and we would be interested if you want/can share the results.


Décembre 5, 2018 à 15:55 #28480

Vous devez être connecté pour répondre à ce sujet.

Log in now