Hi, I’m experimenting with the AmbisonicToolKit and especially the binaural decoders.
I tried several subjects of the IRCAM (using newListen) database and the CIPIC data base (using newCIPIC). But I can not find one that works for me.
So I wonder if I can apply my own customized HRTF. I had one made some time ago. Is there a binaural decoder that can work with custom HRTFs?
@simmania, the short answer is “not directly”.
A longer answer, “sure… if you know what you’re doing.” That is, you could write your own extension to create a new method to load your custom HRIR. I’ll add, to sound good, you really do need to optimize your “raw” HRIRs for Ambisonics. You can read about how we’ve prepared ATK included 3rd party HRIRs here.
We ARE intending to add a tutorial at some point illustrating how to do so given a measured set of HRIRs, using both the ATK and SignalBox.
FYI, a medium term goal is be able to import SOFA encoded data for this purpose.
I’m affraid it is out of my reach to do that myself. When will this functionality be implemented?
What will be a good generic HRTF?
One of the IRCAM database or one of the CIPIC data base. And which index?
We are creating a VR experience and want to use binaural audio. This should sound as good as possible for as much people as possible. Customized HRTFs are not possible yet. So what would be a good generic HRTF for this?
We don’t have definitive timeline. As a first step we’re looking at bringing reading SOFA to SC.
My preference is to use the default (subjectID = 4) Duda spherical model. This is just a sphere of average head size.
The CIPIC library includes two KEMAR dummy heads. From the documentation:
The CIPIC HRTF database provides two KEMAR dummy measurements. subjectID = 21 is the large pinnae dummy and subjectID = 165 is the small pinnae dummy.
I’d suggest trying the three I’ve mentioned above…