RonanQ

Quick question, is it possible to make use of the hoalibrary's hoa.binaural object in conjunction with icst ambisonics ambiencode object using a first order? If so, how? I've been trying to create a headphone recreation of a first order system to mix with without the need of a 5.1 setup all the time and I can't seem to figure out how to get it to sound correct - as in there is no binaural heard, just a heavily lobsided stereo effect. I notice the harmonics for binaural only take 3 and have tried all possible combinations - is there an additional mathematical step required? 

hoalibrary-icst-tools

1st : ICST tools don't use the same normalization, I think you should use the factor sqrt(2)/2 on all the ambisonics channels except the W channel. In fact, it depends on what kind of normalization is used by ICST (I think under 3rd order is the Furse-Malham), HOA objects don't use because there's no limit for the order.

2nd : Binaural is based on artificial heads that should be different from yours, that's why binaural never works as well as they say in papers. you should try for large pinnae.

3rd : There was a pi/2 offset with hoa.binaural~, the problem solved in the last version of HOA.

If you use ambisonics for a 2D spatialization only, you should only use HOA it should be easier.  Look at the last version, there are a lot of new tools and UI objects to facilitate the sound spatialization.

But, what if we have 3d trajectories? Is there a way to represent source height with filters, even if it is slightly crude?

But, what if we have 3d trajectories?  Is there a way to represent source height with filters, even if it is slightly crude?


If it's not hurry, you can wait the 3D version of Hoa that is in progress. Otherwise, you can do it yourself. It's not optimized but it should work (at least for the first decomposition order).

1 - Download the HRTF data base of your choice (IRCAM, CIPIC or MIT are free )

2 - Decode the sound field for the best configuration you can (look at the platonic solids).

3 - Select the responses that match with your decoding (or the nearest ones if the azimuth and the elevation don't match exactly). You should have two set of responses, one for the left ear and one for the right (sometimes you should use the left ear responses and inverse the azimuth for the right ear).

4 - For each ear, filter the decoding signals with the matching responses (you can use [buffir~]) then sum the result.

1 - Download the HRTF data base of your choice (IRCAM, CIPIC or MIT are free )
2 - Decode the sound field for the best configuration you can (look at the platonic solids).
3 - Select the responses that match with your decoding (or the nearest ones if the azimuth and the elevation don't match exactly). You should have two set of responses, one for the left ear and one for the right (sometimes you should use the left ear responses and inverse the azimuth for the right ear).
4 - For each ear, filter the decoding signals with the matching responses (you can use [buffir~]) then sum the result.


I am just trying to do the same think nowadays. I am using ICST ambisonics for 1st order but I want to implement binaural in it. I am just wondering did you use hoa.binaural object in ICST patch or did you try something else?

Hi RONANQ
I am just trying to do the same think nowadays. I am using ICST ambisonics for 1st order but I want to implement binaural in it. I am just wondering did you use hoa.binaural object in ICST patch or did you try something else?

hoalibrary & icst tools