1
$\begingroup$

I am trying to use a 5.1.2 speaker configuration (left front, center, right front, left rear, right rear, left front height, right front height and Low-Frequency) to create a phantom sound at a certain location. I am looking for an algorithm which takes the 3D coordinate and current velocity of the emitter to calculate sound level at each speaker.

I have been doing research on this and what I found was use Head Related Transfer Function to calculate the sound pressure at right / left pinna, and use cross-talk cancellation to eliminate cross-talk effect from multiple speakers. However, I suppose HRTF is meant for creating virtual sound from headphones which does not help from my perspective. Can someone give some ideas on this? Thanks!

$\endgroup$
1
$\begingroup$

I believe that Crosstalk Cancellation is not the method you are looking for (based on the fact that you are using more than 2 loudspeakers for sound reproduction). You have not described exactly what the problem is that you are currently experiencing, and I therefore provide you with the following as general advice and guidance:

A binaural recording is generated utilising HRTFs that describe a particular listeners head's filtering (due to head, pinna, torso etc) of the original sound source. When played back over a pair of loudspeakers (or in your case many) the crosstalk paths from loudspeakers to the actual listeners ears degrade the virtual image, which is why Crosstalk Cancellation filters are typically employed (to overcome this problem of "crosstalk" from each loudspeaker to each ear). A great, simple, explanation of this can be found here http://pcfarina.eng.unipr.it/Aurora/crostalk.htm, however if you wish to delve deaper into Crosstalk Cancellation theory then I would recommend this paper for you: http://asa.scitation.org/doi/abs/10.1121/1.3206660

If you are extending your playback system to a larger number of loudspeakers, then depending on your desired listening position/area you may wish to consider alternative soundfield reproduction methods such as Ambisonics or Wave Field Synthesis. You therefore don't have to rely on an individuals HRTF, nor Crosstalk Cancellation etc... Since you have an irregular loudspeaker placement you may wish to consult research such as https://www.researchgate.net/profile/Yukio_Iwaya2/publication/268328126_High_order_Ambisonic_decoding_method_for_irregular_loudspeaker_arrays/links/54f5044e0cf2ba6150645b70.pdf for more information on how to design your playback algorithms.

I hope this is the help you were after?!

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.