3D-Audio Class
University of Paris-Sud — Year 2 CS-Master (TC-2)
Theoretical and practical material for the Fundamentals of Virtual and Augmented Reality class (TC2, Master 2 Research in Human-Computer Interaction).
Pure Data practical instruction (2012)
Here is the binaural mixing program to complete. Try to finish it for this Thursday, before the next practical. Please feel free to contact me through email if you have any questions.
I also included a WAV file of an audio example to be processed. It is a voice that I recorded some years ago. The recording is completely raw without any effect. You can use any kind of other sounds that you wish as long as it is a mono wave file without effects.
Some information:
Stages (1), (2) and (3) are finalized. You just need to add a stage (4), which will handle the filtering.
You should first use a function to read the arraySrc, like tabplay~.
To apply the filter you have to use a convolution. You may apply it in the time or the frequency domain. The frequency domain is usually more efficient as it allows the use of the FFT (Fast Fourier Transform). Although FIR~ is chiefly designed to apply a Finite Impulse Response filter, it is also capable of applying a simple FFT convolution efficiently.
An important point is that HRTF does not provide the intensity decrease with the distance following the Inverse Square Law. You should then use the dist_m variable from the stage (2) to apply this effect in the stage (4).
To use a variable in Pd you can use the function r (which stands for "receive") followed by the variable name.
The last stage should be a dac~ module to transmit the sound to your audio interface.
Don't forget to tick the compute audio option in the main application window to hear some sound.
These explanations are rather long, but the stage (4) is rather small and straightforward if you follow these steps.
Extra:
-
If you wish to observe the shape of the HRTF (HRIR actually) file, you can use Audacity to import it:
-
File>Import>Raw Data... -
Use: 32 bit float; Little-endian; mono; 16 bytes offset; 44100 sample rate
Good luck to you all!