[PD] earplug~ or FIR~ for HRTF?

[Claude Heiland-Allen]

Nicholas Mariette wrote:
> IEM's bin_ambi is a sophisticated binaural rendering system via higher 
> order Ambisonic spatialisation to virtual speaker arrays.
> http://iem.at/Members/noisternig/bin_ambi
> 
> FIR~ is very CPU intensive for HRTF processing.  It's much better to use 
> frequency domain convolution with fft~ & ifft~ which is the case with 
> bin_ambi.

I've had a look through the files in the pure-data svn at sf.net, but I 
can't seem to find a complete example of how to use it.  It seems very 
complicated to understand how all the pieces fit together.

So if anyone has a complete example for something like this:

     source1~ position1 source2~ position2 ....
      |        |         |        |
     [some combination of iem voodoo objects ...]
      |  |
     [dac~]
      :  :
     binaural stereo

I would appreciate it a great deal.

> I've listed some publicly available HRTF sets on my blog here:
> http://blog.soundsorange.net/2005/11/09/hrtf/

Thanks, that's a useful resource (some of the links are dead by now, 
though).

> Nick


Claude


> 
> http://soundsorange.net
> 
> 
> On Sep 4, 2008, at 4:09 PM, Claude Heiland-Allen wrote:
> 
>> Hi all,
>>
>> I'm considering whether to use the earplug~ HRTF external or some
>> alternative implementation of binaural spatialization.
>>
>> On the plus side, earplug~ seems quite easy to use.
>>
>> On the down side, earplug~ seems not to handle any sample-rate stuff at
>> all, so I'm wondering which samplerate the impulse responses in the
>> earplug_data.txt are sampled at (because obviously it would be somewhat
>> weird to use them at another rate...).
>>
>> I don't really need the interpolation provided by earplug~ either, so I
>> guess I could save some CPU by using FIR~ directly, but then I need to
>> obtain some HRIR's (ideally both at 44100Hz and 48000Hz sample rates).
>>
>> So: has someone already made a binaural processing patch using FIR~ and
>> some set of impulse responses and would care to share it?
>>
>>
>> Some context for the curious:  I'm trying to auralize 4D space using 4D
>> room simulation, with an "ear" made up of several co-solid listening
>> points in the 4D space, each mapped to a 3D source point for binaural
>> processing.
>>
>> Thanks for any tips,
>>
>>
>> Claude
>> -- 
>> http://claudiusmaximus.goto10.org
>>
>> _______________________________________________
>> Pd-list at iem.at mailing list
>> UNSUBSCRIBE and account-management -> 
>> http://lists.puredata.info/listinfo/pd-list
>