diff options
Diffstat (limited to 'examples/05-cw_binaural~_creation_arguments.pd')
-rw-r--r-- | examples/05-cw_binaural~_creation_arguments.pd | 24 |
1 files changed, 24 insertions, 0 deletions
diff --git a/examples/05-cw_binaural~_creation_arguments.pd b/examples/05-cw_binaural~_creation_arguments.pd new file mode 100644 index 0000000..2ab4bba --- /dev/null +++ b/examples/05-cw_binaural~_creation_arguments.pd @@ -0,0 +1,24 @@ +#N canvas 855 359 391 423 10; +#X text 105 18 cw_binaural~ creation arguments; +#X text 21 50 each instance of the external has 3 creation arguments +with defaults values; +#X text 21 83 1st arg: is the length of the impulse response used for +filtering \, default value is 128; +#X text 22 115 2nd arg: is the filtering method: default value is 'RIFF' +\, which means that the filtering will be done in temporal domain: +this method has a huge computational cost \, but 0 ltency. The other +supported filtering method is 'FFT' \, it lowers the filtering cost +\, implies more latency \, and require to use the external in a block +\, examples will be given later.; +#X text 23 198 3rd arg: if set to 'nodelay' then interpolation is done +directly on the HRIR \, and results in less computations \, but results +in a not so good itd estimation in the interpolated filter. Otherwise +\, we proceed to a minimalphase and pure delay decomposition of the +HRIR. The user may chose the fractional delay algorithm used between +'Hermite4' \, '6points' \, 'linear' \, and 'nofractional'. Default +value for the 3rd arg is 'Hermite4' \, '6points' is also a good method +but more expensive. 'linear' and 'nofractional' should not be used +\, except for didactic purposes.; +#X obj 74 359 cw_binaural~ 128 RIFF Hermite4; +#X text 36 378 instantiation of the external with its defaults creation +arguments; |