Forums > MaxMSP

pfft~ phase?

December 19, 2007 | 8:19 pm

can anyone help with my phase issues. im trying to make adjustments to the phase of each individual frequency bin passed out by a cartopol~ object.This is located within a pfft~ subpatch size 1024 and an overlap of 2. The values from the phase outlet of the cartopol~ are passed into a buffer using poke~ (indexed using the bin index from fftin~ object). The issue that i’ve got is that the phase at each index seem to vary every time i open and reopen the buffer,i am passing a sine wave into the pfft~ and i thought id have 512 bins with set phases for each bin. is my understanding wrong and can anyone give me any pointers.


December 19, 2007 | 10:30 pm

On 19 Dec 2007, at 20:19, ben Gilman wrote:

> The issue that i’ve got is that the phase at each index seem to
> vary every time i open and reopen the buffer,i am passing a sine
> wave into the pfft~ and i thought id have 512 bins with set phases
> for each bin. is my understanding wrong and can anyone give me any
> pointers.

your understanding is not quite right. the phase in each fft frame is
supposed to change corresponding
to the frequency of your oscillator the window size of the fft and
the overlap factor.
if you think about it, this is logical. i would expect a constant
phase value only for oscillator frequencies that
are integer multiples of the fft size.
note also, that the phases of those bins, that contain minimal to no
energy, will act more or less randomly.
you might want to monitor the contents of the buffer with the
waveform object to easier observe the phase changes.
volker.


December 20, 2007 | 1:23 pm

that makes total sense. i think i need to find an alternative method for achieving my desired result this route seems far too complex for my understanding of max/msp. im fairly new to max/msp and im undertaking a noise canceling project, i need to cancel the incoming signal with an inverted phase version of the original. issues arose with the latency of passing one of the split signals into the pfft~ object. im not sure how do deal with this latency such that it maximizes the cancellation process back in the main patcher. any thoughts


December 20, 2007 | 4:34 pm


December 20, 2007 | 11:38 pm


Viewing 5 posts - 1 through 5 (of 5 total)