giorgioesse

Hi, I'm looking for a way to do resynthesis from a live signal using spectrum analysis and I'm using zsa.freqpeak~ for that, BUT I need a long time integration of the fft data, like 1 or 2 minutes.

Hi, I'm looking for a way to do resynthesis from a live signal using spectrum analysis and I'm using zsa.freqpeak~ for that, BUT  I  need  a long time integration of the fft data, like 1 or 2 minutes.
I'm stuck at this problem ...  
Ideas?

Hi, I'm looking for a way to do resynthesis from a live signal using spectrum analysis and I'm using zsa.freqpeak~ for that, BUT  I  need  a

help-needed-for-long-time-integration-spectrum-analysis

Do you want the peaks averaged over 1 or 2 minutes? You could stick with your current method, record the peaks you have and average over 1-2 minutes...

What do you want to do exactly?
Do you want the peaks averaged over 1 or 2 minutes? You could stick with your current method, record the peaks you have and average over 1-2 minutes...

what about a "simple" pfft and then you use [+=~ ] to add up the values?

@Roman, I've tried that but it doesn't do what I need... it sums everything and I get all the peaks at 1

@Jean-Francois
yes, but how do i record the peaks? 

@Roman, I've tried that but it doesn't do what I need... it sums everything and I get all the peaks at 1
I need to average the peaks over time.



If I understand your question correctly, there are many options to "record" the peaks, like for instance:

- you could graph the peaks (multislider in scrolling mode)

- build a list with the peaks (maybe [zl group])

If I understand your question correctly, there are many options to "record" the peaks, like for instance:
- you could graph the peaks (multislider in scrolling mode)
- build a list with the peaks (maybe [zl group])
- write them into a [coll]


help needed for long time integration spectrum analysis