I'm looking for a utility which gives me from a sample (or .pvx  
analysis file) the N most prominent frequencies and their amplitudes.
Does anyone have a tip for this? (I'm on OSX)
Or can this be done in Csound?
Thanks -

	joachim

have u seen this?

http://www.klingbeil.com/spear/


joachim heintz wrote:
> 
> I'm looking for a utility which gives me from a sample (or .pvx  
> analysis file) the N most prominent frequencies and their amplitudes.
> Does anyone have a tip for this? (I'm on OSX)
> Or can this be done in Csound?
> Thanks -
> 
> 	joachim
> 
> 
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
> 
> 


-----
http://www.phasetransitions.net
http://www.facebook.com/pages/hermetic-music/38525750672
http://www.myspace.com/hermeticmusic


-- 
View this message in context: http://www.nabble.com/fft-utility-tp22196987p22215110.html
Sent from the Csound - General mailing list archive at Nabble.com.

I do this in the feedback instrument in ImproSculpt4. I use Python to
sort the list of frequencies by amplitude. The code is in the files:
/comp/feedbackInstrument.py
/inc/instr_feedback.inc
/inc/adaptivefilter_pvs.inc

I know this is not a complete and ready version of what you need
(sorry), but it might provide a model if you want to build something.
To outline the process:
- pvs opcodes in Csound write freq and amp data to ftables
- Python reads the tables
- Python search the tables (as lists) to find the frequencies with
strongest amplitudes (see the method
FeedbackInstrument.getFreqAndAmpsForDisplay in feedbackInstrument.py).

best
Oeyvind


2009/2/25 joachim heintz :
> I'm looking for a utility which gives me from a sample (or .pvx analysis
> file) the N most prominent frequencies and their amplitudes.
> Does anyone have a tip for this? (I'm on OSX)
> Or can this be done in Csound?
> Thanks -
>
>        joachim
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>

thanks, yes, but as far as i see i can't select the N most prominent  
bins when exporting the analysis file, can i?
	j



Am 26.02.2009 um 01:20 schrieb Tim Mortimer:

>
> have u seen this?
>
> http://www.klingbeil.com/spear/
>
>
> joachim heintz wrote:
>>
>> I'm looking for a utility which gives me from a sample (or .pvx
>> analysis file) the N most prominent frequencies and their amplitudes.
>> Does anyone have a tip for this? (I'm on OSX)
>> Or can this be done in Csound?
>> Thanks -
>>
>> 	joachim
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe
>> csound"
>>
>>
>
>
> -----
> http://www.phasetransitions.net
> http://www.facebook.com/pages/hermetic-music/38525750672
> http://www.myspace.com/hermeticmusic
>
>
> -- 
> View this message in context: http://www.nabble.com/fft-utility-tp22196987p22215110.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"

from memory (i havent done this in a while ... i'm ALMOST on the verge of
getting back into some computer music again ... been rather distracted by
the real world of late...) 

i use SPEAR to find the most prominent bins by sight. 
select them. 
then select inverse & delete the other 999 or thereabouts...

you can then save the data in a .txt file - basicaly it reads like SDIF data
- each "tracked partial" has a running ID & freq value

i then use python to analyse (boring stuff, min freq/amp, max freq/amp, mean
etc...)
& export further .txt's to use as masks, or additive synthesis datasets via
GEN23 etc... 

i havent experimented with Victor Lazzarini's TRACKS stuff... that might be
another more direct route to the same types of outcomes?

-----
http://www.phasetransitions.net
http://www.facebook.com/pages/hermetic-music/38525750672
http://www.myspace.com/hermeticmusic


-- 
View this message in context: http://www.nabble.com/fft-utility-tp22196987p22244331.html
Sent from the Csound - General mailing list archive at Nabble.com.

It should be clarified that the detection of the most prominent bins
is not in itself the same as partial tracking, which is a bit more
complex (this is what spears, Tracks opcodes and ATS do).

Another option would be to use pvstencil to pass bins above a certain
threshold only and then write the result to a table which is
periodically written to a text file. I haven't tried but it should be
possible. Or you could write the result to a file using pvsfwrite.

Best
Peiman



2009/2/27 Tim Mortimer :
>
> from memory (i havent done this in a while ... i'm ALMOST on the verge of
> getting back into some computer music again ... been rather distracted by
> the real world of late...)
>
> i use SPEAR to find the most prominent bins by sight.
> select them.
> then select inverse & delete the other 999 or thereabouts...
>
> you can then save the data in a .txt file - basicaly it reads like SDIF data
> - each "tracked partial" has a running ID & freq value
>
> i then use python to analyse (boring stuff, min freq/amp, max freq/amp, mean
> etc...)
> & export further .txt's to use as masks, or additive synthesis datasets via
> GEN23 etc...
>
> i havent experimented with Victor Lazzarini's TRACKS stuff... that might be
> another more direct route to the same types of outcomes?
>
> -----
> http://www.phasetransitions.net
> http://www.facebook.com/pages/hermetic-music/38525750672
> http://www.myspace.com/hermeticmusic
>
>
> --
> View this message in context: http://www.nabble.com/fft-utility-tp22196987p22244331.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>

Thanks to all for the suggestions, especially to Luis Antunes Pena who  
sent me a .csd on this subject. This showed me that what I want can  
well be done in Csound. I'd like to share the results:
1) a .csd which takes a .pvx file as input;
2) a bash script which takes an audio file as input.
The program looks for the N strongest amplitudes (one for each bin)  
and prints them, together with the corresponding frequency, time and  
the number of the bin.
Thanks for any corrections -

	joachim





Am 27.02.2009 um 15:00 schrieb peiman khosravi:

> It should be clarified that the detection of the most prominent bins
> is not in itself the same as partial tracking, which is a bit more
> complex (this is what spears, Tracks opcodes and ATS do).
>
> Another option would be to use pvstencil to pass bins above a certain
> threshold only and then write the result to a table which is
> periodically written to a text file. I haven't tried but it should be
> possible. Or you could write the result to a file using pvsfwrite.
>
> Best
> Peiman
>
>
>
> 2009/2/27 Tim Mortimer :
>>
>> from memory (i havent done this in a while ... i'm ALMOST on the  
>> verge of
>> getting back into some computer music again ... been rather  
>> distracted by
>> the real world of late...)
>>
>> i use SPEAR to find the most prominent bins by sight.
>> select them.
>> then select inverse & delete the other 999 or thereabouts...
>>
>> you can then save the data in a .txt file - basicaly it reads like  
>> SDIF data
>> - each "tracked partial" has a running ID & freq value
>>
>> i then use python to analyse (boring stuff, min freq/amp, max freq/ 
>> amp, mean
>> etc...)
>> & export further .txt's to use as masks, or additive synthesis  
>> datasets via
>> GEN23 etc...
>>
>> i havent experimented with Victor Lazzarini's TRACKS stuff... that  
>> might be
>> another more direct route to the same types of outcomes?
>>
>> -----
>> http://www.phasetransitions.net
>> http://www.facebook.com/pages/hermetic-music/38525750672
>> http://www.myspace.com/hermeticmusic
>>
>>
>> --
>> View this message in context: http://www.nabble.com/fft-utility-tp22196987p22244331.html
>> Sent from the Csound - General mailing list archive at Nabble.com.
>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"

joachim,

this is really nice - thanks for writing it and sharing it.  can't  
wait to show it to my DSP students next week!

-dr.B.

Dr. Richard Boulanger
Professor of Electronic Production and Design
Music Technology Division
Berklee College of Music
1140 Boylston Street
Boston, MA 02135
rboulanger@berklee.edu
http://csounds.com/boulanger
617-747-2485 (office)



On Mar 11, 2009, at 10:10 AM, joachim heintz wrote:

> Thanks to all for the suggestions, especially to Luis Antunes Pena  
> who sent me a .csd on this subject. This showed me that what I want  
> can well be done in Csound. I'd like to share the results:
> 1) a .csd which takes a .pvx file as input;
> 2) a bash script which takes an audio file as input.
> The program looks for the N strongest amplitudes (one for each bin)  
> and prints them, together with the corresponding frequency, time and  
> the number of the bin.
> Thanks for any corrections -
>
> 	joachim
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>
>
>
> Am 27.02.2009 um 15:00 schrieb peiman khosravi:
>
>> It should be clarified that the detection of the most prominent bins
>> is not in itself the same as partial tracking, which is a bit more
>> complex (this is what spears, Tracks opcodes and ATS do).
>>
>> Another option would be to use pvstencil to pass bins above a certain
>> threshold only and then write the result to a table which is
>> periodically written to a text file. I haven't tried but it should be
>> possible. Or you could write the result to a file using pvsfwrite.
>>
>> Best
>> Peiman
>>
>>
>>
>> 2009/2/27 Tim Mortimer :
>>>
>>> from memory (i havent done this in a while ... i'm ALMOST on the  
>>> verge of
>>> getting back into some computer music again ... been rather  
>>> distracted by
>>> the real world of late...)
>>>
>>> i use SPEAR to find the most prominent bins by sight.
>>> select them.
>>> then select inverse & delete the other 999 or thereabouts...
>>>
>>> you can then save the data in a .txt file - basicaly it reads like  
>>> SDIF data
>>> - each "tracked partial" has a running ID & freq value
>>>
>>> i then use python to analyse (boring stuff, min freq/amp, max freq/ 
>>> amp, mean
>>> etc...)
>>> & export further .txt's to use as masks, or additive synthesis  
>>> datasets via
>>> GEN23 etc...
>>>
>>> i havent experimented with Victor Lazzarini's TRACKS stuff... that  
>>> might be
>>> another more direct route to the same types of outcomes?
>>>
>>> -----
>>> http://www.phasetransitions.net
>>> http://www.facebook.com/pages/hermetic-music/38525750672
>>> http://www.myspace.com/hermeticmusic
>>>
>>>
>>> --
>>> View this message in context: http://www.nabble.com/fft-utility-tp22196987p22244331.html
>>> Sent from the Csound - General mailing list archive at Nabble.com.
>>>
>>>
>>>
>>> Send bugs reports to this list.
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>