[Csnd] lpinterp
| Date | 2011-05-13 10:07 |
| From | menno |
| Subject | [Csnd] lpinterp |
Hi Victor, a while ago i was completing the example from the manual for lpinterp and thought perhaps with the latest changes it would work now, so i tried again but it still blows up. First i analyzed 2 files, using the pole type and 30 poles: menno@muziek:~$ csound -U lpanal -a -p30 fox.wav fox30poles.lpc time resolution is 1000.000 ns PortAudio real-time audio module for Csound virtual_keyboard real time MIDI plugin for Csound PortMIDI real time MIDI plugin for Csound 0dBFS level = 32768.0 Csound version 5.13 beta (double samples) May 12 2011 libsndfile-1.0.21 util lpanal: Using pole storage method audio sr = 44100, monaural opening WAV infile fox.wav analysing 121569 sample frames (2.8 secs) 607 lpc frames written to fox30poles.lpc menno@muziek:~$ csound -U lpanal -a -p30 flute.aiff flute30poles.lpc time resolution is 1000.000 ns PortAudio real-time audio module for Csound virtual_keyboard real time MIDI plugin for Csound PortMIDI real time MIDI plugin for Csound 0dBFS level = 32768.0 Csound version 5.13 beta (double samples) May 12 2011 libsndfile-1.0.21 util lpanal: Using pole storage method audio sr = 44100, monaural opening AIFF infile flute.aiff analysing 115506 sample frames (2.6 secs) 577 lpc frames written to flute30poles.lpc Then i used this completed example from the manual: |
| Date | 2011-05-13 13:02 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] lpinterp |
Thanks for pointing this out, I'll take a look. Victor On 13 May 2011, at 10:07, menno wrote: > Hi Victor, > a while ago i was completing the example from the manual for > lpinterp and > thought perhaps with the latest changes it would work now, so i > tried again > but it still blows up. > First i analyzed 2 files, using the pole type and 30 poles: > > menno@muziek:~$ csound -U lpanal -a -p30 fox.wav fox30poles.lpc > time resolution is 1000.000 ns > PortAudio real-time audio module for Csound > virtual_keyboard real time MIDI plugin for Csound > PortMIDI real time MIDI plugin for Csound > 0dBFS level = 32768.0 > Csound version 5.13 beta (double samples) May 12 2011 > libsndfile-1.0.21 > util lpanal: > Using pole storage method > audio sr = 44100, monaural > opening WAV infile fox.wav > analysing 121569 sample frames (2.8 secs) > 607 lpc frames written to fox30poles.lpc > menno@muziek:~$ csound -U lpanal -a -p30 flute.aiff flute30poles.lpc > time resolution is 1000.000 ns > PortAudio real-time audio module for Csound > virtual_keyboard real time MIDI plugin for Csound > PortMIDI real time MIDI plugin for Csound > 0dBFS level = 32768.0 > Csound version 5.13 beta (double samples) May 12 2011 > libsndfile-1.0.21 > util lpanal: > Using pole storage method > audio sr = 44100, monaural > opening AIFF infile flute.aiff > analysing 115506 sample frames (2.6 secs) > 577 lpc frames written to flute30poles.lpc > > Then i used this completed example from the manual: > > > |
| Date | 2011-05-14 22:36 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] lpinterp |
I think the situation with lpinterp is that it is liable to blow up, depending on the combination of sources. This is because while the two individual analyses might have been protected from producing unstable filter poles, the interpolation of these can introduce instability that leads the whole filter to blow up. So it is not a bug, but something inherent to the process. Victor On 13 May 2011, at 10:07, menno wrote: > Hi Victor, > a while ago i was completing the example from the manual for > lpinterp and > thought perhaps with the latest changes it would work now, so i > tried again > but it still blows up. > First i analyzed 2 files, using the pole type and 30 poles: > > menno@muziek:~$ csound -U lpanal -a -p30 fox.wav fox30poles.lpc > time resolution is 1000.000 ns > PortAudio real-time audio module for Csound > virtual_keyboard real time MIDI plugin for Csound > PortMIDI real time MIDI plugin for Csound > 0dBFS level = 32768.0 > Csound version 5.13 beta (double samples) May 12 2011 > libsndfile-1.0.21 > util lpanal: > Using pole storage method > audio sr = 44100, monaural > opening WAV infile fox.wav > analysing 121569 sample frames (2.8 secs) > 607 lpc frames written to fox30poles.lpc > menno@muziek:~$ csound -U lpanal -a -p30 flute.aiff flute30poles.lpc > time resolution is 1000.000 ns > PortAudio real-time audio module for Csound > virtual_keyboard real time MIDI plugin for Csound > PortMIDI real time MIDI plugin for Csound > 0dBFS level = 32768.0 > Csound version 5.13 beta (double samples) May 12 2011 > libsndfile-1.0.21 > util lpanal: > Using pole storage method > audio sr = 44100, monaural > opening AIFF infile flute.aiff > analysing 115506 sample frames (2.6 secs) > 577 lpc frames written to flute30poles.lpc > > Then i used this completed example from the manual: > > > |
| Date | 2011-05-15 10:30 |
| From | menno |
| Subject | [Csnd] Re: lpinterp |
Hello Victor,
i found another Csound thread about the pops and crackles:
http://csound.1045644.n5.nabble.com/LPC-random-bursts-bug-td1119318.html
so i resampled 2 soundfiles to 11kHz, cut off all possible more all less
silences...
menno@muziek:~$ csound -U lpanal -a -p50 -h250 11kHzflute_chopped.wav
11kHzflute_chopped.lpc
time resolution is 1000.000 ns
PortAudio real-time audio module for Csound
virtual_keyboard real time MIDI plugin for Csound
PortMIDI real time MIDI plugin for Csound
0dBFS level = 32768.0
Csound version 5.13 beta (double samples) May 13 2011
libsndfile-1.0.21
util lpanal:
Using pole storage method
audio sr = 11025, monaural
opening WAV infile 11kHzflute_chopped.wav
analysing 20661 sample frames (1.9 secs)
82 lpc frames written to 11kHzflute_chopped.lpc
menno@muziek:~$ csound -U lpanal -a -p50 -h250 11kHzfox.wav 11kHzfox.lpc
time resolution is 1000.000 ns
PortAudio real-time audio module for Csound
virtual_keyboard real time MIDI plugin for Csound
PortMIDI real time MIDI plugin for Csound
0dBFS level = 32768.0
Csound version 5.13 beta (double samples) May 13 2011
libsndfile-1.0.21
util lpanal:
Using pole storage method
audio sr = 11025, monaural
opening WAV infile 11kHzfox.wav
analysing 30392 sample frames (2.8 secs)
121 lpc frames written to 11kHzfox.lpc
And run the example with lpinterp/lpslot again. It blows up. i have not had
any sounding result with the lpinterp/lpslot - apart from a short explosion
trying to break my speakers :)
menno
--
View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4397365.html
Sent from the Csound - General mailing list archive at Nabble.com.
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-15 11:03 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] Re: lpinterp |
The problem is that pole interpolation can create unstable filters. I
have tried here and got various results with stable and unstable
filters.
A stability test might be placed in lpinterp, but that might open up a
series of other problems, though.
Some combinations of sounds create unstable combinations of filters,
while others don't. Also the number of poles seem to have an effect on
this. More poles have a tendency to create instability, but also
produce in general better analyses. In general, pole-storage is worse
in quality than coefficient-storage (the default).
Regards
Victor
On 15 May 2011, at 10:30, menno wrote:
> Hello Victor,
>
> i found another Csound thread about the pops and crackles:
> http://csound.1045644.n5.nabble.com/LPC-random-bursts-bug-td1119318.html
>
> so i resampled 2 soundfiles to 11kHz, cut off all possible more all
> less
> silences...
>
> menno@muziek:~$ csound -U lpanal -a -p50 -h250 11kHzflute_chopped.wav
> 11kHzflute_chopped.lpc
> time resolution is 1000.000 ns
> PortAudio real-time audio module for Csound
> virtual_keyboard real time MIDI plugin for Csound
> PortMIDI real time MIDI plugin for Csound
> 0dBFS level = 32768.0
> Csound version 5.13 beta (double samples) May 13 2011
> libsndfile-1.0.21
> util lpanal:
> Using pole storage method
> audio sr = 11025, monaural
> opening WAV infile 11kHzflute_chopped.wav
> analysing 20661 sample frames (1.9 secs)
> 82 lpc frames written to 11kHzflute_chopped.lpc
>
> menno@muziek:~$ csound -U lpanal -a -p50 -h250 11kHzfox.wav
> 11kHzfox.lpc
> time resolution is 1000.000 ns
> PortAudio real-time audio module for Csound
> virtual_keyboard real time MIDI plugin for Csound
> PortMIDI real time MIDI plugin for Csound
> 0dBFS level = 32768.0
> Csound version 5.13 beta (double samples) May 13 2011
> libsndfile-1.0.21
> util lpanal:
> Using pole storage method
> audio sr = 11025, monaural
> opening WAV infile 11kHzfox.wav
> analysing 30392 sample frames (2.8 secs)
> 121 lpc frames written to 11kHzfox.lpc
>
> And run the example with lpinterp/lpslot again. It blows up. i have
> not had
> any sounding result with the lpinterp/lpslot - apart from a short
> explosion
> trying to break my speakers :)
>
> menno
>
>
> --
> View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4397365.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
> Send bugs reports to the Sourceforge bug tracker
> https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body
> "unsubscribe csound"
>
Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-15 11:31 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] Re: lpinterp |
Actually, it's worse than that: the code tries to interpolate pole
phases, which is something not really advisable to do. I'll try to
see if I can put a bit of sanity in it.
Victor
On 15 May 2011, at 11:03, Victor Lazzarini wrote:
> The problem is that pole interpolation can create unstable filters.
> I have tried here and got various results with stable and unstable
> filters.
> A stability test might be placed in lpinterp, but that might open up
> a series of other problems, though.
>
> Some combinations of sounds create unstable combinations of filters,
> while others don't. Also the number of poles seem to have an effect
> on this. More poles have a tendency to create instability, but also
> produce in general better analyses. In general, pole-storage is
> worse in quality than coefficient-storage (the default).
>
> Regards
>
> Victor
>
> On 15 May 2011, at 10:30, menno wrote:
>
>> Hello Victor,
>>
>> i found another Csound thread about the pops and crackles:
>> http://csound.1045644.n5.nabble.com/LPC-random-bursts-bug-td1119318.html
>>
>> so i resampled 2 soundfiles to 11kHz, cut off all possible more all
>> less
>> silences...
>>
>> menno@muziek:~$ csound -U lpanal -a -p50 -h250 11kHzflute_chopped.wav
>> 11kHzflute_chopped.lpc
>> time resolution is 1000.000 ns
>> PortAudio real-time audio module for Csound
>> virtual_keyboard real time MIDI plugin for Csound
>> PortMIDI real time MIDI plugin for Csound
>> 0dBFS level = 32768.0
>> Csound version 5.13 beta (double samples) May 13 2011
>> libsndfile-1.0.21
>> util lpanal:
>> Using pole storage method
>> audio sr = 11025, monaural
>> opening WAV infile 11kHzflute_chopped.wav
>> analysing 20661 sample frames (1.9 secs)
>> 82 lpc frames written to 11kHzflute_chopped.lpc
>>
>> menno@muziek:~$ csound -U lpanal -a -p50 -h250 11kHzfox.wav
>> 11kHzfox.lpc
>> time resolution is 1000.000 ns
>> PortAudio real-time audio module for Csound
>> virtual_keyboard real time MIDI plugin for Csound
>> PortMIDI real time MIDI plugin for Csound
>> 0dBFS level = 32768.0
>> Csound version 5.13 beta (double samples) May 13 2011
>> libsndfile-1.0.21
>> util lpanal:
>> Using pole storage method
>> audio sr = 11025, monaural
>> opening WAV infile 11kHzfox.wav
>> analysing 30392 sample frames (2.8 secs)
>> 121 lpc frames written to 11kHzfox.lpc
>>
>> And run the example with lpinterp/lpslot again. It blows up. i
>> have not had
>> any sounding result with the lpinterp/lpslot - apart from a short
>> explosion
>> trying to break my speakers :)
>>
>> menno
>>
>>
>> --
>> View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4397365.html
>> Sent from the Csound - General mailing list archive at Nabble.com.
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>> https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>> "unsubscribe csound"
>>
>
> Dr Victor Lazzarini
> Senior Lecturer
> Dept. of Music
> NUI Maynooth Ireland
> tel.: +353 1 708 3545
> Victor dot Lazzarini AT nuim dot ie
>
>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
> https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body
> "unsubscribe csound"
>
Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-15 11:35 |
| From | menno |
| Subject | [Csnd] Re: lpinterp |
i see.
And lpinterp/lpslot are to be used with this (unstable-before-you-know-it)
pole filter...
That makes it hard for me to find working examples for the manual :)
I will continue my efforts but with less poles and hope to get a sounding
result.
I think your information should be in the manual, on the lpanal page bit
also on the lpinterp and lpslot pages.
But first let me get some working examples...do you have some for me - that
makes use of the wav or aiff that are in the manual/examples folder?
thanks for explaining!
Menno
--
View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4397444.html
Sent from the Csound - General mailing list archive at Nabble.com.
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-15 11:42 |
| From | menno |
| Subject | [Csnd] Re: lpinterp |
ah, sorry i did not see your second email.
i'll wait a bit for what you may find.
menno
--
View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4397451.html
Sent from the Csound - General mailing list archive at Nabble.com.
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-15 11:43 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] Re: lpinterp |
I think lpinterp is problematic, to say the least. However, pole
interpolation is also used in lpread, so I am not surprised it does
not sound
great.
Victor
On 15 May 2011, at 11:35, menno wrote:
> i see.
> And lpinterp/lpslot are to be used with this (unstable-before-you-
> know-it)
> pole filter...
> That makes it hard for me to find working examples for the manual :)
>
> I will continue my efforts but with less poles and hope to get a
> sounding
> result.
> I think your information should be in the manual, on the lpanal page
> bit
> also on the lpinterp and lpslot pages.
> But first let me get some working examples...do you have some for me
> - that
> makes use of the wav or aiff that are in the manual/examples folder?
>
> thanks for explaining!
> Menno
>
> --
> View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4397444.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
> Send bugs reports to the Sourceforge bug tracker
> https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body
> "unsubscribe csound"
>
Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-17 19:34 |
| From | menno |
| Subject | [Csnd] Re: lpinterp |
Hi Victor, i have done some further testing but still no luck, only the loud pop. I have analyzed fox.wav and flute.aiff every time diminishing the number of poles: foxnew50poles, foxnew34poles, foxnew20poles, foxnew10poles flutenew50poles, flutenew34poles, flutenew20poles, flutenew10poles This was the command: csound -U lpanal -a -pxx flute.aiff flutenewxxpoles.lpc and this the csd: |
| Date | 2011-05-17 19:54 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] Re: lpinterp |
I think it's the combination of those two files that create the problem. On 17 May 2011, at 19:34, menno wrote: > Hi Victor, > > i have done some further testing but still no luck, only the loud pop. > I have analyzed fox.wav and flute.aiff every time diminishing the > number of > poles: > > foxnew50poles, foxnew34poles, foxnew20poles, foxnew10poles > flutenew50poles, flutenew34poles, flutenew20poles, flutenew10poles > > This was the command: csound -U lpanal -a -pxx flute.aiff > flutenewxxpoles.lpc > > and this the csd: > > > |
| Date | 2011-05-18 11:30 |
| From | menno |
| Subject | [Csnd] Re: lpinterp |
it must be. As you stated earlier, pole interpolation is also used in lpread,
so it is hard to predict which combination of files will work with
lpinterp/lpslot. Do you see some logic? What kind of files should work as
far as you can see?
menno
--
View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4406130.html
Sent from the Csound - General mailing list archive at Nabble.com.
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|
| Date | 2011-05-18 11:33 |
| From | Victor Lazzarini |
| Subject | Re: [Csnd] Re: lpinterp |
The code needs some surgery. I will need to spend some time working
with it before it can be made a bit more sane and robust.
On 18 May 2011, at 11:30, menno wrote:
> it must be. As you stated earlier, pole interpolation is also used
> in lpread,
> so it is hard to predict which combination of files will work with
> lpinterp/lpslot. Do you see some logic? What kind of files should
> work as
> far as you can see?
>
> menno
>
> --
> View this message in context: http://csound.1045644.n5.nabble.com/lpinterp-tp4392607p4406130.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
> Send bugs reports to the Sourceforge bug tracker
> https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body
> "unsubscribe csound"
>
Dr Victor Lazzarini
Senior Lecturer
Dept. of Music
NUI Maynooth Ireland
tel.: +353 1 708 3545
Victor dot Lazzarini AT nuim dot ie
Send bugs reports to the Sourceforge bug tracker
https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
|