Hi all,
I'm doing a piece with a viola d'amore (7 string viol) where [sigmund~] is analysing the live input and feeding 48 tracks/partials to a bank of 48 [osc~] for resynthesis and a bit of jiggery-pokery, mainly reversing the amplitudes so the quietest partials are the loudest and vice versa.
The viola is playing long(ish) single notes between 1-3 seconds and the problem I am having is that when sigmund processes the partials they come out rhythmicised/pulses which seems to be a combination of the 'hop' and 'npts' settings. As the values for hop and npts get higher the pulses get longer up until my current huge setting: [sigmund~ -npts 65536 -hop 65536 -npeak 48 tracks] when the partial frequencies balance out into steady lines. I have also tried -hop 131072 which works too but trying that with the -npts completely locks pd.
The problem I have now is that there is a significant delay between the beginning of the processing in comparison to the viola and worse still is the overlap after the viola has stopped playing. Also the viola notes are indeterminate so I can't just set a cutoff to sync the two together. I could use the input of the viola to gate the tail of the processing but I am currently using a combination of [env~] and [lop~] so that as the performer plays his amplitude controls the processing in reverse, so when he's loud the processing is quiet etc. I suppose I can add another layer on top of the end
I am aware that the gap between the two sounding events is caused by the huge window sizes I am using and that there may well be no way around this to get the kind of sound from the processing I'm looking for, i.e. without the pulses, but I am wondering if anyone has any tips and tricks I can apply to try and bring the two sounds, viola and processing, closer together. Or ideally some way of achieving the smoother processing with lower npts and hop settings. This is for a live performance btw.
All good wishes,
Julian
Could you make a simple example patch that demonstrates the rhythmicized/pulsed result that you don't like? Without knowing exactly what's going on in your patch, my first question would be whether or not you're smoothing the frequencies and amplitudes with a [line~] or something else. Maybe sudden changes due to fleeting tracking errors cause glitches that pulse in sync with the analysis period?
With a violin as input you should get a reasonable enough spectrum to get stable harmonic tracking with a window much smaller than 65536. The low G on a violin is 196Hz, so the bin spacing you'd get with -npts 4096 would give plenty of resolution for catching multiples of that fundamental and higher ones, and only 92ms of extra delay. The hop setting you choose will just affect how often the information is updated (and CPU usage too). You should expect to deal with some temporary tracking errors during moments like attacks, where components of the spectrum won't be in a perfectly harmonic relationship.
Going beyond 131072 points using Pd's built-in mayer_fft() has given me garbage as output in the past. Maybe someone else knows something about this limitation?
On Wed, Feb 9, 2011 at 11:41 AM, J bz jbeezez@gmail.com wrote:
Hi all,
I'm doing a piece with a viola d'amore (7 string viol) where [sigmund~] is analysing the live input and feeding 48 tracks/partials to a bank of 48 [osc~] for resynthesis and a bit of jiggery-pokery, mainly reversing the amplitudes so the quietest partials are the loudest and vice versa.
The viola is playing long(ish) single notes between 1-3 seconds and the problem I am having is that when sigmund processes the partials they come out rhythmicised/pulses which seems to be a combination of the 'hop' and 'npts' settings. As the values for hop and npts get higher the pulses get longer up until my current huge setting: [sigmund~ -npts 65536 -hop 65536 -npeak 48 tracks] when the partial frequencies balance out into steady lines. I have also tried -hop 131072 which works too but trying that with the -npts completely locks pd.
The problem I have now is that there is a significant delay between the beginning of the processing in comparison to the viola and worse still is the overlap after the viola has stopped playing. Also the viola notes are indeterminate so I can't just set a cutoff to sync the two together. I could use the input of the viola to gate the tail of the processing but I am currently using a combination of [env~] and [lop~] so that as the performer plays his amplitude controls the processing in reverse, so when he's loud the processing is quiet etc. I suppose I can add another layer on top of the end
I am aware that the gap between the two sounding events is caused by the huge window sizes I am using and that there may well be no way around this to get the kind of sound from the processing I'm looking for, i.e. without the pulses, but I am wondering if anyone has any tips and tricks I can apply to try and bring the two sounds, viola and processing, closer together. Or ideally some way of achieving the smoother processing with lower npts and hop settings. This is for a live performance btw.
All good wishes,
Julian
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On Thu, 10 Feb 2011, William Brent wrote:
I'm doing a piece with a viola d'amore (7 string viol)
Oops, that's not a violin :) I don't know the lowest note on that one offhand...
Is everybody only ever using the default tuning on those things ?
| Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal, QC
Hey William,
Many thanks for offering to have a look at this for me. I'm currently trying to extricate the offending section out of what has now ballooned up into a very involved patch (certainly for me anyway). In the process of doing this I have realised that I am mistakenly constantly sending attacks and decays to the[osc~]'s. It would probably be best for me to just send it 'as is' rather than trying to sort it out. I'm somewhat in over my head with this as it is (maybe not always a bad place to be) and could really do with some advice/help/coding-elegance as my brain is starting to melt.
Mathieu,
If you mean the tuning of the viola, yes it's the standard D Maj tuning, low to high A D A D F# A D.
Very best wishes,
Julian
P.S. Back soon
On 10 February 2011 23:16, Mathieu Bouchard matju@artengine.ca wrote:
On Thu, 10 Feb 2011, William Brent wrote:
I'm doing a piece with a viola d'amore (7 string viol)
Oops, that's not a violin :) I don't know the lowest note on that one offhand...
Is everybody only ever using the default tuning on those things ?
| Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal, QC
Right then,
I have only sent one sample only: '7_45_A_violadamore.wav' (7-string number, 45-midi number, A-note) which you will need to send into sigmund~ - I'm using [throw~ viola_samples] to get it in.
From sigmund~ it goes into '[pd sine_output] then thrown to
[sines] which has a switch~, top right.
I'm using GEM and MSD to create a 'swarm' which has the partials mapped to it so I haven't included my output~ as that would mean including loads of extra files.
So you will need an output~ to [catch~] the 2 track [throws~] [throw~ revsound_L] [throw~ revsound_R]
I suppose it's quite possible that as your not receiving any pan info the audio will be stuck either left or right anyway?
What I was trying to achieve with the [attdec_gen2] inside [sine_rev_gen_2] was a simple attack and decay but what is happening is that they are being triggered all the time, aargh.
I would really like the amp and freq from sigmund~ to be as quick and as smooth as possible, so any assistance with that will be hugely helpful.
As I'm not hugely confident in my programming abilities please point out any obvious mistakes and any and all tips and tricks/elegant solutions gratefully accepted. Really appreciate the help people.
Cheers,
Julian
On 11 February 2011 21:18, J bz jbeezez@gmail.com wrote:
Hey William,
Many thanks for offering to have a look at this for me. I'm currently trying to extricate the offending section out of what has now ballooned up into a very involved patch (certainly for me anyway). In the process of doing this I have realised that I am mistakenly constantly sending attacks and decays to the[osc~]'s. It would probably be best for me to just send it 'as is' rather than trying to sort it out. I'm somewhat in over my head with this as it is (maybe not always a bad place to be) and could really do with some advice/help/coding-elegance as my brain is starting to melt.
Mathieu,
If you mean the tuning of the viola, yes it's the standard D Maj tuning, low to high A D A D F# A D.
Very best wishes,
Julian
P.S. Back soon
On 10 February 2011 23:16, Mathieu Bouchard matju@artengine.ca wrote:
On Thu, 10 Feb 2011, William Brent wrote:
I'm doing a piece with a viola d'amore (7 string viol)
Oops, that's not a violin :) I don't know the lowest note on that one offhand...
Is everybody only ever using the default tuning on those things ?
| Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal, QC
Hi Julian,
I just looked at the patches and I hear what you're talking about now. The erratic changes are because it's really hard to get nice smooth tracks by analyzing this kind of signal. With violin samples I got useable results, but the rougher timbre of your viol has a lot of high frequencies that emerge and disappear unpredictably. I think that getting 48 stable tracks out of it is probably a lost cause...even with only 10 it was pretty bad with sigmund~'s default settings.
The only way I could imagine improving things is to use the flag that shows up with each track list to indicate whether it's a new track or a continuation of an old one. With that information, you could try fading in new tracks on free oscillators in the bank to avoid the sudden discontinuity when frequency jumps by a huge amount. Data from a continuing track could safely be sent to the same oscillator. I guess you could also EQ it to get rid of those unpredictable high frequencies - I plotted the spectrogram and the most troublesome ones are above 2500Hz. Of course, that would also put a big damper on the timbre you're interested in...
I wish I had a great solution for you, but this is a tough one. Assigning frequencies to specific oscillators in your bank based on track flags really seems like the best bet.
On Fri, Feb 11, 2011 at 4:54 PM, J bz jbeezez@gmail.com wrote:
Right then,
I have only sent one sample only: '7_45_A_violadamore.wav' (7-string number, 45-midi number, A-note) which you will need to send into sigmund~ - I'm using [throw~ viola_samples] to get it in. From sigmund~ it goes into '[pd sine_output] then thrown to [sines] which has a switch~, top right.
I'm using GEM and MSD to create a 'swarm' which has the partials mapped to it so I haven't included my output~ as that would mean including loads of extra files.
So you will need an output~ to [catch~] the 2 track [throws~] [throw~ revsound_L] [throw~ revsound_R]
I suppose it's quite possible that as your not receiving any pan info the audio will be stuck either left or right anyway?
What I was trying to achieve with the [attdec_gen2] inside [sine_rev_gen_2] was a simple attack and decay but what is happening is that they are being triggered all the time, aargh.
I would really like the amp and freq from sigmund~ to be as quick and as smooth as possible, so any assistance with that will be hugely helpful.
As I'm not hugely confident in my programming abilities please point out any obvious mistakes and any and all tips and tricks/elegant solutions gratefully accepted. Really appreciate the help people.
Cheers,
Julian
On 11 February 2011 21:18, J bz jbeezez@gmail.com wrote:
Hey William,
Many thanks for offering to have a look at this for me. I'm currently trying to extricate the offending section out of what has now ballooned up into a very involved patch (certainly for me anyway). In the process of doing this I have realised that I am mistakenly constantly sending attacks and decays to the[osc~]'s. It would probably be best for me to just send it 'as is' rather than trying to sort it out. I'm somewhat in over my head with this as it is (maybe not always a bad place to be) and could really do with some advice/help/coding-elegance as my brain is starting to melt.
Mathieu,
If you mean the tuning of the viola, yes it's the standard D Maj tuning, low to high A D A D F# A D.
Very best wishes,
Julian
P.S. Back soon
On 10 February 2011 23:16, Mathieu Bouchard matju@artengine.ca wrote:
On Thu, 10 Feb 2011, William Brent wrote:
I'm doing a piece with a viola d'amore (7 string viol)
Oops, that's not a violin :) I don't know the lowest note on that one offhand...
Is everybody only ever using the default tuning on those things ?
_______________________________________________________________________ | Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal, QC
Hey William,
Ho hum...,
Cheers for having a look though.
I'm not completely clear about the continuation flags? I can see from the help file that they're there but what do they mean?
Perhaps this should be a new thread but why does sigmund have a frequency range of 100,000hz, and what would be a decent useable range from practical experience?
Regards,
Julian
On 13 February 2011 02:50, William Brent william.brent@gmail.com wrote:
Hi Julian,
I just looked at the patches and I hear what you're talking about now. The erratic changes are because it's really hard to get nice smooth tracks by analyzing this kind of signal. With violin samples I got useable results, but the rougher timbre of your viol has a lot of high frequencies that emerge and disappear unpredictably. I think that getting 48 stable tracks out of it is probably a lost cause...even with only 10 it was pretty bad with sigmund~'s default settings.
The only way I could imagine improving things is to use the flag that shows up with each track list to indicate whether it's a new track or a continuation of an old one. With that information, you could try fading in new tracks on free oscillators in the bank to avoid the sudden discontinuity when frequency jumps by a huge amount. Data from a continuing track could safely be sent to the same oscillator. I guess you could also EQ it to get rid of those unpredictable high frequencies - I plotted the spectrogram and the most troublesome ones are above 2500Hz. Of course, that would also put a big damper on the timbre you're interested in...
I wish I had a great solution for you, but this is a tough one. Assigning frequencies to specific oscillators in your bank based on track flags really seems like the best bet.
On Fri, Feb 11, 2011 at 4:54 PM, J bz jbeezez@gmail.com wrote:
Right then,
I have only sent one sample only: '7_45_A_violadamore.wav' (7-string number, 45-midi number, A-note) which you will need to send into sigmund~ - I'm using [throw~ viola_samples] to get it in. From sigmund~ it goes into '[pd sine_output] then thrown to [sines] which has a switch~, top right.
I'm using GEM and MSD to create a 'swarm' which has the partials mapped
to
it so I haven't included my output~ as that would mean including loads of extra files.
So you will need an output~ to [catch~] the 2 track [throws~] [throw~ revsound_L] [throw~ revsound_R]
I suppose it's quite possible that as your not receiving any pan info the audio will be stuck either left or right anyway?
What I was trying to achieve with the [attdec_gen2] inside [sine_rev_gen_2] was a simple attack and decay but what is happening is that they are
being
triggered all the time, aargh.
I would really like the amp and freq from sigmund~ to be as quick and as smooth as possible, so any assistance with that will be hugely helpful.
As I'm not hugely confident in my programming abilities please point out
any
obvious mistakes and any and all tips and tricks/elegant solutions gratefully accepted. Really appreciate the help people.
Cheers,
Julian
On 11 February 2011 21:18, J bz jbeezez@gmail.com wrote:
Hey William,
Many thanks for offering to have a look at this for me. I'm currently trying to extricate the offending section out of what has now ballooned
up
into a very involved patch (certainly for me anyway). In the process of doing this I have realised that I am mistakenly constantly sending
attacks
and decays to the[osc~]'s. It would probably be best for me to just
send it
'as is' rather than trying to sort it out. I'm somewhat in over my head with this as it is (maybe not always a bad place to be) and could really
do
with some advice/help/coding-elegance as my brain is starting to melt.
Mathieu,
If you mean the tuning of the viola, yes it's the standard D Maj tuning, low to high A D A D F# A D.
Very best wishes,
Julian
P.S. Back soon
On 10 February 2011 23:16, Mathieu Bouchard matju@artengine.ca wrote:
On Thu, 10 Feb 2011, William Brent wrote:
I'm doing a piece with a viola d'amore (7 string viol)
Oops, that's not a violin :) I don't know the lowest note on that one offhand...
Is everybody only ever using the default tuning on those things ?
| Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal,
QC
-- William Brent www.williambrent.com
“Great minds flock together” Conflations: conversational idiom for the 21st century
www.conflations.com
I'm not completely clear about the continuation flags? I can see from the help file that they're there but what do they mean?
On every analysis period, several track reports are output in a burst. These provide the frequency and amplitude of what [sigmund~] considers the most important components in the spectrum. It does its best to find continuity between the components in the current frame and those in the previous frame. Of course, sometimes it fails to connect things nicely, sometimes there are new components that weren't there last time, and sometimes old components disappear. That's why each track report has its own flag. If it's a new track that flag will be 1. If it's a continuation of a previously existing track, the flag will be 0. So you can do some patching logic to make sure that continuing track information is always routed to the appropriate oscillator in your bank (i.e., the one that is already tuned to roughly that frequency). If you have a new track with a flag of 1, you'd want to send that track information to a silent (currently free) oscillator in your bank, and fade it up to the desired amplitude so that you don't hear a click from sudden change in frequency.
Perhaps this should be a new thread but why does sigmund have a frequency range of 100,000hz, and what would be a decent useable range from practical experience?
Strange - I never read that part of the help patch carefully. It's even stranger because it actually says the default maxfreq is 1,000,000 Hz, not 100,000. I'll assume there's a good reason for this that I'm just missing. But otherwise I think I'm understanding that setting in the same way you are: it puts a cap on the highest spectrum component that you want [sigmund~] to report. Assuming that's correct, I'd say you're safe capping it at 15kHz. That'll keep the number of oscillators you need practical without losing serious high end detail. But I think the best advice is to trust your ears...with some sounds (like speech) you might not be disturbed if you cap it at 10kHz.
Perhaps this should be a new thread but why does sigmund have a frequency range of 100,000hz, and what would be a decent useable range from practical experience?
Strange - I never read that part of the help patch carefully. It's even stranger because it actually says the default maxfreq is 1,000,000 Hz, not 100,000. I'll assume there's a good reason for this that I'm just missing. But otherwise I think I'm understanding that setting in the same way you are: it puts a cap on the highest spectrum component that you want [sigmund~] to report. Assuming that's correct, I'd say you're safe capping it at 15kHz. That'll keep the number of oscillators you need practical without losing serious high end detail. But I think the best advice is to trust your ears...with some sounds (like speech) you might not be disturbed if you cap it at 10kHz.
The default maximum frequency of a million means that, on most existing audio hardware, there's effectively no limit. Sigmund~ can never report a frequency above the Nyquist, so unless the sample rate is more than two MhZ the Nyquist is the effective limit. It's usually OK to leave it at the default, but for some instruments the pitch output is more stable when it's set lower.
cheers Miller
Thanks guys, that's really helpful - got things a bit straighter.
So can I say for clarification (and I noticed it came up on an earlier thread when (of the top of my head) someone was asking why the attack settings didn't work for tracks and they had to use [env~] - vibrato stabletime minpower growth
Have no effect when in 'tracks' mode./?
Not really sure what to do with the viola d'amore source sounds if they are just not going to be stable. Having the huge hop & npts settings seems kinda counter-productive and also introduces a significant delay. I guess if I want a smoother sound I could always sample and loop a small portion of the viola sound as he plays. My thinking atm is to go with the sound from sigmund 'cos that IS the sound.
BTW, I started out with fiddle~ and sigmund is much more responsive and more importantly better sounding.
Anyway back to the flags and my attacks and decays I guess.
Very best wishes,
Julian
On 16 February 2011 03:16, Miller Puckette msp@ucsd.edu wrote:
Perhaps this should be a new thread but why does sigmund have a
frequency
range of 100,000hz, and what would be a decent useable range from
practical
experience?
Strange - I never read that part of the help patch carefully. It's even stranger because it actually says the default maxfreq is 1,000,000 Hz, not 100,000. I'll assume there's a good reason for this that I'm just missing. But otherwise I think I'm understanding that setting in the same way you are: it puts a cap on the highest spectrum component that you want [sigmund~] to report. Assuming that's correct, I'd say you're safe capping it at 15kHz. That'll keep the number of oscillators you need practical without losing serious high end detail. But I think the best advice is to trust your ears...with some sounds (like speech) you might not be disturbed if you cap it at 10kHz.
The default maximum frequency of a million means that, on most existing audio hardware, there's effectively no limit. Sigmund~ can never report a frequency above the Nyquist, so unless the sample rate is more than two MhZ the Nyquist is the effective limit. It's usually OK to leave it at the default, but for some instruments the pitch output is more stable when it's set lower.
cheers Miller
Good to know about the reason for the 1Mhz maxfreq...
Not really sure what to do with the viola d'amore source sounds if they are just not going to be stable.
I think if you can settle for fewer harmonics and work out the flag logic there's hope yet. Another thing that occurred to me is that since the scratchy noise part of the viol sound is likely what's making the tracking jumpy, you could try to remove the noise. I've been working on an external that lets you rescale the amplitudes of harmonics in a pitched signal. There's an early draft on my website - it's called [pitchEnv~]. Using that, I scaled all the harmonics of your sample to 0, so that all the pitch related components are erased from the signal. The attached wav file is what's left...basically all the scratchy noise.
Then I tried subtracting that noise from the original signal (delayed slightly based on the time it takes to do this analysis), and it did clean up the noise some. It wasn't perfect, but with some work this could produce a signal you send to [sigmund~] that would be more easily tracked out. I might have time to fiddle with that later, but if you download [pitchEnv~] you can give it a shot too.
William
Hey William,
Appreciate the involvement...
Unfortunately I'm on W7 atm. My main patch uses the MSD lib and after what I thought was a trivial problem with using it in Puredyne (my regular distro), last Nov, MSD is still not working for me. Fortunately I bought a 2nd hand lappy with W7 on it early Dec, so have been working in that since. The instrumentalist is a Mac guy but that doesn't help me prepare the piece.
So yeah, a long way round of asking do you have a windows version?
Funnily enough I was nosy'ing around your site a couple of months ago and d/l'd the timbre ID lib for further research. Looks well good:)
Best wishes,
Julian
On 16 February 2011 19:02, William Brent william.brent@gmail.com wrote:
Good to know about the reason for the 1Mhz maxfreq...
Not really sure what to do with the viola d'amore source sounds if they
are
just not going to be stable.
I think if you can settle for fewer harmonics and work out the flag logic there's hope yet. Another thing that occurred to me is that since the scratchy noise part of the viol sound is likely what's making the tracking jumpy, you could try to remove the noise. I've been working on an external that lets you rescale the amplitudes of harmonics in a pitched signal. There's an early draft on my website - it's called [pitchEnv~]. Using that, I scaled all the harmonics of your sample to 0, so that all the pitch related components are erased from the signal. The attached wav file is what's left...basically all the scratchy noise.
Then I tried subtracting that noise from the original signal (delayed slightly based on the time it takes to do this analysis), and it did clean up the noise some. It wasn't perfect, but with some work this could produce a signal you send to [sigmund~] that would be more easily tracked out. I might have time to fiddle with that later, but if you download [pitchEnv~] you can give it a shot too.
William
Ok, a Windows binary of [pitchEnv~] is up on my site:
http://williambrent.conflations.com/pages/research.html#pitchEnv
I hadn't put up Windows or the source yet because I was still in the middle of working on it, but I think it's fine so that's all up now. Let me know off list if you have issues with it.
On Wed, Feb 16, 2011 at 4:03 PM, J bz jbeezez@gmail.com wrote:
Hey William,
Appreciate the involvement...
Unfortunately I'm on W7 atm. My main patch uses the MSD lib and after what I thought was a trivial problem with using it in Puredyne (my regular distro), last Nov, MSD is still not working for me. Fortunately I bought a 2nd hand lappy with W7 on it early Dec, so have been working in that since. The instrumentalist is a Mac guy but that doesn't help me prepare the piece.
So yeah, a long way round of asking do you have a windows version?
Funnily enough I was nosy'ing around your site a couple of months ago and d/l'd the timbre ID lib for further research. Looks well good:)
Best wishes,
Julian
On 16 February 2011 19:02, William Brent william.brent@gmail.com wrote:
Good to know about the reason for the 1Mhz maxfreq...
Not really sure what to do with the viola d'amore source sounds if they are just not going to be stable.
I think if you can settle for fewer harmonics and work out the flag logic there's hope yet. Another thing that occurred to me is that since the scratchy noise part of the viol sound is likely what's making the tracking jumpy, you could try to remove the noise. I've been working on an external that lets you rescale the amplitudes of harmonics in a pitched signal. There's an early draft on my website - it's called [pitchEnv~]. Using that, I scaled all the harmonics of your sample to 0, so that all the pitch related components are erased from the signal. The attached wav file is what's left...basically all the scratchy noise.
Then I tried subtracting that noise from the original signal (delayed slightly based on the time it takes to do this analysis), and it did clean up the noise some. It wasn't perfect, but with some work this could produce a signal you send to [sigmund~] that would be more easily tracked out. I might have time to fiddle with that later, but if you download [pitchEnv~] you can give it a shot too.
William
Good stuff, ok will do.
Thanks again,
Julian
On 17 February 2011 19:19, William Brent william.brent@gmail.com wrote:
Ok, a Windows binary of [pitchEnv~] is up on my site:
http://williambrent.conflations.com/pages/research.html#pitchEnv
I hadn't put up Windows or the source yet because I was still in the middle of working on it, but I think it's fine so that's all up now. Let me know off list if you have issues with it.
On Wed, Feb 16, 2011 at 4:03 PM, J bz jbeezez@gmail.com wrote:
Hey William,
Appreciate the involvement...
Unfortunately I'm on W7 atm. My main patch uses the MSD lib and after
what
I thought was a trivial problem with using it in Puredyne (my regular distro), last Nov, MSD is still not working for me. Fortunately I bought
a
2nd hand lappy with W7 on it early Dec, so have been working in that
since.
The instrumentalist is a Mac guy but that doesn't help me prepare the
piece.
So yeah, a long way round of asking do you have a windows version?
Funnily enough I was nosy'ing around your site a couple of months ago and d/l'd the timbre ID lib for further research. Looks well good:)
Best wishes,
Julian
On 16 February 2011 19:02, William Brent william.brent@gmail.com
wrote:
Good to know about the reason for the 1Mhz maxfreq...
Not really sure what to do with the viola d'amore source sounds if
they
are just not going to be stable.
I think if you can settle for fewer harmonics and work out the flag logic there's hope yet. Another thing that occurred to me is that since the scratchy noise part of the viol sound is likely what's making the tracking jumpy, you could try to remove the noise. I've been working on an external that lets you rescale the amplitudes of harmonics in a pitched signal. There's an early draft on my website - it's called [pitchEnv~]. Using that, I scaled all the harmonics of your sample to 0, so that all the pitch related components are erased from the signal. The attached wav file is what's left...basically all the scratchy noise.
Then I tried subtracting that noise from the original signal (delayed slightly based on the time it takes to do this analysis), and it did clean up the noise some. It wasn't perfect, but with some work this could produce a signal you send to [sigmund~] that would be more easily tracked out. I might have time to fiddle with that later, but if you download [pitchEnv~] you can give it a shot too.
William
-- William Brent www.williambrent.com
“Great minds flock together” Conflations: conversational idiom for the 21st century
www.conflations.com
-
J bz -
Mathieu Bouchard -
Miller Puckette -
William Brent