It seems that plltrack can increase CPU usage at very low input signal levels.
This would appear to be denormals related, but it does not disappear
if I denorm the signal before plltrack. However, if I lower the
tracking signal level threshold (kthresh) to e.g. 10^-20 and manually
denorm the input signal by adding noise at a level of 10^-19, then I
can prevent it from increased CPU usage.

I haven't been able to reproduce the behaviour in a scaled-down csd,
but I am quite sure I'm looking in the right place for the bug, since
- the problem does not appear if I comment out plltrack and replace it
with a static cps value,
- the problem also re-appears if I keep plltrack in there but replace
its output with a static value (meaning that the rest of the
instrument gets the static value for cps, and plltrack is just
processing a cps value that is never used for anything
- I can make the problem go away by fiddling with the kthresh value in
relationg to the noise floor set by a manual denorm
- The problem still appears if I set kthresh to 0.0, The kthresh value
needs to be a value larger than 10^-24, and I need to add noise at a
level above this threshold, then it behaves nicely.

Does this make any sense?
And if it can be fixed it will probably save someone else some time later :-)

best
Oeyvind

-- 

Oeyvind Brandtsegg
Professor of Music Technology
NTNU
7491 Trondheim
Norway
Cell: +47 92 203 205

http://flyndresang.no/
http://www.partikkelaudio.com/
http://soundcloud.com/brandtsegg
http://soundcloud.com/t-emp

------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

can you file an issue and we’ll look at it?
========================
Dr Victor Lazzarini
Dean of Arts, Celtic Studies and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952 

> On 28 Nov 2014, at 12:57, Oeyvind Brandtsegg  wrote:
> 
> It seems that plltrack can increase CPU usage at very low input signal levels.
> This would appear to be denormals related, but it does not disappear
> if I denorm the signal before plltrack. However, if I lower the
> tracking signal level threshold (kthresh) to e.g. 10^-20 and manually
> denorm the input signal by adding noise at a level of 10^-19, then I
> can prevent it from increased CPU usage.
> 
> I haven't been able to reproduce the behaviour in a scaled-down csd,
> but I am quite sure I'm looking in the right place for the bug, since
> - the problem does not appear if I comment out plltrack and replace it
> with a static cps value,
> - the problem also re-appears if I keep plltrack in there but replace
> its output with a static value (meaning that the rest of the
> instrument gets the static value for cps, and plltrack is just
> processing a cps value that is never used for anything
> - I can make the problem go away by fiddling with the kthresh value in
> relationg to the noise floor set by a manual denorm
> - The problem still appears if I set kthresh to 0.0, The kthresh value
> needs to be a value larger than 10^-24, and I need to add noise at a
> level above this threshold, then it behaves nicely.
> 
> Does this make any sense?
> And if it can be fixed it will probably save someone else some time later :-)
> 
> best
> Oeyvind
> 
> -- 
> 
> Oeyvind Brandtsegg
> Professor of Music Technology
> NTNU
> 7491 Trondheim
> Norway
> Cell: +47 92 203 205
> 
> http://flyndresang.no/
> http://www.partikkelaudio.com/
> http://soundcloud.com/brandtsegg
> http://soundcloud.com/t-emp
> 
> ------------------------------------------------------------------------------
> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
> with Interactivity, Sharing, Native Excel Exports, App Integration & more
> Get technology previously reserved for billion-dollar corporations, FREE
> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel


------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
https://lists.sourceforg

oki. done.


2014-11-28 14:09 GMT+01:00 Victor Lazzarini :
> can you file an issue and we'll look at it?
> ========================
> Dr Victor Lazzarini
> Dean of Arts, Celtic Studies and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
>> On 28 Nov 2014, at 12:57, Oeyvind Brandtsegg  wrote:
>>
>> It seems that plltrack can increase CPU usage at very low input signal levels.
>> This would appear to be denormals related, but it does not disappear
>> if I denorm the signal before plltrack. However, if I lower the
>> tracking signal level threshold (kthresh) to e.g. 10^-20 and manually
>> denorm the input signal by adding noise at a level of 10^-19, then I
>> can prevent it from increased CPU usage.
>>
>> I haven't been able to reproduce the behaviour in a scaled-down csd,
>> but I am quite sure I'm looking in the right place for the bug, since
>> - the problem does not appear if I comment out plltrack and replace it
>> with a static cps value,
>> - the problem also re-appears if I keep plltrack in there but replace
>> its output with a static value (meaning that the rest of the
>> instrument gets the static value for cps, and plltrack is just
>> processing a cps value that is never used for anything
>> - I can make the problem go away by fiddling with the kthresh value in
>> relationg to the noise floor set by a manual denorm
>> - The problem still appears if I set kthresh to 0.0, The kthresh value
>> needs to be a value larger than 10^-24, and I need to add noise at a
>> level above this threshold, then it behaves nicely.
>>
>> Does this make any sense?
>> And if it can be fixed it will probably save someone else some time later :-)
>>
>> best
>> Oeyvind
>>
>> --
>>
>> Oeyvind Brandtsegg
>> Professor of Music Technology
>> NTNU
>> 7491 Trondheim
>> Norway
>> Cell: +47 92 203 205
>>
>> http://flyndresang.no/
>> http://www.partikkelaudio.com/
>> http://soundcloud.com/brandtsegg
>> http://soundcloud.com/t-emp
>>
>> ------------------------------------------------------------------------------
>> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
>> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
>> with Interactivity, Sharing, Native Excel Exports, App Integration & more
>> Get technology previously reserved for billion-dollar corporations, FREE
>> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
> with Interactivity, Sharing, Native Excel Exports, App Integration & more
> Get technology previously reserved for billion-dollar corporations, FREE
> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel



-- 

Oeyvind Brandtsegg
Professor of Music Technology
NTNU
7491 Trondheim
Norway
Cell: +47 92 203 205

http://flyndresang.no/
http://www.partikkelaudio.com/
http://soundcloud.com/brandtsegg
http://soundcloud.com/t-emp

------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

thanks.
========================
Dr Victor Lazzarini
Dean of Arts, Celtic Studies and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952 

> On 28 Nov 2014, at 13:21, Oeyvind Brandtsegg  wrote:
> 
> oki. done.
> 
> 
> 2014-11-28 14:09 GMT+01:00 Victor Lazzarini :
>> can you file an issue and we'll look at it?
>> ========================
>> Dr Victor Lazzarini
>> Dean of Arts, Celtic Studies and Philosophy,
>> Maynooth University,
>> Maynooth, Co Kildare, Ireland
>> Tel: 00 353 7086936
>> Fax: 00 353 1 7086952
>> 
>>> On 28 Nov 2014, at 12:57, Oeyvind Brandtsegg  wrote:
>>> 
>>> It seems that plltrack can increase CPU usage at very low input signal levels.
>>> This would appear to be denormals related, but it does not disappear
>>> if I denorm the signal before plltrack. However, if I lower the
>>> tracking signal level threshold (kthresh) to e.g. 10^-20 and manually
>>> denorm the input signal by adding noise at a level of 10^-19, then I
>>> can prevent it from increased CPU usage.
>>> 
>>> I haven't been able to reproduce the behaviour in a scaled-down csd,
>>> but I am quite sure I'm looking in the right place for the bug, since
>>> - the problem does not appear if I comment out plltrack and replace it
>>> with a static cps value,
>>> - the problem also re-appears if I keep plltrack in there but replace
>>> its output with a static value (meaning that the rest of the
>>> instrument gets the static value for cps, and plltrack is just
>>> processing a cps value that is never used for anything
>>> - I can make the problem go away by fiddling with the kthresh value in
>>> relationg to the noise floor set by a manual denorm
>>> - The problem still appears if I set kthresh to 0.0, The kthresh value
>>> needs to be a value larger than 10^-24, and I need to add noise at a
>>> level above this threshold, then it behaves nicely.
>>> 
>>> Does this make any sense?
>>> And if it can be fixed it will probably save someone else some time later :-)
>>> 
>>> best
>>> Oeyvind
>>> 
>>> --
>>> 
>>> Oeyvind Brandtsegg
>>> Professor of Music Technology
>>> NTNU
>>> 7491 Trondheim
>>> Norway
>>> Cell: +47 92 203 205
>>> 
>>> http://flyndresang.no/
>>> http://www.partikkelaudio.com/
>>> http://soundcloud.com/brandtsegg
>>> http://soundcloud.com/t-emp
>>> 
>>> ------------------------------------------------------------------------------
>>> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
>>> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
>>> with Interactivity, Sharing, Native Excel Exports, App Integration & more
>>> Get technology previously reserved for billion-dollar corporations, FREE
>>> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>> 
>> 
>> ------------------------------------------------------------------------------
>> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
>> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
>> with Interactivity, Sharing, Native Excel Exports, App Integration & more
>> Get technology previously reserved for billion-dollar corporations, FREE
>> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
> 
> 
> 
> -- 
> 
> Oeyvind Brandtsegg
> Professor of Music Technology
> NTNU
> 7491 Trondheim
> Norway
> Cell: +47 92 203 205
> 
> http://flyndresang.no/
> http://www.partikkelaudio.com/
> http://soundcloud.com/brandtsegg
> http://soundcloud.com/t-emp
> 
> ------------------------------------------------------------------------------
> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
> with Interactivity, Sharing, Native Excel Exports, App Integration & more
> Get technology previously reserved for billion-dollar corporations, FREE
> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel


------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net