Hmmm,... 
phasors, filters, 1.8,... 

Try this, change the 1.8 to 0.8 or 0.9, 
keep any poles and zeros within the "unit circle". 
Something I've been reading lately in Julius Smith's 
books. 

Or, try running the results thru an Allpass Filter. 
I forget what the Csound opcode for that is. 

Maybe one of those or both will smooth it out. 

With speaker crossover networks, one gets notch 
distortion where the cut-off frequencies overlap. This 
horn thing between consecutive windows seems to be like 
that, but with inverse notch (or horns) in digital domain. 
Not sure if that makes any sense. 



Luck, 
-PBS 


===================================================


From:  	"Jeanette C." 
Reply-To:  	A discussion list for users of Csound 
To:  	CSOUND@LISTSERV.HEANET.IE
Subject:  	Re: [Csnd] Mathematics problem or Csound caveat?
Date:  	Thu 05/11/17 03:55 AM
 

May 11 2017, Partev Barr Sarkissian has written:

> Some window functions have what could be described as
> a horn at each end of the window function.
...
Hi,
I think this might exactly be what happened, due to the wrong optional
parameter to the Gaussian window function. I should have researched what
this parameter does, before experimenting wildly.

Best wishes,

Jeanette

--------
When you need someone, you just turn around and I will be there <3

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

==========================================================

--- encino_man@NETSCAPE.COM wrote:

From:         Partev Barr Sarkissian 
To:           CSOUND@LISTSERV.HEANET.IE
Subject: Re: [Csnd] Mathematics problem or Csound caveat?
Date:         Wed, 10 May 2017 21:26:34 -0700

Some window functions have what could be described as 
a horn at each end of the window function. I forget 
which kind of window. 

A friend of mine was using a window function on a 
continuous data stream. The horn peaks were somehow 
related to overlap between the previous window L-1 and 
the next window L+1. Left him with horns on either end. 

Not sure if or how it relates here, but if it's helpful,...  
 

-PBS 

===========================================


--- julien@MAIL.UPB.DE wrote:

From:         "Jeanette C." 
To:           CSOUND@LISTSERV.HEANET.IE
Subject: Re: [Csnd] Mathematics problem or Csound caveat?
Date:         Thu, 4 May 2017 13:35:31 +0200

May 3 2017, Oeyvind Brandtsegg has written:

> In m_risset_int.udo,
> replace
> kphasor wrap ((kphs * kdirfac) + (iphs_shift * iiter)), 0, 1
> kfm table3 kphasor*(ioctave/10), irise, 1, 0, 0
> with
> kphasor wrap (kphs  + (iphs_shift * iiter)), 0, 1
> kfm table3 kphasor*kdirfac*(ioctave/10), irise, 1, 0, 1
Hey hey Oeyvind,
thank you for that. I amended it slightly, I removed the wraparound
parameter from the table3 opcode to create kfm and instead created a
second phasor signal from the first with:
kfphasor wrap (kphasor *kdirfac)
That way the frequencies stay where they should be. Otherwise the
example with three octaves sounds very surprising. :)

I also found the error, that probably beat me all along: the parameter
to the wndow function. I set it to 1.8, which sounded good in my trials,
but caused the small peak of the high frequencies, when going down
somehow.

Thanks again, now I can finish and upload. :)


Best wishes,

Jeanette
...

--------
When you need someone, you just turn around and I will be there <3

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here




_____________________________________________________________
Netscape.  Just the Net You Need.

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here




_____________________________________________________________
Netscape.  Just the Net You Need.

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

Hi Partev,
thanks for your continued dedication to the problem. Maybe, I expressed myself 
a little badly, but the whole problem is solved. I released the UDO a few days 
ago.
http://juliencoder.de/sound/m_risset_filter-0.5.zip
As you can see in the demo CSD, a option of 1.0 to the Gaussian window is just 
perfect.

Thank you very much again.

Best wishes,

Jeanette

--------
When you need someone, you just turn around and I will be there <3

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here