Re: [Csound] portamento algorithm

Re: [Csound] portamento algorithm

Date	2015-10-15 13:43
From	Tim Mortimer
Subject	Re: [Csound] portamento algorithm
	Hang on, just realised 1/freq might help ... "making a dill of yourself" seems a sure way to immediately solve a problem it seems. i'll work on that, but none the less feel free to jump in. that's probably not the whole story ... -- 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

Date	2015-10-17 02:18
From	Paul Batchelor
Subject	Re: [Csound] portamento algorithm
	> I'm trying to explore some more natural sounding portamento algorithms What sort of portamento algorithms have you looked at prior? There is a Csound journal article Steven Yi wrote on tied notes in Csound, which you should read if you haven't already: http://www.csounds.com/journal/2005fall/tiedNotes.html. There's also a chapter in the Csound book called "Designing Legato Instruments in Csound", which covers portamento. I personally find using a portamento filter on step-wise signals to work out just fine for me, but there are certainly interesting note interpolation methods that could be explored as well! -P On Thu, Oct 15, 2015 at 5:43 AM, Tim Mortimer <mortysrevenge@gmail.com> wrote: Hang on, just realised 1/freq might help ... "making a dill of yourself" seems a sure way to immediately solve a problem it seems. i'll work on that, but none the less feel free to jump in. that's probably not the whole story ... -- 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 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

Date	2015-10-17 20:46
From	Iain McCurdy
Subject	Re: [Csound] portamento algorithm
	A 'human' portamento slide is likely to spend a little time accelerating from the steady pitch to a constant rate of change. Similarly, when arriving at the destination pitch a short deceleration in the rate of change will occur. port/portk on their own can sound a bit 'laggy'; passing a lineto through a port filter with a short half-time can be more convincing. Also worth bearing in mind is that most acoustic instruments, while performing portamento, drop in power slightly. Date: Fri, 16 Oct 2015 18:18:24 -0700 From: ralphbluecoat@GMAIL.COM Subject: Re: [Csound] portamento algorithm To: CSOUND@LISTSERV.HEANET.IE > I'm trying to explore some more natural sounding portamento algorithms What sort of portamento algorithms have you looked at prior? There is a Csound journal article Steven Yi wrote on tied notes in Csound, which you should read if you haven't already: http://www.csounds.com/journal/2005fall/tiedNotes.html. There's also a chapter in the Csound book called "Designing Legato Instruments in Csound", which covers portamento. I personally find using a portamento filter on step-wise signals to work out just fine for me, but there are certainly interesting note interpolation methods that could be explored as well! -P On Thu, Oct 15, 2015 at 5:43 AM, Tim Mortimer <mortysrevenge@gmail.com> wrote: Hang on, just realised 1/freq might help ... "making a dill of yourself" seems a sure way to immediately solve a problem it seems. i'll work on that, but none the less feel free to jump in. that's probably not the whole story ... -- 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 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

Date	2015-10-18 09:35
From	Tim Mortimer
Subject	Re: [Csound] portamento algorithm
	Thanks Ian. I got the basic thing working. there was some slight improvement in longer portamento times, especially in an upward direction. I was intending to add the "accel / decel" at the beginning & end, but hadn't considered the "power" thing ... so that's worth a look ... Anyway, after much headscratching i managed to get the basic idea in place, & have put it aside to move onto to Python stuff today. The porto code is below ... I'm pretty happy with how my "whole system" is emerging now, having finally decided to bite the bullet & implement the thing i first started thinking about 6 or 7 years ago. I've had a bit more programming experience since then. & Pyops really has become essential to the whole process / approach.. I do have a few test bits & pieces on Soundcloud, but really its not much more than "pencil sketches": some generative phrases using Python driving sine waves mainly, & a little bit of emerging sound design ... https://soundcloud.com/timmortimer/sets/2015-2016-ideas-developments /* / -odac sr = 48000 ksmps = 128 nchnls = 2 0dbfs = 1.0 ;;; the "physical model" ;;; IS BETTER ;;; but only really comes into its own on long transitions ... instr 01 istartpitch = p4 iendpitch = p5 iportotime = p6 kcps init istartpitch ;;; stage 1 -> octave conversion iinterval = istartpitch-iendpitch iinterval_oct = iinterval /12 ;;; determine the "normalised exponential range" to be traversed ;;; (application revealed below...) ihigh = 1 ilow = 2^iinterval_oct idenominator = ilow-ihigh ;;; time pointer is normalised into 0 -1 range ktimepointer linseg 0,iportotime,1,0,1 knumerator = ktimepointeridenominator knumerator += 1 kinterval_oct = log2(knumerator) kinterval = kinterval_oct * 12 knote = istartpitch-kinterval kcps cpsmidinn knote printk 0.25,knote asig oscil 0.25,kcps aenv linseg 1,p3-0.5,1,0.5,0 asig = aenv outs asig,asig endin ;;; garden variety portomento ;;; **try normalising & applying power curve to this ... instr 02 knote linseg p4,p6,p5,0,p5 istartpitch = p4 iendpitch = p5 iportotime = p6 kcps init istartpitch iinterval = iendpitch-istartpitch ;;; normalised trajectory ktimepointer linseg 0,iportotime,1,0,1 ;;; shaped / if iinterval < 0 then ;;; starts fast ends slow ;;; => transition high to low ktimepointer = ktimepointer^0.45 endif if iinterval > 0 then ;;; starts slow ends fast ;;; => transition low to high ktimepointer = ktimepointer^2 endif / ktrans = iintervalktimepointer knote = istartpitch + ktrans kcps cpsmidinn knote printk 0.25,knote asig oscil 0.25,kcps aenv linseg 1,p3-0.5,1,0.5,0 asig *= aenv outs asig,asig endin ;;; starts fast ends slow ;;; => transition high to low i1 0 4 84 60 1 i2 5 4 84 60 1 ;;; starts slow ends fast ;;; => transition low to high i1 10 4 60 84 1 i2 15 4 60 84 1 -- View this message in context: http://csound.1045644.n5.nabble.com/Csound-portamento-algorithm-tp5744460p5744582.html Sent from the Csound - General mailing list archive at Nabble.com. 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

Date	2015-10-18 09:39
From	Tim Mortimer
Subject	Re: [Csound] portamento algorithm
	Hi Paul. Yes, I make constant reference to that Steven Yi article. That's the basis of most of my instrument designs now, in one form or other. Even without Portamento... i use tiestatus a lot to implement real-time note on / off functionality ... I'll dig out the csound book too & have a look ... -- View this message in context: http://csound.1045644.n5.nabble.com/Csound-portamento-algorithm-tp5744460p5744583.html Sent from the Csound - General mailing list archive at Nabble.com. 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