Re: [Cs-dev] Inaccuracies in cpspch and cpsoct

Date	2007-08-24 04:58
From	Michael Gogins
Subject	Re: [Cs-dev] Inaccuracies in cpspch and cpsoct
	This would be fine with me too. Regards, Mike -----Original Message----- >From: Anthony Kozar >Sent: Aug 23, 2007 11:43 PM >To: Csound Developer list >Subject: Re: [Cs-dev] Inaccuracies in cpspch and cpsoct > >This is all beginning to sound to me like more work, more complexity, and >more confusion for a what amounts to a HYPOTHETICAL possibility that making >cpsoct/pch more accurate will actual cause somebody's existing music to >sound different. > >I know that I am the one who suggested the possibility, but we are talking >about differences in pitch of about 0.14 cents. This difference is >certainly not detectable in sequences of notes. It MAY cause small >differences in beating and timbres when notes (especially with many >overtones) are sounded together, but equal temperament already causes >significant beats to occur in most chords and this small change is likely to >go unnoticed too. > >The change is most likely to be noticeable only in combinations of long >sustained tones where the old pitches just happened to create a more lively >sound with stronger beats. (My example CSD exploited this possibility to >show how much of a problem I think the lookup table is; it is a pretty >unlikely scenario though ...) > >I believe the best solution is to make the pitch more accurate, and to >provide no backwards "compatibility" kludges. I now believe that to provide >such a mechanism complicates Csound unnecessarily and sets a bad precedent. >Some of the other recent changes to Csound (the new random number generator >or the filter changes) are MUCH more likely to produce noticeably different >results (especially the RNG). We cannot provide alternate sets of opcodes, >extra flags or orchestra syntax, etc. to turn these old versions off and on >without greatly increasing the complexity and manageability of the entire >system. Already, we have lots of users who have trouble getting the >existing opcode libraries to work and many people cannot compile or install >the system. > >The simplest solution for everyone is to keep the version of Csound that you >use to create a piece if you are concerned about future changes. All recent >versions of Csound will remain publicly available for a long time and John >has kept archives going back a lot further at the Bath site. > >Anthony > >Michael Gogins wrote on 8/23/07 4:15 PM: > >> I would just like us to give some thought as to which of these alternatives >> >> -- replace opcode names >> >> -- use legacy opcode library >> >> -- replace up to date names with legacy names >> >> is actually LEAST LIKELY TO GET MESSED UP and EASIEST TO MAINTAIN and EASIEST >> TO USE. >> >> I think the last alternative is the least likely to get messed up and the >> easiest to maintain, but it is not quite the easiest to use. > > >------------------------------------------------------------------------- >This SF.net email is sponsored by: Splunk Inc. >Still grepping through log files to find problems? Stop. >Now Search log events and configuration files using AJAX and a browser. >Download your FREE copy of Splunk now >> http://get.splunk.com/ >_______________________________________________ >Csound-devel mailing list >Csound-devel@lists.sourceforge.net >https://lists.sourceforge.net/lists/listinfo/csound-devel ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net

Date	2007-08-24 14:43
From	"Dr. Richard Boulanger"
Subject	Re: [Cs-dev] Inaccuracies in cpspch and cpsoct
	If ever John can't keep the repository at bath, I will be happy to host ALL the older versions and code at cSounds.com -dB On Aug 23, 2007, at 11:58 PM, Michael Gogins wrote: > This would be fine with me too. > > Regards, > Mike > > -----Original Message----- >> From: Anthony Kozar >> Sent: Aug 23, 2007 11:43 PM >> To: Csound Developer list >> Subject: Re: [Cs-dev] Inaccuracies in cpspch and cpsoct >> >> This is all beginning to sound to me like more work, more >> complexity, and >> more confusion for a what amounts to a HYPOTHETICAL possibility >> that making >> cpsoct/pch more accurate will actual cause somebody's existing >> music to >> sound different. >> >> I know that I am the one who suggested the possibility, but we are >> talking >> about differences in pitch of about 0.14 cents. This difference is >> certainly not detectable in sequences of notes. It MAY cause small >> differences in beating and timbres when notes (especially with many >> overtones) are sounded together, but equal temperament already causes >> significant beats to occur in most chords and this small change is >> likely to >> go unnoticed too. >> >> The change is most likely to be noticeable only in combinations of >> long >> sustained tones where the old pitches just happened to create a >> more lively >> sound with stronger beats. (My example CSD exploited this >> possibility to >> show how much of a problem I think the lookup table is; it is a >> pretty >> unlikely scenario though ...) >> >> I believe the best solution is to make the pitch more accurate, >> and to >> provide no backwards "compatibility" kludges. I now believe that >> to provide >> such a mechanism complicates Csound unnecessarily and sets a bad >> precedent. >> Some of the other recent changes to Csound (the new random number >> generator >> or the filter changes) are MUCH more likely to produce noticeably >> different >> results (especially the RNG). We cannot provide alternate sets of >> opcodes, >> extra flags or orchestra syntax, etc. to turn these old versions >> off and on >> without greatly increasing the complexity and manageability of the >> entire >> system. Already, we have lots of users who have trouble getting the >> existing opcode libraries to work and many people cannot compile >> or install >> the system. >> >> The simplest solution for everyone is to keep the version of >> Csound that you >> use to create a piece if you are concerned about future changes. >> All recent >> versions of Csound will remain publicly available for a long time >> and John >> has kept archives going back a lot further at the Bath site. >> >> Anthony >> >> Michael Gogins wrote on 8/23/07 4:15 PM: >> >>> I would just like us to give some thought as to which of these >>> alternatives >>> >>> -- replace opcode names >>> >>> -- use legacy opcode library >>> >>> -- replace up to date names with legacy names >>> >>> is actually LEAST LIKELY TO GET MESSED UP and EASIEST TO MAINTAIN >>> and EASIEST >>> TO USE. >>> >>> I think the last alternative is the least likely to get messed up >>> and the >>> easiest to maintain, but it is not quite the easiest to use. >> >> >> --------------------------------------------------------------------- >> ---- >> This SF.net email is sponsored by: Splunk Inc. >> Still grepping through log files to find problems? Stop. >> Now Search log events and configuration files using AJAX and a >> browser. >> Download your FREE copy of Splunk now >> http://get.splunk.com/ >> _______________________________________________ >> Csound-devel mailing list >> Csound-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/csound-devel > > > > > ---------------------------------------------------------------------- > --- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a > browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Csound-devel mailing list > Csound-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/csound-devel ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net