EMUSIC-L Digest Volume 54, Issue 11 This issue's topics: (fwd) YAMAHA DX,SY series FM emulator for PC Chowning (6 messages) copy protection fm parameterization FM synthesis (2 messages) FM synthesis DX-7 simulator (3 messages) FM synthesis in Software (3 messages) my FM synthesis mystery Your EMUSIC-L Digest moderator is Joe McMahon . You may subscribe to EMUSIC-L by sending mail to listserv@american.edu with the line "SUB EMUSIC-L your name" as the text. The EMUSIC-L archive is a service of SunSite (sunsite.unc.edu) at the University of North Carolina. ------------------------------------------------------------------------ Date: Fri, 9 Jul 1993 08:50:00 EDT From: Jeff Harrington Subject: (fwd) YAMAHA DX,SY series FM emulator for PC HELLO ! I` m working on a public domain emulator of YAMAHA series DX... series SY...,V50 and so on .. That FM emulator will work on a personal computer IBM compatible (minimum 486 i think),and makes 16 or 8 bit samples for use with mods or better sample-sequencer. I` ve tried to find the file format of the saved parameters but in the manual of such Keyboards there` s not too much for working. If someone has something about DX7 ,DX21 ,V50, SY77,SY99 please tell me as soon as possible,expecially about the file structure,the bit resolution of the oscillators ,and if someone has a file version of the preset samples (for the SY) Please post me brm@paola.dei.unipd -- (*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*) (*) IdEAL ORDER Psychic TV (*) Zappin' CBS Evening News (*) (*) idealord@dorsai.dorsai.org (*) Since 1983! (*) (*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*)(*) ------------------------------ Date: Fri, 2 Jul 1993 09:16:19 CDT From: crispen Subject: Re: Chowning mbartkow@GWENDU.ENST-BRETAGNE.FR >Is there anybody on the world, who can help me to find a book >"Foundations of Sound Synthesis" by John Chowning ? These are the only references I have for Chowning: J. Chowning "The Synthesis of Complex Audio Spectra by Means of Frequency Modulation" J. Audio Eng. Soc. 21, no. 7 (1973):526-534 [This is the standard reference.] J. Chowning and D. Bristow "FM Theory and Applications by Musicians for Musicians" Yamaha Music Foundation, Tokyo, 1986 ISBN 4-636-17482-8 I don't claim this is complete, though. If you were referring to something else, then I'd like to know about it, too. +-------------------------------+--------------------------------------+ | Rev. Bob "Bob" Crispen | Music should not be held responsible | | crispen@foxy.boeing.com | for the people who listen to it. | +-------------------------------+--------------------------------------+ ------------------------------ Date: Fri, 2 Jul 1993 14:47:57 EDT From: ronin Subject: chowning n fm um... almost... the book you're looking for is Foundations of Computer Music, edited by Curtis Roads. it is a collection of Computer Music Journal reprints, and Chowning's article is included, as well as some extensions. as for sample->fm parameters, it has i believe been done. get in touch with IRCAM. that is, if you want the utility itself, rather than the fun and experience of figuring it out yourself. if you do go that route, btw, (the homebrew way, that is), let us know what strategy you decide on. i think the last thing discussed here (by IDEALORD@DORSAI.DORSAI.ORG - you might want to contact him directly) had something to do with genetic-algorithm search of optimal parameters via bessel function lookup. -----------< Cognitive Dissonance is a 20th Century Art Form >----------- Eric Harnden (Ronin) or The American University Physics Dept. 4400 Mass. Ave. NW, Washington, DC, 20016-8058 (202) 885-2748 (with Voice Mail) ---------------------< Join the Cognitive Dissidents >------------------- ------------------------------ Date: Fri, 2 Jul 1993 15:17:44 EDT From: mbartkow@GWENDU.ENST-BRETAGNE.FR Subject: Chowning Hi! Is there anybody on the world, who can help me to find a book "Foundations of Sound Synthesis" by John Chowning ? -------------------------------------------------------- Maciej Bartkowiak, Technical University of Poznan / POLAND Temporarily at Ecole Nationale Superieure des Telecommunications de Bretagne / FRANCE ------------------------------ Date: Fri, 2 Jul 1993 18:38:29 EDT From: mbartkow@GWENDU.ENST-BRETAGNE.FR Subject: Re: Chowning ** >Is there anybody on the world, who can help me to find a book ** >"Foundations of Sound Synthesis" by John Chowning ? ** These are the only references I have for Chowning: ** J. Chowning ** "The Synthesis of Complex Audio Spectra by Means of Frequency ** Modulation" ** J. Audio Eng. Soc. 21, no. 7 (1973):526-534 ** [This is the standard reference.] ----> yes, but in "Foundations..." (where nb. this paper is reprinted) there are his newer concepts of more complexive FM synthesis, which I'm looking for... ** J. Chowning and D. Bristow ** "FM Theory and Applications by Musicians for Musicians" ** Yamaha Music Foundation, Tokyo, 1986 ** ISBN 4-636-17482-8 ----> thanx for this, it sounds interesting, especialy because of YAMAHA's activity in the field of FM. In addition: I have found an interesting paper about reconstruction algorithm for FM natural sound simulating. If anybody is still interested in FM synthesis, but in software means, please to contact. I'm going to work at smart program for converting sampled sound to DX's programm parameters. cheers- Maciej ------------------------------ Date: Tue, 6 Jul 1993 01:17:14 EDT From: Graham Collins Subject: Re: Chowning In a previous article, mbartkow@GWENDU.ENST-BRETAGNE.FR () says: >Hi! > >Is there anybody on the world, who can help me to find a book >"Foundations of Sound Synthesis" by John Chowning ? You might consider contacting Dartmouth University as I believe he teaches there. Graham Collins -- If the environment is all that it's cracked up to be, why are the homeless complaining? P.J. O'Rourke ------------------------------ Date: Tue, 6 Jul 1993 16:17:21 -0600 From: Mike Bishop Subject: Re: Chowning On Tue, 6 Jul 1993, Graham Collins wrote: > You might consider contacting Dartmouth University as I believe he > teaches there. > Graham Collins I thought that it was Stanford. ------------------------------ Date: Thu, 1 Jul 1993 10:23:09 U From: Terry Barnum Subject: Re: EMUSIC-L Digest - 29 Jun >I can't find (well three different bookshops couldn't find) that locus >classicus, _Elements of Computer Music_ by Richard Moore, in either the >British or American _Books In Print_. Try Mix Bookshelf: 6400 Hollis St., #12-B Emeryville, CA 94608 vox 800.233.9604 vox 510.653.3307 fax 510.653.3609 terry@voltar.imagitrek.gte.com ------------------------------ Date: Wed, 14 Jul 1993 14:12:30 EDT From: ronin Subject: fm parameterization as has been discussed, the issue with anything-to-fm conversion is one of fit measurement. the first decision is which domain to operate in, time or frequency. personally, i never tried it in the time domain, believing formant content to be critical, and suspecting that perceptable differences in content might be outside the meaningful threshold of any kind of graphic comparison technique. unfortunately, this opens a whole new can of invertebrates... having discarded graphic comparison, it then becomes necessary to quantify 'good enough' for numeric comparison of spectral analyses. this leads to a requirement for psychoacoustic testing, which i began with a colleague some years ago. that project went the way of all graduate work, but the ideas, papers, and software are all still sitting around, if you find merit in the approach and want to talk about it. with exogenous impetus, i think i might be able to reconstruct the core of the argument. -----------< Cognitive Dissonance is a 20th Century Art Form >----------- Eric Harnden (Ronin) or The American University Physics Dept. 4400 Mass. Ave. NW, Washington, DC, 20016-8058 (202) 885-2748 (with Voice Mail) ---------------------< Join the Cognitive Dissidents >------------------- ------------------------------ Date: Tue, 6 Jul 1993 18:19:17 EDT From: mbartkow@GWENDU.ENST-BRETAGNE.FR Subject: Re: FM synthesis In answer of my: > If anybody is still interested in FM > synthesis, but in software means, please to contact. I'm going to work > at smart program for converting sampled sound to DX's programm > parameters. johndu@world.std.com writes: ] I'm still interested in FM software synthesis! If you do develop ] the program you mentioned, I'd be very interested in it. Will it be ] written ] for the Macintosh? In C? <> ] Thanks -- John Duesenberry ] johndu@world.std.com The program will be written in PASCAL (first), in C (maybe) and I am going to give it to the LIST as a source code (I am using actually PC486/50 and VAX station). I have already said here, it'll take me a lot of time (it's very complicated), so please wait. I'm an electronic engineer, working at the Technical University of Poznan /POLAND, not a musician. I do not have a lot of standard equipment (self-made rather). First I am waiting for the DX-7 simulator on the LIST's file server (because I don't have one for my own) If anybody could help me to find: 1) a precision DX-7 description (program parameter's function, especially the real modulation depth values and precision frequency table) 2) full technical specification 3) some good, spectacular sample examples It will be very helpfull. best regards- Maciej Bartkowiak temporarily: mbartkow@gti.enst-bretagne.fr in POLAND: mbartkow@cygnus.ieitpp.poz.edu.pl ------------------------------ Date: Thu, 8 Jul 1993 08:07:27 -0400 From: Andy Farnell Subject: Re: FM synthesis > johndu@world.std.com writes: > > ] I'm still interested in FM software synthesis! If you do develop > ] the program you mentioned, I'd be very interested in it. Will it be > ] written > ] for the Macintosh? In C? <> > ] Thanks -- John Duesenberry > ] johndu@world.std.com > > The program will be written in PASCAL (first), in C (maybe) and I am > going to give it to the LIST as a source code (I am using actually > PC486/50 and VAX station). I have already said here, it'll take me a lot > of time (it's very complicated), so please wait. > This is good stuff guys, I hope you will keep me posted on the developments. I hope you realise this is a NON TRIVIAL problem. How are you going yo analyse the samples? Will you map a spectral map using Bessel functions onto a parameter matrix for a fixed architecture? Can you guaruntee that it will find the best parameter set. How will you extract envelope features and handle significant discontinuities How will it handle vibrato and unpitched sounds such as drums Perhaps we should talk about this if you are really serious, it would be a good point for exchanging some ideas. I am working on a totally generic system which will map a sample onto ANY synthesis architecture as long as it is known in advance. This is a long running research/PhD project. You might like to look at: De Poli et al. (ed C.Roads) Representations of Music Signals. MIT press Andy Farnell Computing and Cognition Bournemouth University England ------------------------------ Date: Fri, 9 Jul 1993 08:42:45 -0400 From: idealord Subject: Re: FM synthesis DX-7 simulator Maciej continues -> > Jeff Harrington writes: > > > have you seen Andrew Horner's research using GA's and > > neural nets to do this (CSound not DX7). > > I have not seen and am very interested. Could you tell > something more ? > I've been "paraphrasing" his research on this list for about 3 months in a totally incompetent fashion - I'm sure - so I'll continue - I got a pre-publication article after talking with someone who attended last summer's CERL sound group workshop - actually two articles - one is on the use of genetic algorithms to match FM C:M indices and envelopes and the other is on his use of neural nets to accomplish the same thing. He's using Csound instruments not DX7 patches ;-( but his GA work was very provocative. The technique is pretty simple and would be easy to simulate with software simulation of a DX7 (or even better a SY77) and samples. He uses simple 1 and 2 modulator FM - so to use a 6 op FM would probably require much more time for the simulated annealing stuff to settle. My GA work has been primarily as a front end to my counterpoint expert system to find spacings between periods which would produce more combinations of counterpoint - and it is slow... on my Amiga 1000... Start with a random population of FM patches - do a FFT or other type of spectrum analysis which would allow for matching comparisons - critique the FM patches with regard to closeness of match - let the FM patches which came close reproduce. > > WHAT DX7 software simulator? Who wrote it and where can I > > get it? Is source available? > > I just sent GET EMUSIC FILELIST to the > listserv@american.edu, and got a list of avaliable > discussion archives, MIDIspec packages, D50 libraries, and > others, like "Small Algorithmic Composition Program" and > just "DXTUTOR"-DX7 Simulator . Unfortunately, that > last is not still avaliable (the INFOFORM too) > > Maciej Bartkowiak > Hmmm... does anybody know what this DXTUTOR do? Jeff Harrington idealord@dorsai.dorsai.org ------------------------------ Date: Fri, 9 Jul 1993 09:31:32 -0400 From: Joe McMahon Subject: Re: FM synthesis DX-7 simulator I've dropped a note to its author about the DX simulator; I'll report anything he has to tell me about it. --- Joe M. ------------------------------ Date: Fri, 9 Jul 1993 12:20:27 EDT From: mbartkow@GWENDU.ENST-BRETAGNE.FR Subject: Re: FM synthesis DX-7 simulator Jeff Harrington writes: > have you seen Andrew Horner's research using GA's and > neural nets to do this (CSound not DX7). I have not seen and am very interested. Could you tell something more ? > WHAT DX7 software simulator? Who wrote it and where can I > get it? Is source available? I just sent GET EMUSIC FILELIST to the listserv@american.edu, and got a list of avaliable discussion archives, MIDIspec packages, D50 libraries, and others, like "Small Algorithmic Composition Program" and just "DXTUTOR"-DX7 Simulator . Unfortunately, that last is not still avaliable (the INFOFORM too) Maciej Bartkowiak ------------------------------ Date: Wed, 7 Jul 1993 12:14:06 -0400 From: idealord Subject: Re: FM synthesis in Software Mr. Bartkowiak writes -> > In answer of my: > > > If anybody is still interested in FM > > synthesis, but in software means, please to contact. I'm going to work > > at smart program for converting sampled sound to DX's programm > > parameters. Sounds great - have you seen Andrew Horner's research using GA's and neural nets to do this (CSound not DX7). > The program will be written in PASCAL (first), in C (maybe) and I am > going to give it to the LIST as a source code (I am using actually > PC486/50 and VAX station). I have already said here, it'll take me a lot > of time (it's very complicated), so please wait. > > I'm an electronic engineer, working at the Technical University of Poznan > /POLAND, not a musician. > > I do not have a lot of standard equipment (self-made rather). First I am > waiting for the DX-7 simulator on the LIST's file server (because I don't > have one for my own) > WHAT DX7 software simulator? Who wrote it and where can I get it? Is source available? Jeff Harrington idealord@dorsai.dorsai.org ------------------------------ Date: Fri, 9 Jul 1993 06:51:22 -0400 From: Andy Farnell Subject: Re: FM synthesis in Software > > Sounds great - have you seen Andrew Horner's research using GA's and neural > nets to do this (CSound not DX7). > Do you have any details of this research, has anything been published? Andy Farnell Computing and Cognition Bournemouth University England ------------------------------ Date: Fri, 9 Jul 1993 08:44:05 -0400 From: idealord Subject: Re: FM synthesis in Software > > > > > Sounds great - have you seen Andrew Horner's research using GA's and neural > > nets to do this (CSound not DX7). > > > > Do you have any details of this research, has anything been published? I've got a pre-publication of two articles - supposed to be published in CMJ soon... Jeff Harrington idealord@dorsai.dorsai.org ------------------------------ Date: Fri, 9 Jul 1993 16:20:08 EDT From: mbartkow@GWENDU.ENST-BRETAGNE.FR Subject: my FM synthesis mystery O'KAY I see it's getting warm around the subject (I didn't expect it, I swear. I was going to begin the stuff in some months, when I'd come back to POLAND, where I left all the papers) So here are some hints (sorry for my poor english, I'm learning french just now, it's completely different) * the first question is the best parameter set: Of course I don't figure to find it (which is the criterion? It's stupid to put human hearing-decicing in a feedback loop of the optimization algorithm) * I wasn't going to use Bessel functions at all, because I'd work in time domain rather. But concidering another's research subjects it's gonna be quite interesting. I would be grateful for any informations (as well as NN based methods) * My first step would be to analyse the sample in spot of analytic signal model. Then you can extract the envelope, the phase trend (-> carrier frequency) and the instantaneous phase variation (->the modulator). Thus far you can obtain whole nested modulator fixed chain. This would be good start point for the optimization process in discrete parameter space. * Using feedbacks and multiplied sources (like in DX7 most algorithms) you can obtain a nonlinear equation set for the optimization. It's of course more difficult, but much more interesting too. I figure, prof.J.M.Chowning has considered the effects of feedbacks in FM synthesis in his later papers, that's why I'm still looking for his "Foundations of Sound Synthesis" and "Foundations of Computer Music" (there are two books, I'm sure) I emphasize, the idea of analytic signal processing is not mine. It has been published in some papers (a.a. in IEEE Transactions, which I already mentioned) regards, ------------------------------ End of the EMUSIC-L Digest ******************************