EMUSIC-L Digest Volume 49, Issue 02 This issue's topics: Additive Synthesis (12 messages) Cmusic and friends (2 messages) Additive synthesis 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: Tue, 16 Feb 1993 14:46:00 PST From: BISHOP_M@PLU.BITNET Subject: Additive Synthesis Hello, I was wondering if anyone out there is doing any research on additive sunthesis. It seems to me that it is the purest form, but yet hard to harness. I haven't seen any articles on it or mention of it in industy circles for a while. The Kawai K5 was the only commercial Additive synthesizer that I know of but it is not made any more. Does anyone know if another one would come out? Mike Bishop BISHOP_M@PLU.BITNET ------------------------------ Date: Tue, 16 Feb 1993 22:59:26 PST From: mifriese@CLN.ETC.BC.CA Subject: Re: Additive Synthesis > I was wondering if anyone out there is doing any research on additive >sunthesis. It seems to me that it is the purest form, but yet hard to harness. >I haven't seen any articles on it or mention of it in industy circles for a >while. The Kawai K5 was the only commercial Additive synthesizer that I know of >but it is not made any more. Does anyone know if another one would come out? Eirikur Hallgrimss writes: >If you are seriously into additive, you should look into packages that generate >sound through DSP on a computer. I believe that the "Turbosynth" program by Digidesign (for the Macintosh and possibly the Atari ST) features additive synthesis algorithms which can be used to generate samples subsequently sent to a sampler. One source is "Soundware" - 415-328-5773. **************************** Michael Friesen North Peace Secondary School Fort St. John, B.C. WorkVox: 604 785 4429 HomeVox: 604 787 9856 ------------------------------ Date: Wed, 17 Feb 1993 00:55:45 EST From: Eirikur Hallgrimsson 17-Feb-1993 0054 Subject: RE: Additive Synthesis Additive synthesis is not quite that rare. You haven't heard of the Seiko DS250/DS310, the Kurzweil 150 (way old), and the user wave feature on the Kawai K3. You won't hear much about additive synthesis or new machines of that sort. It's a picky, complicated, way to make a sound. Anything other than a static timbre is very hard to do. Sample players sound fine to most ears, and have much simpler user interfaces. If you are seriously into additive, you should look into packages that generate sound through DSP on a computer. Cmusic is one such. Eirikur ------------------------------ Date: Wed, 17 Feb 1993 08:19:05 PST From: metlay Subject: RE: Additive Synthesis > >> Additive synthesis is not quite that rare. You haven't heard of the Seiko >> DS250/DS310, the Kurzweil 150 (way old), and the user wave feature on the > Kawai >> K3. > >Hi, >I think that in The K3 the wave are looped single-cycle only .. could you >please elaborate about the user wave feature. >Additive synthesis is also possible (with some limitations) in the Korg DSS1 >were you can draw your own waves and on the EMAXII. Ignoring the additive capabilities of the Prophet VS, which is as much an "additive" machine as the DSS-1 in THAT respect.... I am shocked and appalled at this group, for dancing around the MEAT of the additive issue for so long without going for the metaphorical jugular! Has no one deigned to remember DIGITAL MOONSCAPES? BEAUTY IN THE BEAST? Or any other works of Our Lady of Eternal Perfectionism? >> It's a picky, complicated, way to make a sound. Anything other than a static >> timbre is very hard to do. Sample players sound fine to most ears, and have >> much simpler user interfaces. > >Complicated .. I agree .. but static timbre ... I don't . I do. True additive synthesis has evolving envelopes for EACH hamronic. You can't do this on a K5, but you CAN on the machines that Wendy Carlos used almost exclusively until SOB2k came along.... Shame, my friends! Shame! You have forgotten the ORIGINAL, no-compromise additive digital monster-- the DKI Synergy! Wired up to a Kaypro 4, it was a full-featured 32-voice additive synthesis workstation, a Crumar GDS with no knobs. La Carlos used them, and the rackmounted MuLogix Slave 32 that was their offspring, for over a decade before discovering the joys (?) of MIDI overkill. None of this K5 booshit... this is the real stuff, equalled only by the better computer-based packages. But would *I* do it? Hell no. Additive synthesis is the musical equivalent of building a ship in a bottle. With toothpicks. -- mike metlay | atomic city | "We'll be handing our money over to p. o. box 81175 | the Enemy, but at least they have a pittsburgh pa 15217-0675 | more varied menu." metlay@netcom.com | (sgt. metlay) ------------------------------ Date: Wed, 17 Feb 1993 09:14:19 PST From: metlay Subject: RE: Additive Synthesis >Oh goshhhhh shame on me for sure .. how could I have forgot the beasts of the >beasts !!! >But anyway Mike don't forget to mention the subtle difference in price between >the considered items ! Not so much any more, Andrea. If you can find a working Synergy, they're not taht expensive. >> But would *I* do it? Hell no. Additive synthesis is the musical equivalent >> of building a ship in a bottle. With toothpicks. > >Sometimes is nice to use mini-micro-tools to became mini-macro-crazy ! I'm mini-macro-crazy all by myself, thanks. -- mike metlay | atomic city | "We'll be handing our money over to p. o. box 81175 | the Enemy, but at least they have a pittsburgh pa 15217-0675 | more varied menu." metlay@netcom.com | (sgt. metlay) ------------------------------ Date: Wed, 17 Feb 1993 09:35:07 EST From: Ilia Subject: Additive synthesis >> If you are seriously into additive, you should look into packages th at >> generate sound through DSP on a computer. Cmusic is one such. > > May I ask how many harmonics can you have and how many env generators,with thi > kind of software ? And can you have waves different from sinus ? Cmusic is an acoustic compiler, which means it calculates all digital data prior to playing it. It has no real-time constraints, and hence, the number of oscillators involved in synthesis is limited only by the memory of your machine and your patience. I don't know the number of envelope generators provided by Cmusic, but since its output is merely a sequence of samples, and since Cmusic is an extention of C, it should be possible for you to add the desired features yourself, provided you have some programming experience. Ilia. ------------------------------ Date: Wed, 17 Feb 1993 21:13:00 PST From: BISHOP_M@PLU.BITNET Subject: Additive synthesis Ilia, Where can I get a copy of Cmusic? Is it "FTPable". Is it public domain or do I need to purchase it? It sounds like a VERY intriguing bit of software. Thanks a bunch, Mike Bishop BISHOP_M@PLU.BITNET ------------------------------ Date: Wed, 17 Feb 1993 09:55:30 GMT From: Andy Farnell Subject: Re: Additive Synthesis > > Hello, > I was wondering if anyone out there is doing any research on additive > sunthesis. It seems to me that it is the purest form, but yet hard to harness. > I haven't seen any articles on it or mention of it in industy circles for a > while. The Kawai K5 was the only commercial Additive synthesizer that I know of > but it is not made any more. Does anyone know if another one would come out? > > Mike Bishop > BISHOP_M@PLU.BITNET > Ah now there's a word I haven't seen on these pages since I don't know when. Not 'additive', 'synthesiser', 'hello', but 'research'. As it happens I am starting work on a fairly massive HCI project involving additive and several other synthesis forms. What we expect to acheive within a couple of years is a high level graphical synthesiser interface which allows visual drag and drop, knob twiddling editing of an 'abstract synthesiser'. Because there are so many synthesis forms, all of which are in theory (where you make provision for small matters like having an infinite number of certain components) mathmatically equivilent, it is possible to map from a sound representation using one hardware specific parameter set onto another. An abstract synthesiser is simply a virtual device which has a generic abstract sound representation (which is a set of points in the phase space of n-dimentional structural, timbral and performance parameters. Of course all synthesis programming is the domain of expert knowledge about how to create a sound and unless you are simply experimenting and playing around on a piece of hardware you really need a hell of a lot of deep knowledge about the tools you have in order to form a mental representation both of the current hardware state and the goal state. Abstract sound manipulation makes the process hardware transparent so that you only need to have high level knowledge about a how to focus on your goal state. The rest is done using a mixture of expert system shells and neural networks (self organising Kohonen classifiers) - but that is the tricky bit. With that short expanation dispatched I shall move onto what I really want to say, or rather ask.. I am curious just how many other researchers are out there on this net. Yes there are lots of users, composers and musos as well I know, but is there anyone else out there into some real synthesis fundamentals. Going back to Marks question pertaining to the K5 - this is an extremely powerful device on paper but it suffers two small problems, firstly the quality of the sources is not good. I suspect they originate from 8 bit samples in an internal wavetable and are a bit noisy. Secondly it is a bit of a bastard to program unless you are used to additive synthesis. (It is very difficult to program a good performance profile for a sound where you have 128 harmonics and only 8 envelopes to asign to them.) It originally failed (in England at least) to sell widely due to some pretty sad factory presets and a lack of internal effects (which most of its immediate competition had and which was a real novelty selling point at that time). I think it is an overlooked diamond of a synth. I heard rumors about 3 or 4 years ago about some software to convert samples into K5 parameter sets and also DX7 FM parameter sets into K5 formats. I don't know if this was ever a commercial product but the guy who told me was a MT researcher at York University at the time. Anyway it shouldn't be too hard to at least partly automate the process. All you need is a pretty basic FFT program to feed your samples into and set the harmonics to match for the K5. Extracting the envelopes might be a little more tricky, especially mapping them nicely onto the 6 stages or whatever the K5 has. FM parameter translation is a tad more difficult since it involves inverse Bessel functions and Laplace transformations. Happy synthesising Andy.James.(flange)Farnell Department of Computing and Cognition Bournemouth University England ------------------------------ Date: Wed, 17 Feb 1993 09:52:34 -0500 From: "Joseph D. McMahon" Subject: Re: Additive Synthesis > What we expect to acheive within a couple of years is a high > level graphical synthesiser interface which allows visual drag and drop, knob > twiddling editing of an 'abstract synthesiser'. Now this is an interesting project. Would you mind keeping us up to date on how things work out? At the inception of EMUSIC-L, we had quite a lot of discussion of a proposed timbre description language, but we bogged down on how one would accurately describe the actual effect of circuit-board layout on a Moog-style filter (for instance). What is your abstraction model for defining synthesis techniques? Are you attempting to sound like specific synthesizers, or are you going for generic methods? Will you be implementing some of the more esoteric synthesis methods (Karplus-Strong and automata synthesis)? --- Joe M. ------------------------------ Date: Wed, 17 Feb 1993 10:44:45 MET From: Andrea TONI Subject: RE: Additive Synthesis > Additive synthesis is not quite that rare. You haven't heard of the Seiko > DS250/DS310, the Kurzweil 150 (way old), and the user wave feature on the Kawai > K3. Hi, I think that in The K3 the wave are looped single-cycle only .. could you please elaborate about the user wave feature. Additive synthesis is also possible (with some limitations) in the Korg DSS1 were you can draw your own waves and on the EMAXII. > It's a picky, complicated, way to make a sound. Anything other than a static > timbre is very hard to do. Sample players sound fine to most ears, and have > much simpler user interfaces. Complicated .. I agree .. but static timbre ... I don't . In the K5 for example the most common sounds available are a mere conversion of samples and the 4(8) Envelope generator plus 1 for the DDF and 1 for the DDA are never used. If you start to add reasonable amount of envelops the sound will certainly "move" .. now as I said it is complicated and it is a lot of work .. but it works ! The problem is that in theory you need as many env generator as the harmonics (128 on the K5 ! .. but 256 would be even better .. !). Of course to handle such a beast would be a nightmare ! > If you are seriously into additive, you should look into packages that generate > sound through DSP on a computer. Cmusic is one such. May I ask how many harmonics can you have and how many env generators,with this kind of software ? And can you have waves different from sinus ? Ciao, ------------------------------------------------ENV---ENV---ENV---------------- Andrea TONI (andrea@sihp03.si.estec.esa.nl) | | | Planetary and Space VCO-->VCF-->VCA--> DUCATI 900SS Science Division (SI) | | | ------------------------------------------------LFO---LFO---LFO---------------- ------------------------------ Date: Wed, 17 Feb 1993 17:53:32 MET From: Andrea TONI Subject: RE: Additive Synthesis > I do. True additive synthesis has evolving envelopes for EACH hamronic. You > can't do this on a K5, but you CAN on the machines that Wendy Carlos used > almost exclusively until SOB2k came along.... I do agree on the fact that true additive synthsis requires separate envs for every harmonics .. but even if you can't do this on a K5 I still believe that the K5 doesn't sound static if you use properly the few ones that it has ! > Ignoring the additive capabilities of the Prophet VS, which is as much an > "additive" machine as the DSS-1 in THAT respect.... Point taken ! > Shame, my friends! Shame! You have forgotten the ORIGINAL, no-compromise > additive digital monster-- the DKI Synergy! Wired up to a Kaypro 4, it > was a full-featured 32-voice additive synthesis workstation, a Crumar > GDS with no knobs. La Carlos used them, and the rackmounted MuLogix > Slave 32 that was their offspring, for over a decade before discovering > the joys (?) of MIDI overkill. None of this K5 booshit... this is the > real stuff, equalled only by the better computer-based packages. Oh goshhhhh shame on me for sure .. how could I have forgot the beasts of the beasts !!! But anyway Mike don't forget to mention the subtle difference in price between the considered items ! > But would *I* do it? Hell no. Additive synthesis is the musical equivalent > of building a ship in a bottle. With toothpicks. Sometimes is nice to use mini-micro-tools to became mini-macro-crazy ! Ciao, ------------------------------------------------ENV---ENV---ENV---------------- Andrea TONI (andrea@sihp03.si.estec.esa.nl) | | | Planetary and Space VCO-->VCF-->VCA--> DUCATI 900SS Science Division (SI) | | | ------------------------------------------------LFO---LFO---LFO---------------- ------------------------------ Date: Thu, 18 Feb 1993 14:05:28 EST From: Ilia Subject: Additive synthesis Mike Bishop writes: > Ilia, > Where can I get a copy of Cmusic? Is it "FTPable". Is it public domai > or do I need to purchase it? It sounds like a VERY intriguing bit of software. > To my knowledge, Cmusic is not a public domain software, and it may be purchased from UCSD as a part of their Music Development Kit. This information came from the book by F.R. Moore, one of the developers of Cmusic. I personally never had any hands-on experience with the program, and my information is based entirely upon what I read about it. Ilia. ------------------------------ Date: Fri, 19 Feb 1993 10:33:06 +0000 From: Ferdinando Villa Subject: Cmusic and friends Cmusic requires a $250 lcense fee from CARL. Anyway, there are two great packages which are available for free. I've been told they are at least as powerful as Cmusic. The one I have and enjoy a lot is csound, which runs on SPARCs, DEC, NEXT and others but also has a Mac version which works niftly. It allows so many things that I don't want to mention any here - but you can retrieve the FAQ from the comp.music newsgroup which discusses the whole issue. On the right hardware it can operate in real time. I use it on a Mac LC with coprocessor and it's usable. I plan to build software to export waveforms to my TG500. csound is available from cecelia.media.mit.edu. The second package I know of (never tried) is cmix, also available at a number of ftp sites. I heard pretty good opinions about it. It is also free and comes in precompiled versions for SUNs, NEXTs and IRIX. Hope this helps, ferdinando -- Ferdinando Villa, dr. Institute of Ecology Direct phone: +39-521-905615 University of Parma FAX: +39-521-905665 Viale delle Scienze e.mail: villa@sparrow.bio.unipr.it 43100 Parma, Italy villa@aida.eng.unipr.it ------------------------------ Date: Fri, 26 Feb 1993 13:03:30 GMT From: J P Fitch Subject: Re: Cmusic and friends Csound is also available for 80386/80486 and for 80286 and it ftp'able from ftp.bath.ac.uk in directory pub/jpff ==John ffitch ------------------------------ Date: Wed, 17 Feb 1993 09:35:49 -0500 From: "Joseph D. McMahon" Subject: Additive synthesis Andrea TONI asks about the limitations of cmusic. The limitations are mostly those imposed by memory size and how much code you are willing to write and how long you are willing to wait for results. Cmusic is totally computationally-based; the sound generation is done by calculating the final output and running it directly into a D-to-A converter. It gives you unparalleled control over every aspect of the sound, but it's a very nonphysical way of making music. There's no such thing as an improvised cmusic solo. :-) --- Joe M. ------------------------------ Date: Sun, 28 Feb 1993 08:23:57 GMT From: rayhana cherif Subject: CSound Books ? I'm looking for publications or books on CSound OTHER THAN the CSound manual by B. Vercoe ? In addition, I would like to know if there is any list of publications (for the last few years) of the Centre for Computer Research in Music and Acoustics (CCRMA) - Stanford. Your help is highly appreciated. Please email answer to: rayhana@byron.u.washington.edu Thanks, ------------------------------ End of the EMUSIC-L Digest ******************************