EMUSIC-L Digest Volume 51, Issue 11 This issue's topics: "Quest of the Holly Interface" Sound Morphing (14 messages) 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, 23 Apr 1993 05:16:18 CDT From: Mark Borcherding Subject: Re: "Quest of the Holly Interface" I use a Voyetra v24s with 2 midi in/4 midi out with Cakewalk windows with no problem. I have gotten driver updates to fix problems but I think thats always going to be the case until everyone understands windows enhanced mode. Voyetra does have a MPU option which I do not have. I do have and use the SMPTE option which is on the v24s. Note: I also have a soundblaster pro with midi installed and running at the same time. Mark Borcherding. ------------------------------ Date: Tue, 13 Apr 1993 18:54:58 CDT From: John Eichenseer Subject: Re: K2000 FM etc. >I seem to remember something like this being explained to me when I >saw my first K250 in '84. Is this possible with the K2000? Is it >possible at all? If so, just what envelope are you modulating with? >Some sort of "peak apmlitude over the past few miliseconds" or the raw >sample (seems unlikely)? I don't tjink you can do this sort of thing with the K2000, or any synth. This sounds like the kind of sophisticated processing that you can do in csound or low-level DSP programming. >Or do you translate the samples into the >frequency domain and modulate there and translate the result back This is called Fourier analysis and resynthesis, which the K2000 does not do. Fancy resynthesis hardware is pretty expensive stuff; you can do limited fourier editing with software packages like Alchemy. There are a million ways to modulate sounds, from vocoding to AM to FM to pitch shifting. The Kurzweil does some great stuff, but there is no magic "modulate sound 1 with sound 2" function. The fanciest DSP stuff added by the sampling option appears to be the time stretching/compressing. Lots to work with, though. Heh - I wish there was a sampler with a magical "Transmogrify sound 1 into sound 2 over n seconds" button. A friend of mine was hoping to do this with neural nets; I don't think he got very far. This kind of transformation would be a great way to create sound stories, plots that evolve from one element to the next. I have thought about this approach to composition often, but it is a very complex issue from a technology standpoint. The way I tend to see it is, moving from one sound to another is just as difficult and artistic a process as moving from one musical (pitch-based) idea to another, and therefore is an endeavor that will elude comprehensive analysis. Anyway, -jhno ------------------------------ Date: Wed, 14 Apr 1993 10:04:32 EDT From: ronin Subject: transmogrify how is 'transmogrification' distinct from stepwise interpolation, or even just simple linear mixing with shifting coefficients? as for all the myriad ways of modulation, keep in mind that most are subclasses of a few very basic types. frequency shifting, for instance, is just amplitude modulation with upper and lower sidebands separated out. fm is really a variant of phase modulation, when implemented digitally. vocoding can be considered as subaudio am of source signal spectral bands. finally, realtime resynthesis is, from what i've gathered, sort of how dcc works. maybe. i'm still reading up... anyway... doesn't the k2000 do nonlinear waveshaping? that's something that could be put under the heading of 'modulate sound 1 with sound 2' and still not strictly be fm. -----------< 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: Wed, 14 Apr 1993 09:31:34 PDT From: metlay Subject: Re: K2000 FM etc. John Eichenseer says: > I wish there was a sampler with a magical "Transmogrify sound 1 into sound > 2 over n seconds" button. This is much easier to accomplish with synthesizers, of course; Jim Binkley, on his home-based release JADE GUITAR, had a piece where a DX7's output was being modified from one patch to another via sysex over the course of a song, so it started out as one timbre and gradually became another. -- mike metlay * atomic city * box 81175 pgh pa 15217-0675 * metlay@netcom.com --------------------------------------------------------------------------- I'll survive. Indeed I will. But boy will I ever be confused. (s. kellogg) ------------------------------ Date: Wed, 14 Apr 1993 13:02:20 -0400 From: Joe McMahon Subject: Re: K2000 FM etc. >John Eichenseer says: >> I wish there was a sampler with a magical "Transmogrify sound 1 into sound >> 2 over n seconds" button. > >This is much easier to accomplish with synthesizers, of course; Jim >Binkley, on his home-based release JADE GUITAR, had a piece where >a DX7's output was being modified from one patch to another via sysex >over the course of a song, so it started out as one timbre and gradually >became another. I think what John would like is the aural equivalent of a graphic "morph". Hmmm, now I wonder. If you took a graph of a waveform (or its Fourier transform, I'm not picky) and DID a graphical morph to a completely different waveform, what would it sounds like? Probably a MicroWave. --- Joe M. "Primal scream therapy you can dance to?" (A. Schabtach) ------------------------------ Date: Thu, 15 Apr 1993 10:00:45 +0200 From: Patrice Tarabbia Subject: transmogrify :>John Eichenseer says: :>> I wish there was a sampler with a magical "Transmogrify sound 1 into sound :>> 2 over n seconds" button. :> :>This is much easier to accomplish with synthesizers, of course; Jim :>Binkley, on his home-based release JADE GUITAR, had a piece where :>a DX7's output was being modified from one patch to another via sysex :>over the course of a song, so it started out as one timbre and gradually :>became another. : :I think what John would like is the aural equivalent of a graphic "morph". :Hmmm, now I wonder. If you took a graph of a waveform (or its Fourier :transform, I'm not picky) and DID a graphical morph to a completely :different waveform, what would it sounds like? Probably a MicroWave. : --- Joe M. The common "volume crossfade" function does it okay for me. Patrice tarabbia@cenatls.cena.dgac.fr ------------------------------ Date: Thu, 15 Apr 1993 10:49:14 EDT From: JASON VANTOMME Subject: Re: EMUSIC-L Digest - 13 Apr 1993 to 14 Apr 1993 > Joe McMahon says: > > John Eichenseer says: > > I wish there was a sampler with a magical "Transmogrify sound 1 > > into sound 2 over n seconds" button. This is much easier to > > accomplish with synthesizers, of course; Jim Binkley, > > on his home-based release JADE GUITAR, had a piece where > > a DX7's output was being modified from one patch to > > another via sysex over the course of a song, so it started > > out as one timbre and gradually became another. > > > I think what John would like is the aural equivalent of a > graphic "morph". Hmmm, now I wonder. If you took a graph of > a waveform (or its Fourier transform, I'm not picky) and > DID a graphical morph to a completely different > waveform, what would it sounds like? Probably a > MicroWave. --- Joe M. > This idea has been implemented by a former grad student in Dartmouth's Electroacoustic program named Martin McKinney. If I remember correctly, it was implemented on the NeXT. I don't know if he is still there, but a "finger@dartmouth.edu" does show an account registered to him. If no response from him, you may want to contact Larry Polansky (also at Dartmouth), who is the authority on this problem of musical morphology. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Jason D. Vantomme Faculty of Music, McGill University Montreal, PQ CANADA email: vantomme@music.mcgill.ca =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ------------------------------ Date: Thu, 15 Apr 1993 17:56:36 GMT From: "Luis E. Scheker" Subject: Re: transmogrify In article <9304150800.AA04483@geant.cenatls.cena.dgac.fr> Patrice Tarabbia writes: > :I think what John would like is the aural equivalent of a graphic "morph". > :Hmmm, now I wonder. If you took a graph of a waveform (or its Fourier > :transform, I'm not picky) and DID a graphical morph to a completely > :different waveform, what would it sounds like? Probably a MicroWave. > : --- Joe M. > > The common "volume crossfade" function does it okay for me. A sonic morph would sound much different than a volume crossfade in that somewhere in between the transition from sound 1 to sound 2 there would be a fusion (not simply a layering, as with crossfading) of the two. Now suppose you take this fused sound (e.g. a piano-choir) and morph it into something else (say, a trumpet). You would then have a hybrid piano-choir -- trumpet that would sound unlike anything possible with crossfading. If sonic morphing were ever realized, it would represent the aural equivalent of selective breeding. This is a truly astounding prospect. ------------------------------ Date: Thu, 15 Apr 1993 16:49:03 -0500 From: Brian Adamson Subject: Re: transmogrify % In article <9304150800.AA04483@geant.cenatls.cena.dgac.fr> % Patrice Tarabbia writes: % % > :I think what John would like is the aural equivalent of a graphic "morph". % > :Hmmm, now I wonder. If you took a graph of a waveform (or its Fourier % > :transform, I'm not picky) and DID a graphical morph to a completely % > :different waveform, what would it sounds like? Probably a MicroWave. % > : --- Joe M. % > % > The common "volume crossfade" function does it okay for me. % % A sonic morph would sound much different than a volume crossfade in % that somewhere in between the transition from sound 1 to sound 2 there % would be a fusion (not simply a layering, as with crossfading) of the % two. Now suppose you take this fused sound (e.g. a piano-choir) and % morph it into something else (say, a trumpet). You would then have a % hybrid piano-choir -- trumpet that would sound unlike anything possible % with crossfading. If sonic morphing were ever realized, it would % represent the aural equivalent of selective breeding. This is a truly % astounding prospect. % % Yes, there could be a distinct difference between graphically morphing a sound and a volume crossfade ... it would depend somewhat on the graphical morphing algorithm. A simple example to visualize and consider: Morphing a single cycle of a 1 kHz sinusoid into a 2 kHz sinusoid would slide through the frequencies, with spectral components from 1 kHz to 2 kHz with sideband components depending on the rate of morph ... (In my mind I visualize the 1 kHz to 2 kHz morph as a spring compressing to a shorter wavelength) A volume crossfade would only contain the distinct 1 & 2 kHz spectral components. However, someone else might visualize the morph differently. So it is unclear what effect graphical morphing would have, depends on the algorithm used. Even for visual applications, morphing isn't a cut-and-dried process ... the user selects reference points at his discretion to produce the morphin effect he or she desires. It certainly would be interesting to see/hear what might happen. - Brian A. ------------------------------ Date: Fri, 16 Apr 1993 07:41:11 +0000 From: Nick Rothwell Subject: Re: K2000 FM etc. >Jim >Binkley, on his home-based release JADE GUITAR, had a piece where >a DX7's output was being modified from one patch to another via sysex >over the course of a song, so it started out as one timbre and gradually >became another. A nice idea but only works generically on some architectures. On a DX7, a change of algorithm will sound pretty abrupt unless the changeover is highly structured with respect to the invariants in the voice architecture. Try this kind of thing on a Wavestation and, well, *fzzt*, "WELCOME TO THE KORG WAVESTATION." Nick Rothwell | cassiel@cassiel.demon.co.uk CASSIEL Contemporary Music/Dance | cassiel@cix.compulink.co.uk ------------------------------ Date: Fri, 16 Apr 1993 01:20:17 PDT From: Mike Friesen Subject: Musical Morphing &%$#:> (many messages re. timbral/wavetable morphing...) I've been tossing this idea around for a few years now, but haven't cudgeled myself into writing the up the code to do it. As has been suggested already, my solution was to use sysex messages. The algorithm I was looking at would allow the user to define "start" and "end" patches; whereupon the software would calculate, over a specified time frame (OR specified number of note events) a "delta sequence" - a list of timestamped sysex messages. Given the synths for which I'm inclined to try it (Roland J-106, various FM instruments) the problem lies primarily in handling non-linear values. Algorithm and transposition are the two real nasties here: you simply can't interpolate values. The solution I devised was to give the user a 'break parameter' option whereby the discrete event could be selectively placed within a 'delta sequence'. Experience and experimentation would determine the best placement. If my math (and experience in FM programming) is up to snuff, you should be able to get some unearthly things happening when you start mucking about with the modulator/carrier frequencies. You'll get from one sound to another without being timbrally sure of just how you got there... kinda like the sonic equivalent of hyperspace. One of these days I'll have to snag MAX and see if I can get it happening. ******************************************************** Michael Friesen North Peace Secondary School Fort St. John, British Columbia, Canada I've got a piece of brain lodged in my head! - Mr. Gumby ******************************************************** ------------------------------ Date: Fri, 16 Apr 1993 07:55:53 EDT From: Doug Walker Subject: Re: Musical Morphing > >&%$#:> (many messages re. timbral/wavetable morphing...) > >I've been tossing this idea around for a few years now, but haven't >cudgeled myself into writing the up the code to do it. As has been >suggested already, my solution was to use sysex messages. The algorithm I >was looking at would allow the user to define "start" and "end" patches; >whereupon the software would calculate, over a specified time frame (OR >specified number of note events) a "delta sequence" - a list of timestamped >sysex messages. > >Given the synths for which I'm inclined to try it (Roland J-106, various FM >instruments) the problem lies primarily in handling non-linear values. >Algorithm and transposition are the two real nasties here: you simply can't >interpolate values. The solution I devised was to give the user a 'break >parameter' option whereby the discrete event could be selectively placed >within a 'delta sequence'. Experience and experimentation would determine >the best placement. > >If my math (and experience in FM programming) is up to snuff, you should be >able to get some unearthly things happening when you start mucking about >with the modulator/carrier frequencies. You'll get from one sound to >another without being timbrally sure of just how you got there... kinda >like the sonic equivalent of hyperspace. > >One of these days I'll have to snag MAX and see if I can get it happening. > > > >******************************************************** >Michael Friesen North Peace Secondary School > Fort St. John, British Columbia, Canada >I've got a piece of brain lodged in my head! - Mr. Gumby >******************************************************** > > Hi Michael: This isn't in answer to your note above. Are you a composer? Here on the National Capital FreeNet in Ottawa (the second largest in the world after 2 months online!) there is a New Music SIG which deals with contemporary music matters of a technical, philosophical, or just informative nature, We have a lot of postings of an international nature about microtonal music especially, since I've been into that for 30 years, although I'm not very known here in Canada (or anywhere else!). I'm the SIGop and it would be great having you drop by electronically to post something like the above message. I see by your school address that we have something else in common.My day job is in education, teaching grade 7 late French Immersion. Hope to hear from you! Regards Doug ------------------------------ Date: Fri, 16 Apr 1993 11:11:04 -0400 From: Joe McMahon Subject: Re: transmogrify >So it is unclear what effect graphical morphing would have, depends on the >>algorithm used. Even for visual applications, morphing isn't a cut-and-dried >>process... the user selects reference points at his discretion to produce the >>morphing effect he or she desires. Yes. Certainly one has a lot of options here: assign "close" points (i.e., choose peaks and valleys in the two waveforms that would be close to one another if the waveforms were overlaid) or "distant" points (randomly select places in the waveforms to be mapped into one another). I'm pretty sure the "close" version is what the MicroWave does to interpolate between two waveforms; sample 0 maps to sample 0, etc, with a linear interpolation between wave 1 and wave 2. I don't think anything does the "distant" version (random sample X maps to random sample Y). I suppose I should have written this up before spilling my guts on the net... :-) --- Joe M. "Primal scream therapy you can dance to?" (A. Schabtach) ------------------------------ Date: Fri, 16 Apr 1993 08:18:30 PDT From: metlay Subject: Re: K2000 FM etc. >>Jim >>Binkley, on his home-based release JADE GUITAR, had a piece where >>a DX7's output was being modified from one patch to another via sysex >>over the course of a song, so it started out as one timbre and gradually >>became another. > >A nice idea but only works generically on some architectures. On a DX7, a >change of algorithm will sound pretty abrupt unless the changeover is >highly structured with respect to the invariants in the voice architecture. >Try this kind of thing on a Wavestation and, well, *fzzt*, "WELCOME TO THE >KORG WAVESTATION." The more they overthink the plumbing, the easier it is to stop up the drain. Now THIS would be a good place for something like a computer-based sound architecture, I would think. You don't have to worry about some bozotic little CPU chip going er er er if you ask it to do something odd..... -- mike metlay * atomic city * box 81175 pgh pa 15217-0675 * metlay@netcom.com --------------------------------------------------------------------------- I'll survive. Indeed I will. But boy will I ever be confused. (s. kellogg) ------------------------------ Date: Fri, 16 Apr 1993 11:35:16 EDT From: wbf@ALUXPO.ATT.COM Subject: Re: Sound Morphing Joe M: > I'm pretty sure the "close" version is what the MicroWave does to > interpolate between two waveforms; sample 0 maps to sample 0, etc, with a > linear interpolation between wave 1 and wave 2. What if the samples are of different lengths? And what about looping? -- Bill Fox * Fox's Den Recording Studio * Nazareth, PA * wbf@alux1.att.com ------------------------------------------------------------------------ "You don't have to worry about some bozotic little CPU chip going er er er if you ask it to do something odd....." - Metlay ------------------------------ End of the EMUSIC-L Digest ******************************