EMUSIC-L Digest Volume 34, Issue 01 This issue's topics: Studies and research Additive Synthesis information (2 messages) another list archives (2 messages) Computing in Musicology Electronic Music Studio Facilities (2 messages) filter project:1 filter project:2 Spectromorphological nonsense.... (5 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: Tue, 5 Nov 1991 19:52:00 From: UEBRU@BRFAPESP.BITNET Subject: Additive Synthesis information Hi. I wish programs, for music aplications in additive synthesis. For example : - how obtain functions amplitude and frequency for this synthesis, by segment discrete of sound. Thankz Nando UEBRU@BRFAPESP.BITNET ------------------------------ Date: Fri, 8 Nov 1991 05:51:00 From: UEBRU@BRFAPESP.BITNET Subject: Information additive synthesis Hi everybody. I need programs or algoritms for extract functions Ampitude and Phase in Short time Fourier Transform. Thank you very much. Nando Marar Unesp-Bauru-Sao Paulo Brasil My addresss : UEBRU@BRFAPESP.BITNET ------------------------------ Date: Wed, 20 Nov 1991 16:51:51 From: Russ Gelinas Subject: another list I'd just like to point out that there is another list devoted to music research. Topics cover the psychology of sound perception, artificial intelligence wrt to music composition, and other heady things like that. I believe Paul Lansky (who's been mentioned on this list) occasionally posts to the music research list. Here's what I have for an address: music-research-request%prg.oxford.ac.uk@nsfnet-relay.ac.uk I believe that there is now a US distributor. *v Russ Gelinas OPAL/ESP UNH ------------------------------ Date: Fri, 22 Nov 1991 12:22:57 From: "Joseph D. McMahon" Subject: Re: archives Yes, we do need an EMUSIC-L rules-and-regs file and an info file as well. I'm reviewing the LISTSERV database stuff to see if it's useful for non- BITNETters. We have archives running from last March to the current date. This is subject to disk availability at AU. I also have offline copies of the old digests, going almost all the way back to EMUSIC-L's conception. I haven't decided the best way to handle those; maybe when/if I get my A/UX box, I'll be able to set up a better method of accessing the back issues. --- Joe M. ------------------------------ Date: Fri, 22 Nov 1991 16:53:57 From: 1ronin Subject: archives sorry, joe, for stepping on your toes here... a couple of people have asked me a similar question by private mail, and i figure this might just be generally useful, so i am replying publicly... you can send commands to the listserver one of two ways. 1) if your system supports realtime interactive messaging, you send a message like TELL LISTSERV AT AUVM (...) or MSG LISTSERV AT AUVM (...) or maybe even MSG LISTSERV@AMERICAN.AUVM.EDU (...) depending on how you gate to this group, where (...) is the body of the command you're sending the server. 2) otherwise, send a normal mail message addressed to LISTERV@AUVM with whatever other domain thingies you need to add, like .BITNET or whatever, in which the text of the message is the command you're sending to the server. archive commands of use are INDEX and GET. both require the name of the list that you're talking about. GET also requires the name of the file you want. so... TELL LISTSERV AT AUVM INDEX EMUSIC-L will obtain a listing of the currently arhived logs. TELL LISTSERV AT AUVM GET EMUSIC-L LOGXXXXX will get you the log you want, where XXXXX is some date coding. it occurs to me that, with emusic-l now hooked up to netnews, we need to post a standard info file (with some kind of 'ignore this if you already know it' header) every few weeks or so. -------------------< Cognitive Dissonance is an Art Form >--------------- Eric Harnden (Ronin) or The American University Physics Dept. Washington, D.C ------------------------------ Date: Mon, 4 Nov 1991 08:40:09 From: Eleanor Selfridge Subject: Computing in Musicology Order Form~Individual/Domestic and Overseas PREPAID ORDERS ONLY Computing in Musicology Volume 7, 1991 Edited by Walter B. Hewlett and Eleanor Selfridge-Field Current Applications: Historical Musicology, Simulation of Historical Repertories, Encoding Tablatures; Representation and Analysis Software, Cognitive Approaches, Fractals; Systems for Visually Impaired Musicians Regular Features: Annual Survey of Notation Software (set pieces by Brahms, Mahler, Ravel); Software for Automatic Recognition; Recent Publications, Symposia, Theses; Making Music (product news) 182 pages Fully indexed ISBN 0-936943-06-8 ISSN 1057-9478 $20.00 Choose items: ISBN Qty. Year Price Each Total Amt. 0-936943-06-8 1991 $20.00 0-936943-05-X 1990 $18.00 0-936943-04-1 1989 $15.00 0-936943-03-3 1988 $12.00 0-936943-02-5 1987 $10.00 Total number _____ Subtotal __________________________ Shipping: $2/1st copy; $1 each additional [required] _________________ Air supplement: Europe $7/copy; Asia $9/copy [optional] ______________ Total enclosed __________________________ Select method of prepayment: ____ Check [$US drawn on US bank] ____ VISA ____ MasterCard Card # /__/__/__/__/__/__/__/__/__/__/__/__/__/__/__/__/ Expiration Date_________________________ Cardholder Signature_________________________________ Ship to: ____________________________________________________________ ____________________________________________________________ ____________________________________________________________ ____________________________________________________________ ____________________________________________________________ Send order and prepayment to: Center for Computer Assisted Research in the Humanities 525 Middlefield Road, Suite 120 Menlo Park, California 94025-3443, USA DO NOT REMIT BY E-MAIL: WE NEED YOUR CHECK or YOUR CARD NO./SIGNATURE FAX: (415) 329-8365 Telephone: (415) 322-3307 Federal ID # 94-3038594 Institutions may submit a purchase order #:__________________ ------------------------------ Date: Thu, 7 Nov 1991 04:17:00 From: FMUS1184%Vega@SLU.BITNET Subject: Electronic Music Studio Facilities Tom Erbe-- Your inquiry about computer/electronic music programs was not addressed to me, but it is one of the things I have seen on the list that interests me particularly. We have a small synthesizer lab at SLU and are planning to join the twentieth century before it ends by implementing a digital electronic/ computer lab and expanding our curriculum in electronic music (at present only two courses and individual study projects.) I would be happy to receive ideas from other programs already in existence, or from others like me who are planning improvements. Bob Priez Music Dept, Southeastern La. University (BitNet:BOBPRIEZ@SLU) ------------------------------ Date: Fri, 8 Nov 1991 00:47:33 From: Tom Erbe Subject: Re: Electronic Music Studio Facilities Bob Priez (BOBPRIEZ@SLU.BitNet) Writes: >We have a small synthesizer lab at SLU and are planning to join the >twentieth century before it ends by implementing a digital electronic/ >computer lab and expanding our curriculum in electronic music (at present >only two courses and individual study projects.) > >I would be happy to receive ideas from other programs already in existence, >or from others like me who are planning improvements. Here are a few ideas on setting up a good electronic/computer music studio. If I were to give a shopping list, it would go something like this (in a loose order of priority): 0) Complete interconnection of all machines (MIDI, audio, ethernet, SMPTE, etc.) 1) Digidesign SoundTools with DAT machine and Mac II 1) Computer for alogrithmic composition, performance, other real-time tasks with sampler or other open synthesizers 2) Multitrack recording studio with as many Neumanns as you can afford 3) UN*X machine, either a NeXT or SGI Indigo (UN*X without sound is limiting) I have attempted to organize the studio here so it is very accessible and flexible. We have a lot of users (students, faculty, visiting composers) and it has been important to set up as many independant workstations/studios as our budget permits. If many rooms are available, this is best. Otherwise, buy a lot of headphones. Tom Erbe * Technical Director * Center for Contemporary Music * Mills College tom@mills.edu * Mills College, Oakland, CA 94613 * (510) 430-2191 ------------------------------ Date: Fri, 1 Nov 1991 03:52:53 From: ronin Subject: filter project:1 On Wed, 30 Oct 1991 13:29:02 -0800 Tom Erbe said: >A sine wave is a real bad choice to test a filter with in that all of its >energy is at one frequency, and a filter can only change the amplitude of >that frequency. Something with a full spectrum like white noise or a >pulse wave will reveal much more about the filter. Using an external >signal is a good idea. > a sine wave is a bad choice if you only take one measurement. a common measurement technique is take an effectively infinite series of data points by synchronising a chart recorder to a frequency sweep passed through the filter. we've lent our chain-driven chart recorder to NRL, so we are taking sets of 1/4-octave measurements, both with fixed-filter/variable tone, and variable filter/fixed tone. at the moment, we're doing one round of each and will interpolate the intermediate permutations. we will then take enough data points to prove or disprove the validity of the interpolation. one of the things i would like to prove to myself is that a given oscillator frequency/filter cutoff ratio does (or does not) provide the same amplitude and phase results regardless of the arbitrary frequencies. we will also do a noise burst test at specific cutoff intervals. >BTW, does the American University have an electronic and/or computer music >program? If so, what kinds of things are you doing? > we teach Sound Synthesis as part of the Audio Technology Program, which is administered by the Physics Dept. we currently have three projects under way, all of which are pursued as extracurricular or independent study by our undergraduates. these include this filter analysis project, a computer-music workstation exploratory project, and an intermediate MIDI programming project. we also offer the Music and Technology Program, an interdisciplinary program including studies in Audio Technology, Performing Arts, and Computer Science. this is geared more toward composers and performers who want to learn how to do real work with contemporary musical tools. -------------< Extremism in the Pursuit of Good Noise is no Vice >------- Eric Harnden (Ronin) or The American University Physics Dept. Washington, D.C ------------------------------ Date: Tue, 12 Nov 1991 10:48:22 From: ronin Subject: filter project:2 our first data set on the moog was submitted to the Journal of Irreproducible Results... then we reviewed our procedures and did it again. we obtained a very nice data set for a 250 Hz sine wave, with the filter set to 1/4 octave positions under external dc control. when converted, we obtain what seems, in the earliest analysis, to be a clean 18 dB/octave rolloff characteristic, suggesting a three-pole filter design. quote none of this, now. these are just the preliminaries. we actually have the filter diagrams around here, and will be doing some cross-checking. but so far, we like it. btw... we also took rough phase data. this is interesting, but slightly more difficult to characterise. we took the data by feeding the original signal and the filtered signal to two channels of a scope, and triggering on the input signal. we could then measure the apparent delay time between a given peak on input and at output. the delay time as a percentage of the period of the signal, multiplied by 360, is the phase change in degrees. when the signal frequency is well below the cutoff point of the filter, the signal is inverted, and swings slowly towards 0 degrees as it nears the cutoff. when the signal frequency goes above the cutoff point, its phase goes rapidly negative. we will probably find that there is in fact a direct correlation between absolute phase difference and attenuation. we'll see. our next data set on the moog will be to fix the filter cutoff point and sweep the input frequency. if predictions (according to our faith in the reasonableness of design characteristics) hold up, then we will be able to show that only one of these types of data sets is necessary to completely characterize any filter. for jollies, we will also do the sawtooth and noise burst tests next week. we've called up roland and oberheim, and are starting to try to figure out where to inject and tap signals to and from the lopass filter on a matrix 6 and a juno 106. if anyone out there has the techspecs on these thingies and can provide this information more immediately, i would be much obliged. the next installment will report on the second moog data set, and on whether a limited data set and interpolation can provide usable information. stay tuned. -------------------< Cognitive Dissonance is an Art Form >--------------- Eric Harnden (Ronin) or The American University Physics Dept. Washington, D.C ------------------------------ Date: Wed, 20 Nov 1991 18:31:14 From: Bryan Basham Subject: Computer Music > [stuff you cannot do in MIDIland at all - > frequency-domain manipulation, spectral morphologies, What is spectral morphology? > flying things > through fictitious rooms at 180 miles/hour, etc.] For a while I was working on a compostional idea I called Sonic Motion. This "flying things" about idea sounds similar. Do you have actual code? Reply to me directly (Basham@AIO.JSC.NASA.GOV) if you think we should take this level of detail off-line. -Bryan ------------------------------ Date: Thu, 21 Nov 1991 11:05:08 CST From: Gregory Taylor Subject: Spectromorphological nonsense.... >>[stuff you cannot do in MIDIland at all - >>frequency-domain manipulation, spectral morphologies, > What is spectral morphology? I'm sure that someone may be better at doing this simply and clearly than I, but I'll try: It involves an approach to structure in computer which takes the distribution of spectral energy of a piece of digitized sound and then either uses that "architecture" of the sound to construct the piece or actively alters the spectral distribution in some way to acheive its ends. The pretty much normal way to analyze the spectral components of said sound is to perform a Fourier Transform on the signal. What you get is similar to the old "Sonogram" that you always see in the speech analysis books - the signal appears as a series of narrow "bins", containing a representation of the original sound rendered as a series of sine wave components whose sum is an approximation of the original signal. There are some kinds of spectral [frequency domain] manipulations that allow one to "magnify" things like attack transients by stretching them in time without altering [lowering] their frequency components. You might want to hunt up a copy of Roger Reynolds' "Transfigured Wind" for one of the beeg ultrafamous examples of this. Trevor Wishart tends to do a lot of this kind of work. There's a CD on the Virgin classics label called VOX, which collects his work of that sort. VOX 5 is a good example of the technique. It seems that spectromorphological stuff seems very big in the UK computer music world these days [the Birmingham outfit BEAST, et. al.]. Some composers use the information derived from spectral information as the raw material for compositional systems as well. Kaaija Saariaho has written a number of electro-acoustic works in which she takes a sound or signal, and the uses the analytical data mapped onto a set of parameters to produce the orchestral score. There's a quite nice CD of her work on Finlandia records which includes at least one piece of this variety [I think it's called "Lichtbogen"]. The final bit of ridiculous teaser for this stuff is anecdotal, and involves a piece of equipment that I cannot possibly afford at this point [although it *is* cheap for that it does and is]. The folks at IRCAM in Paree have developed their basic workstation around a NeXT with an add-in board with two i860s in it. The sucker runs MAX doing *real-time* DSP stuff [no kidding. It is a *monster*.]. However, they also have this lovely little graphic spectrgraphic editor that takes a sound file and gives you a spectral representation of it. You can then treat it like a file one would have in, say, McDraw, save that the outcome is playable as audio.... Draw a squiggly little rope around one of the partials of a sound and move it horizontally or vertically and then paste a squished version of same over the top of it....use a box or a shape as a filter [everything only in the box is played] or do the reverse....punch out a partial or a block of frequencies on the image of the sound. It was one of the most wonderful several minutes I've ever spent with a computer. But I'm rambling. Sorry. I hope this helps. > > flying things > > through fictitious rooms at 180 miles/hour, etc.] > For a while I was working on a compostional idea I called Sonic Motion. > This "flying things" about idea sounds similar. Do you have actual code? > Reply to me directly (Basham@AIO.JSC.NASA.GOV) if you think we should > take this level of detail off-line. > > -Bryan > > PS - Hey, server dude, I have sent out a couple of messages in the past > three days, but have not seen them come back through the EMUSIC-L > list. I know they get through, because people have responded to > my notes. What gives? > ------------------------------ Date: Thu, 21 Nov 1991 15:44:18 CST From: Gregory Taylor Subject: Re: Spectromorphological nonsense.... Pardon me, but it was me, not the estimable Mr. Basham. Secondly, there was not any MIDI involved in what I was doing. The conversions were independent, in that there weren't any MAX MIDI objects in sight. And finally, you *are* correct about running some of the things I described on the Mac. You can, if you're so inclined, run csound on the Mac [but one seems to need to wait a trifle longer]. My final choice to acquire the NeXT came down to having a cube available at less money than I'd have paid for a Mac II [the famous Businessland fire sale], and the fact that the DSP board wasn't necessary. What I *was* trying, however foolishly, to suggest was that the NeXT potentially fills a gap in terms of being a UNIX system which runs most of the currently available stuff for traditional "computer music" and whose MusicKit and MIDI toolbox stuff make it a potentially useful box to the first luck soul or company who brings out a MIDI sequencer for the MIDI whiners [I just go back to the Mac for that, since the religious arguments seem a senseless waste of time]. Sorry for the confusion. I'd be curious, though, to know why you'd tack the noun nonsense onto the subject matter. With regards, Gregory Taylor ------------------------------ Date: Thu, 21 Nov 1991 11:12:18 -0800 From: latta@SCAM.BERKELEY.EDU Subject: Re: Spectromorphological nonsense.... The DSP-controlled-by-MIDI stuff on the IRCAM machine, to which Bryan Basham referred, is a perfect example of MIDI being involved in what one may consider "impossible in MIDIland". It illustrates the point that MIDI is *symbolic*, not meant to be associated with the actual audio generation at all, whether it be pressing a key to make a synth squawk, or setting a complex spectral morphological process in action. Whatever you want to control, you can slap MIDI (or better protocols) onto it. By the way, the Max/DSP stuff he described works perfectly well on the any Macintosh with 5Meg of RAM and a supported DSP board, like, say, the oh-so-cheap Audiomedia. -C ------------------------------ Date: Fri, 22 Nov 1991 11:20:56 +0000 From: Nick Rothwell Subject: Re: Spectromorphological nonsense.... >However, they >also have this lovely little graphic spectrgraphic editor that takes >a sound file and gives you a spectral representation of it. Reminds me of something I saw on the Indigo. A snatch of dialogue, with a fourier graph of the thing. No big deal. Click and drag with the mouse, and the fourier picture rotates in realtime in 3D. Hmm, nice. Now, drag along the 3D image, a 2-dimensional coloured plane scrubs through the image, playing as it goes. Cute. Seriously cute. Nick. ------------------------------ Date: Tue, 5 Nov 1991 09:42:38 PST From: jonathan berger Subject: summer program in music technology The Center for Studies in Music Technology (CSMT) at Yale University will be offering a summer course in computer music. The six week program directed by Jonathan Berger, will include programming in Max for the Macintosh and NeXT computers, sound synthesis and signal processing, real time control and performance interaction, and more. In addition to classes, guest lectures and concerts, students will design and execute individual projects in composition or research using CSMT's facilities which include NeXT's with the IRCAM/Ariel workstation, Macs with Sound Tools, MIDI and recording studios. The course will be offered between June 29 and August 7. Full college credit is given to participants. Some scholarships are available. For more information and application materials please contact: Prof. Penelope Laurans Summer and Special Programs Yale University PO Box 2145 Yale Station New Haven CT 06520 for information on CSMT's facilities contact: Prof. Jonathan Berger The Center for Studies in Music Technology Yale School of Music 96 Wall Street New Haven CT 06520 jberger@alice.music.yale.edu ------------------------------ End of the EMUSIC-L Digest ******************************