From gmk@santafe.edu Wed Nov 9 22:40:01 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA04326; Wed, 9 Nov 94 22:40:00 -0600 Received: from santafe (santafe.santafe.edu) by sfi.santafe.edu (4.1/SMI-4.1) id AA12534; Wed, 9 Nov 94 21:29:19 MST Received: by santafe (4.1/SMI-4.1) id AA07112; Wed, 9 Nov 94 21:29:42 MST Date: Wed, 9 Nov 1994 21:29:37 -0700 (MST) From: Gottfried Mayer-Kress Subject: Observing Chaotic Systems To: icad@santafe.edu Cc: Gottfried Mayer-Kress Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO I would like to provide some background for my comments during the ICAD session on "Sound synthesis and composition applied to observing chaotic system" as they are related to some of my own work: 1. The procedure of creating a meaningful continuous signal from a discrete chaotic orbit is non-trivial. One of the problems is to avoid artifacts that would lead to low-frequency modulations that could be mistaken as meaningful structures. We have introduced a very general methodology based on suspensions of diffeomorphisms (G. Mayer-Kress, H.Haken, An Explicit Construction of a Class of Suspensions and Autonomous Differential Equations for Diffeomorphisms in the Plane, Commun. Math. Phys. 111, 63-74, 1987 ) and made this paper available to the presenter with many explanations and instructions in early 1993. 2. The statement that for the Chua oscillator one can obtain a smooth transitions between sounds with broad spectra and those with discrete spectra is at least misleading: Chaotic systems such as the Chua-attractor are (in general) structurally unstable, i.e. a small movement in parameter-space can lead to a qualitatively different attractor, which generically will have a qualitatively different spectrum. That is one of the pitfalls of chaotic systems but also an example of their enormous richness, which can be applied if one carefully explores the dynamical properties. We have a paper in preparation, where we demonstrate with the help of the Chua system, how one can use convex sums of vector-fields in order to carefully construct transitions between sounds of different spectra in a manner that is maximally smooth. (R. Mettin, G. Mayer-Kress, Chaotic Attractors from Convex Sums of Vector Fields, (in progress, see http://www.ccsr.uiuc.edu/People/gmk/Papers/ConV-1.ps for a draft version, comments are welcome) Again, that procedure requires many careful considerations and on should not be misled, that this smooth transition is inherent in the system. 3. Graphical control-interfaces are useless in a scientific context, unless they have a meaningful mapping between parameters and properties of the attractor and thereby also properties of the observed sound. We have presented examples of those interfaces at NOLTA 93 (G. Mayer-Kress, I. Choi, R. Bargar, Sound Synthesis and Music Composition using Chua's Oscillator, Proc, NOLTA93, Hawaii, Dec.5-10, 1993, see also http://www.ccsr.uiuc.edu/People/gmk/Papers/NOLTA93/nolta111/nolta111.html and http://www.ccsr.uiuc.edu/People/gmk/Projects/ChuaSoundMusic/ for general background of the Chua System in relation to sound- synthesis and music (including sound examples). A careful mapping of such a manifold by hand is very tedious and CPU intensive but it provides direct insights into audio properties of the system which can be used to design paths for the construction of sounds of desired qualities, such as properties of natural sounds. We are exploring possibilities of using neural nets and genetic algorithms to facilitate those tasks and are looking for serious collaborators in that area. Gottfried Mayer-Kress Center for Complex Systems Research, Department of Physics 3025 Beckman Institute, 405 N Mathews, Urbana, Il 61801 gmk@pegasos.ccsr.uiuc.edu (NeXT-Mail),gmk@santafe.edu URL: http://www.ccsr.uiuc.edu/~gmk/, http://www.santafe.edu/~gmk/ (217)-244-5877 (voice/fax modem),x8371(fax), x1994 (msg) From mps@acs.uswest.com Wed Nov 16 14:34:59 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA03329; Wed, 16 Nov 94 14:34:58 -0600 Received: from uswat.advtech.uswest.com by sfi.santafe.edu (4.1/SMI-4.1) id AA20512; Wed, 16 Nov 94 13:12:12 MST Received: from acs.uswest.com ([137.108.22.26]) by uswat.advtech.uswest.com (8.6.9/8.6.9) with SMTP id NAA26524 for ; Wed, 16 Nov 1994 13:13:26 -0700 Received: from reliant by acs.uswest.com (4.1/SMI-4.1) id AA04309; Wed, 16 Nov 94 14:13:04 CST Received: from avenger by reliant (NX5.67e/NX3.0M) id AA02222; Wed, 16 Nov 94 14:13:31 -0600 Message-Id: <9411162013.AA02222@reliant> Received: by avenger (NX5.67d/NX3.0X) id AA01396; Wed, 16 Nov 94 14:13:39 -0600 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3 v116.1) Received: by NeXT.Mailer (1.116.1) From: Michael Pelz-Sherman Date: Wed, 16 Nov 94 14:13:37 -0600 To: icad@santafe.edu Subject: "Arty" - facts Status: O The most interesting part of Dr. Meyer-Kress' fascinating diatribe, to me, was his claim that (I'm paraphrasing here) the pretty pictures and beautiful(?) sounds generated by rendering the output of chaotic systems visible and audible are "mere artifacts" of the rendering process, and not to be trusted as scientifically valid representations of Reality(tm). If we extend the implications of this statement, we cannot trust _any_ transduction of signals from discrete form into analog; thus, the sounds coming out of my speaker when I play a CD are "mere artifacts" of the digital-to-audio conversion process, etc... Perhaps I'm missing the boat here, but it seems to me that such artifacts are unavoidable with any mapping from discrete to continuous signals. Clearly, one very important activity of those working in the fields of visualization and sonification must necessarily be the search for representation techniques which minimize artifacts and maximize the human perception system's ability to perceive the important information in the signal. Artists are frequently very tuned in on an intuitive level to this sort of thing, provided they understand what is important about the signal in the first place. The job of the scientist when communicating with an artist or composer, then, is to help us help them by explaining what they feel is _important_ about their data. Without such information, it's impossible to separate the artifacts from the real thing. --- Michael Pelz-Sherman | "The man that hath no music in himself, mpelzshe@ucsd.edu | Nor is not moved with concord of sweet sounds, UCSD Music Dept. | Is fit for treasons, stratagems, and spoils. PhD Candidate | The motions of his spirit are dull as night, Software Engineer, | And his affections dark as Erebus. Integrity Solutions | Let no such man be trusted. Mark the music." St. Paul, Minnesota | Phone: 612-683-0228 | William Shakespeare, NeXTMail OK! | "The Merchant of Venice" From mspboss!zappa!mps@uunet.uu.net Thu Nov 17 23:58:27 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA01193; Thu, 17 Nov 94 23:58:26 -0600 Received: from relay3.UU.NET by sfi.santafe.edu (4.1/SMI-4.1) id AA09020; Thu, 17 Nov 94 22:50:49 MST Received: from uucp3.UU.NET by relay3.UU.NET with SMTP id QQxqmx17288; Fri, 18 Nov 1994 00:52:06 -0500 Received: from mspboss.UUCP by uucp3.UU.NET with UUCP/RMAIL ; Fri, 18 Nov 1994 00:52:17 -0500 Received: by mspboss (NX5.67d/NX3.0M) id AA26705; Thu, 17 Nov 94 23:34:59 -0600 Received: by zappa (NX5.67d/NX3.0M) id AA27317; Thu, 17 Nov 94 23:36:55 -0600 Date: Thu, 17 Nov 94 23:36:55 -0600 From: Michael Pelz-Sherman Message-Id: <9411180536.AA27317@zappa> Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: mspboss!uunet!santafe.edu!icad@uunet.uu.net Subject: Sound is Qualitative, Vision is Quantitative (?) Status: O It occurs to me that the readily apparent gap between Dr. Meyer-Kress and his artistic collaborators may be symptomatic of a larger problem with the entire field of sonification: The problem is, essentially, the subject of this message: Sound is Qualitative, Vision is Quantitative. More explicitly, humans seem to use their sense of hearing to derive "qualities" of things around us: the reverberent characteristics of a room tell us about not only its size, but the surface of its walls and perhaps roughly about its shape. We recognize in the sound of someone's voice not only their identity, but also their state of mind, and perhaps (as ICAD '94 attendee Norma Crivici will tell you) about their physical and emotional health. We may not be able to make accurate measurements this way, although as I'm sure our blind colleagues can tell us, the potential for humans to develop our sense of hearing is probably vastly underestimated. (Anyone who saw said colleagues get around on those Anasazi cliffs would have to admit that!) The "quality-oriented" nature of sound is not a liability, is it an incredibly valuable feature. Those who fail to use sound for what it's good for, and insist on forcing the round peg into a square hole, I suspect will have a very hard time ever producing anything in the way of successful experiments with sonification. Michael Pelz-Sherman UCSD Music Dept. mpelzshe@ucsd.edu From gmk Thu Nov 17 01:07:40 1994 Date: Thu, 17 Nov 94 01:07:40 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Being explicit about intentions / "Arty" - facts Status: O Michael, I think your last message is very thoughtful and of general interest for the icad audience. Thus I take the freedom to forward it to this list together with my short reply: I find it interesting that you mention Feyerabend: I like his view a lot and it is actually the main reason why I think that in order to work in an interdisciplinary collaboration one has to keep the categories and concepts separated. I agree with you completely that with the "scientific method" alone one cannot discover anything new. Coming from complex systems and chaos theory I also would never suggest to look at data with linear displays. The key is "heuristics": By looking at the data in a new way you can hope to discover a new aspect of the system that was hidden to the eyes of the experts. I have been so many times in situations where the "experts" told me that my ideas were crazy and why. In most cases they were correct and I had to admit it. In a few cases I could tell the experts something new. Today I am one of the "experts" and I do know a lot about non-linear dynamical systems and chaos and their properties. And you will have noticed that I don't hesitate to share my knowledge about why "the Jimi-Hendrix-sounds that come out of the speakers are 'wrong' in the context of the old Spanish guitar". I know that I can convince (almost) everyone why it is not correct to render a periodic system with an aperiodic sound. That does not mean that errors in the history of science have never led to new discoveries. To do that usually requires a genius; but unfortunately geniuses are rare and fools are many. --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ Begin forwarded message: Date: Thu, 17 Nov 94 00:02:50 -0600 From: Michael Pelz-Sherman To: mspboss!uunet!pegasos.ccsr.uiuc.edu!gmk@uunet.uu.net (Gottfried Mayer-Kress), mspboss!uunet!listen.com!kramer@uunet.uu.net (Gregory Kramer) Subject: Re: Being explicit about intentions / "Arty" - facts Greg, Gottfried, and Robin: Now that this email discussion on sound design & sonification is becoming calmer - more general and less about particular people's ICAD presentations - I'd like to respond to you both (privately) about our recent postings. Feel free to post this to the ICAD list if you are so inclined - I'll trust your judgement as to whether the rest of the group would be interested in reading it. If you are not interested in continuing this discussion, please let me know. I think Dr. Meyer-Kress and I are in complete agreement that we must know what it is we are trying to represent before we can say whether we have represented it "correctly". For example, you must know in advance before you hear the distorted Spanish guitar what it's _supposed_ to sound like. Had you no prior knowledge of the original sound source, your reaction might be altogether different! (I happen to really _love_ Jimi Hendrix!) Ultimately, the yardstick by which we judge the representation is in our heads, not "out there" in the Universe. What does this mean in terms of using traditional scientific methods to discover the "truth" about data? If we know prior to our investigation that we're looking for a particular behavior X, and we design our apparatus specifically to _produce_ behavior X, we're bound to end up with (guess what?) behavior X no matter what the "truth" of the data says. There is an excellent book that I have found very influential in my questioning of the supremacy of the scientific method: "Against Method", by Paul Feyerabend. It was recommended to me by my dissertation advisor, Prof. George E. Lewis at UCSD. Finally, this discussion has me wondering if there is such a thing as a perfectly "linear" auditory display - one that introduces no distortion whatsoever into the data. I'm tempted to answer that no, there is not; furthermore, I do not believe there is such a thing as a completely "impartial" visual display, although concepts of "distortion" in visual display are much better understood. Quantitative measurements such as ratios, for example, are much more effectively conveyed graphically than sonically. An rectangular block that is twice as tall as another may be easily distinguished from one that is four times as tall, yet were these ratios (2:1, 4:1) rendered in sound, they would have a great deal in common, being octave-separated instances of the same "pitch class". Sonic renderings tend to place special importance on periodicity in both the pitch domain and the rhythmic domain. The "temporal alignment" of musical sounds seems to catch our attention very strongly. Users of auditory displays must be aware of this fact, which makes the notion of a completely "linear" auditory display seem impossible. On the other hand, we may exploit our sensitivity to this parameter to great effect. I would suspect, for example, that most people can be trained to bring two sounds into quite accurate temporal alignment with one another. I can imagine doing very simple experiments like this with a computer and a MIDI synth; set up two metronomic pulses, let subjects tweak the tempos (but without quantitative displays of the actual tempo in b.p.m.) until they seem to be identical, and record their success. Has anyone conducted such experiments? -Michael From mps@acs.uswest.com Fri Nov 18 15:05:04 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA01753; Fri, 18 Nov 94 15:05:02 -0600 Received: from uswat.advtech.uswest.com by sfi.santafe.edu (4.1/SMI-4.1) id AA16256; Fri, 18 Nov 94 13:43:29 MST Received: from acs.uswest.com ([137.108.22.26]) by uswat.advtech.uswest.com (8.6.9/8.6.9) with SMTP id NAA09366; Fri, 18 Nov 1994 13:44:40 -0700 Received: from reliant by acs.uswest.com (4.1/SMI-4.1) id AA18277; Fri, 18 Nov 94 14:44:18 CST Received: from avenger by reliant (NX5.67e/NX3.0M) id AA10788; Fri, 18 Nov 94 14:44:43 -0600 Message-Id: <9411182044.AA10788@reliant> Received: by avenger (NX5.67d/NX3.0X) id AA04907; Fri, 18 Nov 94 14:46:11 -0600 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3 v116.1) Received: by NeXT.Mailer (1.116.1) From: Michael Pelz-Sherman Date: Fri, 18 Nov 94 14:46:09 -0600 To: malcolm@interval.com Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Cc: icad@santafe.edu Status: O Thanks for the info, Malcolm. My posting was intentionally stated in the extreme in order to stimulate just such discussion. I'm sure there are many exceptions to this "rule", the longer-winded version of which being, "The auditory perception mechanism is better suited to many kinds of _qualitative_ information display than is the visual perception mechanism, whereas the visual perception mechanism is better suited to many kinds of _quantitative_ information display than the auditory perception mechanism." I'm not sure how much stock to put in experiments like the one you cite. One can devise experiments to prove or disprove almost anything. Just for kicks, though, I guess I'll take a look at the Perception & Psychophysics article. BTW - is there some research where the "JND" for tempo is determined to be 3-5%? Or are they simply applying the JND's for pitch to tempo? So... does this research imply that auditory displays should be based on popular song melodies? ;^) --- Michael Pelz-Sherman | "The man that hath no music in himself, mpelzshe@ucsd.edu | Nor is not moved with concord of sweet sounds, UCSD Music Dept. | Is fit for treasons, stratagems, and spoils. PhD Candidate | The motions of his spirit are dull as night, Software Engineer, | And his affections dark as Erebus. Integrity Solutions | Let no such man be trusted. Mark the music." St. Paul, Minnesota | Phone: 612-683-0228 | William Shakespeare, NeXTMail OK! | "The Merchant of Venice" Begin forwarded message: X-Sender: malcolm@interval.interval.com Date: Fri, 18 Nov 1994 10:12:09 -0700 To: uunet!interval.com!withgott@uunet.uu.net (Meg Withgott), uunet!interval.com!audio@uunet.uu.net, uunet!interval.com!vision@uunet.uu.net, uunet!cs.cornell.edu!zabih@uunet.uu.net, uunet!santafe.edu!icad@uunet.uu.net From: mspboss!uunet!interval.com!malcolm@uunet.uu.net (Malcolm Slaney) Subject: Re: Sound is Qualitative, Vision is Quantitative (?) [I'm not on the ICAD mailing list, but I was forwarded this note and thought it was worth pointing out this bit of data on pitch....] >Date: Thu, 17 Nov 94 23:36:55 -0600 >From: Michael Pelz-Sherman >To: mspboss!uunet!santafe.edu!icad@uunet.uu.net >Subject: Sound is Qualitative, Vision is Quantitative (?) > >Sound is Qualitative, >Vision is Quantitative. More explicitly, humans seem to use their sense of >hearing to derive "qualities" of things around us: the reverberent >characteristics of a room tell us about not only its size, but the surface of >its walls and perhaps roughly about its shape. The recent literature on perfect pitch says we are all very good at remembering a pitch (or tempo), i.e. we all have perfect pitch. This seems more than just a qualitative task. When talking about perfect pitch there are three issues: 1) Can you remember the pitch value? 2) Can you produce it? 3) Can you label it (i.e. that's a little flat of G above middle C)? Dan Levitt's study showed that people are VERY good at tasks 1 and 2. Part 3, which is usually what people call perfect pitch, doesn't seem very important to me. Is this what you mean by being quantitative? I've attached an abstract for a recent talk of Dan's. I can't find a electronic abstract for the paper in the databases. -- Malcolm Memory for the Tempo of Familiar Songs (Presented at the Stanford CCRMA Hearing Seminar) Daniel Levitin (University of Oregon and CCRMA) Perry Cook (CCRMA) In a previous study ("Absolute Memory for Musical Pitch," Perception & Psychophysics, October, 1994) I presented evidence that people tend to remember the actual pitches of familiar melodies, suggesting that some form of latent absolute pitch may be resident in a significant percentage of people. Perry Cook and I are now gathering evidence that people also seem to remember - with a great degree of accuracy - the actual tempos of familiar songs. Subjects asked to sing songs they know well from memory tend to sing within a JND of the proper tempo (3-5%). Perry and I will discuss this work, the methodology, findings, and control studies we performed to control for possible artifacts. From malcolm@interval.com Fri Nov 18 18:15:32 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA01954; Fri, 18 Nov 94 18:15:31 -0600 Received: from fred.interval.com by sfi.santafe.edu (4.1/SMI-4.1) id AA18687; Fri, 18 Nov 94 17:09:32 MST Received: by fred.interval.com id AA03737; Fri, 18 Nov 94 16:10:47 PST Received: by interval.interval.com id AA12734; Fri, 18 Nov 94 16:10:44 PST X-Sender: malcolm@interval.interval.com Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 18 Nov 1994 16:10:34 -0700 To: Michael Pelz-Sherman From: malcolm@interval.com (Malcolm Slaney) Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Cc: icad@santafe.edu Status: RO At 1:46 PM 11/18/94, Michael Pelz-Sherman wrote: >"The >auditory perception mechanism is better suited to many kinds of _qualitative_ >information display than is the visual perception mechanism, whereas the >visual >perception mechanism is better suited to many kinds of _quantitative_ >information display than the auditory perception mechanism." I'm not sure I'd buy this. From my dictionary Quantitative: Expressed as the measurement of quantity Qualitative: Concerning quality Quality: A characteristic or attribute; property Ignoring the popular definition, properties or attributes can be numerical. Perhaps you mean that audition is better at categorical decisions (as was pointed out by a colleague)? It seems to me that neither vision or audition are good at quantitative assements. I certainly can't tell how big the moon is by looking at it. Color perception is inherently qualitative. We do seem to be good at making comparisons: this tone is higher pitched then this one, this car is longer then the ruler. (Or the visual illusion which has two lines of the same length, but the arrow heads at the end either point in or out... and the resulting lines look very different in size.) >BTW - is there some research where the "JND" for tempo is determined to be >3-5%? Or are they simply applying the JND's for pitch to tempo? Depending on the task, the JNDs for pitch are less than 1%. You'd have to talk to Dan Levitin (levitin@ccrma.stanford.edu) to get the tempo references. The pitch work is solid and good. I recommend it. The temp work is in progress and hasn't been written up yet. I think he's still working on getting the story right. >So... does this research imply that auditory displays should be based on >popular song melodies? ;^) Gee, I bet nobody at ICAD as thought of using songs as a dimension. You could use the age of the composer, religeousness (from traditional church hymns to Madonna), or even ..... :-). -- Malcolm From malcolm@interval.com Fri Nov 18 12:24:37 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA01626; Fri, 18 Nov 94 12:24:14 -0600 Received: from fred.interval.com by sfi.santafe.edu (4.1/SMI-4.1) id AA14024; Fri, 18 Nov 94 11:11:28 MST Received: by fred.interval.com id AA22045; Fri, 18 Nov 94 10:12:38 PST Received: by interval.interval.com id AA27555; Fri, 18 Nov 94 10:12:19 PST X-Sender: malcolm@interval.interval.com Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 18 Nov 1994 10:12:09 -0700 To: withgott@interval.com (Meg Withgott), audio@interval.com, vision@interval.com, zabih@cs.cornell.edu, icad@santafe.edu From: malcolm@interval.com (Malcolm Slaney) Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Status: O [I'm not on the ICAD mailing list, but I was forwarded this note and thought it was worth pointing out this bit of data on pitch....] >Date: Thu, 17 Nov 94 23:36:55 -0600 >From: Michael Pelz-Sherman >To: mspboss!uunet!santafe.edu!icad@uunet.uu.net >Subject: Sound is Qualitative, Vision is Quantitative (?) > >Sound is Qualitative, >Vision is Quantitative. More explicitly, humans seem to use their sense of >hearing to derive "qualities" of things around us: the reverberent >characteristics of a room tell us about not only its size, but the surface of >its walls and perhaps roughly about its shape. The recent literature on perfect pitch says we are all very good at remembering a pitch (or tempo), i.e. we all have perfect pitch. This seems more than just a qualitative task. When talking about perfect pitch there are three issues: 1) Can you remember the pitch value? 2) Can you produce it? 3) Can you label it (i.e. that's a little flat of G above middle C)? Dan Levitt's study showed that people are VERY good at tasks 1 and 2. Part 3, which is usually what people call perfect pitch, doesn't seem very important to me. Is this what you mean by being quantitative? I've attached an abstract for a recent talk of Dan's. I can't find a electronic abstract for the paper in the databases. -- Malcolm Memory for the Tempo of Familiar Songs (Presented at the Stanford CCRMA Hearing Seminar) Daniel Levitin (University of Oregon and CCRMA) Perry Cook (CCRMA) In a previous study ("Absolute Memory for Musical Pitch," Perception & Psychophysics, October, 1994) I presented evidence that people tend to remember the actual pitches of familiar melodies, suggesting that some form of latent absolute pitch may be resident in a significant percentage of people. Perry Cook and I are now gathering evidence that people also seem to remember - with a great degree of accuracy - the actual tempos of familiar songs. Subjects asked to sing songs they know well from memory tend to sing within a JND of the proper tempo (3-5%). Perry and I will discuss this work, the methodology, findings, and control studies we performed to control for possible artifacts. From gmk Mon Nov 21 00:01:02 1994 Date: Mon, 21 Nov 94 00:01:02 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Cc: Malcolm Slaney< malcolm@interval.com> Status: O Malcolm Slaney< malcolm@interval.com> wrote: > It seems to me that neither vision or audition are good at quantitative > assements. I certainly can't tell how big the moon is by looking at it. > That is certainly correct, but you probably would have no problem if I showed you a picture with 13 oranges to give me the correct number of oranges (alright, make that 13 orange dots...) in the picture. If you played me a sequence of five notes at different pitches I probably would have a hard time telling you if they are 5 or 6 or 4 different pitches. But if you play the same 5 notes 100 times and at the 101'st time you would drop one note or change one of them I am pretty confident that I would notice that because it sounds "different", even if you play the sequence very fast. On the other hand if I would show you a sequence of 100 slides, each with 13 (or maybe 5) oranges and at the 101'st slide I would leave one out or add one, I would expect that you would have to have a very good concentration to catch that. I think this sort of monitoring/alerting function of the auditory system is most important for scientific applications. For the 92 ICAD paper we have produced examples where we have periodic (up to 8) sequences on one stereo channel and intermittent sequences (periodic for a long time and every so often a chaotic burst) on the other channel. You can listen to some of them on the ICAD-92 CD or at: http://www.ccsr.uiuc.edu/People/gmk/Papers/ICAD92/AnnotEx.html --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From malcolm@interval.com Mon Nov 21 16:56:24 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA04544; Mon, 21 Nov 94 16:55:55 -0600 Received: from fred.interval.com by sfi.santafe.edu (4.1/SMI-4.1) id AA09989; Mon, 21 Nov 94 15:41:48 MST Received: by fred.interval.com id AA06429; Mon, 21 Nov 94 14:42:55 PST Received: by interval.interval.com id AA17672; Mon, 21 Nov 94 14:42:53 PST X-Sender: malcolm@interval.interval.com Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 21 Nov 1994 14:42:40 -0700 To: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress), icad@santafe.edu From: malcolm@interval.com (Malcolm Slaney) Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Status: O At 11:01 PM 11/20/94, Gottfried Mayer-Kress wrote: >That is certainly correct, but you probably would have no problem if I >showed you a picture with 13 oranges to give me the correct number of oranges >(alright, make that 13 orange dots...) in the picture. I agree that... but I can't look at the picture and count all the oranges in one instance. Instead I have to look at each one and enumerate them. Likewise, if you played 13 tones to me, one at a time, I could certainly count each of them. Perhaps the real difference is in information flow and/or processing style. I recently led a discussion at Stanford on the following book chapter "A Critique of Pure Vision", by P. Churchland, V.S. Ramachandran and T. Sejnowski. In: Large-Scale Neuronal Theories of the Brain, by C. Koch and J. Davis, MIT Press, 1994. This chapter argues that vision is inherently an active process. We don't see all thirteen oranges and then count them. Instead we see then one at a time, search out the next one, and count it. I tried to draw parallels between this work and audition. I thought I was successful, but this discussion argues that maybe there really are differences. I don't think it's a quantitative/qualititative issue, perhaps it's another dimension. Michael Pelz-Sherman, in a separate note, argues that with vision we can easily interpret graphs, and there is no auditory equivalent. That's true... but graphs require a lot of interaction. Something that is hard to do in audition (at least for non-musicians like me.) -- Malcolm From mspboss!zappa!mps@uunet.uu.net Mon Nov 21 20:00:34 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA04716; Mon, 21 Nov 94 20:00:33 -0600 Received: from relay3.UU.NET by sfi.santafe.edu (4.1/SMI-4.1) id AA12935; Mon, 21 Nov 94 18:49:59 MST Received: from uucp3.UU.NET by relay3.UU.NET with SMTP id QQxrbb10857; Mon, 21 Nov 1994 20:51:18 -0500 Received: from mspboss.UUCP by uucp3.UU.NET with UUCP/RMAIL ; Mon, 21 Nov 1994 20:51:27 -0500 Received: by mspboss (NX5.67d/NX3.0M) id AA01487; Mon, 21 Nov 94 19:42:51 -0600 Received: by zappa (NX5.67d/NX3.0M) id AA14822; Sun, 20 Nov 94 11:51:41 -0600 Date: Sun, 20 Nov 94 11:51:41 -0600 From: Michael Pelz-Sherman Message-Id: <9411201751.AA14822@zappa> Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: mspboss!uunet!santafe.edu!icad@uunet.uu.net Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Status: O Malcolm Slaney wrote: > I'm not sure I'd buy this. I don't really buy it either, at least not entirely. Again, my statement that vision is better suited to numerical judgements while sound is better suited to qualitative assessment is more like a "rule of thumb" than an axiomatic "truth". Your counter-example of color perception is a case in point. > It seems to me that neither vision or audition are good at quantitative > assements. I certainly can't tell how big the moon is by looking at it. Certainly without any point of reference, everything is relative. (Of course, if you could measure the intensity of light being reflected by the moon, you might be able to roughly estimate its distance, right? Then you might be able to get a fix on how big it is...) At least with a visual representation, such as a graph, we can derive a fairly accurate numeric value for a given data point by referring to the numbers along the graph's vertices. With an auditory display, numeric quantities are extremely difficult, if not impossible, to represent. My main point is that the world of sound is extremely rich in potential for presenting qualitative information. Rather than trying to force sound to into a role it does not seem to be cut out for, as many scientists in the field of AD seem to be trying to do, I suggest we should use sound to represent the high-level qualities about our data that matter to us. Michael Pelz-Sherman UCSD Music Dept. mpelzshe@ucsd.edu From mspboss!zappa!mps@uunet.uu.net Tue Nov 22 01:00:31 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA05034; Tue, 22 Nov 94 01:00:30 -0600 Received: from relay3.UU.NET by sfi.santafe.edu (4.1/SMI-4.1) id AA14449; Mon, 21 Nov 94 23:49:52 MST Received: from uucp4.UU.NET by relay3.UU.NET with SMTP id QQxrbv07748; Tue, 22 Nov 1994 01:51:02 -0500 Received: from mspboss.UUCP by uucp4.UU.NET with UUCP/RMAIL ; Tue, 22 Nov 1994 01:50:43 -0500 Received: by mspboss (NX5.67d/NX3.0M) id AA06188; Tue, 22 Nov 94 00:37:42 -0600 Received: by zappa (NX5.67d/NX3.0M) id AA15703; Tue, 22 Nov 94 00:25:26 -0600 Date: Tue, 22 Nov 94 00:25:26 -0600 From: Michael Pelz-Sherman Message-Id: <9411220625.AA15703@zappa> Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: mspboss!uunet!interval.com!malcolm@uunet.uu.net (Malcolm Slaney) Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Cc: mspboss!uunet!santafe.edu!icad@uunet.uu.net Status: O > Michael Pelz-Sherman, in a separate note, argues that with vision we can > easily interpret graphs, and there is no auditory equivalent. That's > true... but graphs require a lot of interaction. Something that is hard to > do in audition (at least for non-musicians like me.) Well now, Malcom's comment gives me an idea: what if you added to a "classical" auditory display (a la Gregory Kramer's "Some Organizing Principles for Representing Data with Sound", in Auditory Display, Ed. Gregory Kramer, Addison Wesley 1994) an "interactive" control device - say a mouse click - which would cause the system to "display" the current numeric value of the display using a speech synthesizer? One could also imagine sound being useful in designing a "talking calculator", which would provide auditory feedback while one entered numbers. 'Might be interesting to see what effect this would have on both accuracy and speed? Perhaps another key issue here, following up on Malcom's point that vision may be a more "active" and sound a more "passive" perceptual process (I'm putting words in Malcom's mouth here - note: I suspect this is highly culturally determined and is NOT predetermined by biology), is the fact that the concept of number and measurement itself is a highly abstract, _constructed_ notion. It takes years of training to recognize numbers and even more to recognize the importance of number as applied to various concepts: age, weight, temperature, etc. Contrast this with more intuitively-understood "qualitative" judgements - the tone of someone's voice may be instantly recognized as threatening or friendly, even to a newborn infant. Many animals clearly respond to "tone of voice" in certain predictable ways. It is the ability of sound to convey these qualities that I find interesting, and yet there seems to be relatively little research in the AD literature I've looked at so far on the subject. One example that moves in this direction is Gaver's work on Auditory Icons which makes use of physical modelling to convey the inherent "qualities" of a sound. I wonder: what is Xerox doing with Mr. Gaver and his research results? Michael Pelz-Sherman UCSD Music Dept. mpelzshe@ucsd.edu From gmk Tue Nov 22 15:01:03 1994 Date: Tue, 22 Nov 94 15:01:03 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Re: Sound is Qualitative, Vision is Quantitative (?) Status: O At Mon, 21 Nov 1994 14:42:40 -0700 Malcolm Slaney wrote: > >That is certainly correct, but you probably would have no problem if I > >showed you a picture with 13 oranges to give me the correct number of oranges > >(alright, make that 13 orange dots...) in the picture. > > I agree that... but I can't look at the picture and count all the oranges > in one instance. Instead I have to look at each one and enumerate them. > Malcolm, thanks for that clarification: my point is that this "mental counting" is typically done more accurately with visual information. In my example a more accurate sonic example would be to play 5 sounds simultaneously which certainly would make the task even more complicated. Thus we can use the trick of "separation of time-scales" that was also talked about at the ICAD meeting: we can spread the simultaneous appearance of the elements of our object out in a "short" time-interval a fraction of a second, say. It could be even extended up to about 4 seconds and still be perceived as one individual pattern, since it is stored in working memory. Still I would claim that the quality of that sonic pattern would be much more dominant than its quantitative properties: It should be easier to re-member if a certain pattern has been heard before than to mentally count the number of different tones in the pattern. I think that would also agree with the difference in the neuronal organization of pattern processing mechanisms in the visual vs. auditory system. For example I would expect that it is easier to count the 13 dots when they are presented over one second simultaneously compared to showing them in the same time sequentially. I would be interested to see if the synchronization events that I played during ICAD would be better audible when they are not played precisely in synchrony but with a short variation in onset-time, on top of a fixed longer time-scale at which the notes are played. (This is related to a discussion I had with Michael Pelz-Sherman at ICAD). At Sun, 20 Nov 94 11:51:41 -0600 Michael Pelz-Sherman writes > My main point is that the world of sound is extremely rich in > potential for presenting qualitative information. Rather > than trying to force sound to into a role it does not seem to be cut > out for, as many scientists in the field of AD seem to be trying to > do, I suggest we should use sound to represent the high-level > qualities about our data that matter to us. > Michael, I think you pretty well summarized the main message from our ICAD'92 paper (see http://www.ccsr.uiuc.edu/People/gmk/Papers/ICAD92/ICAD92.html) and I think perception of these "high-level qualities" is related to the human capability of producing and enjoying music. > more intuitively-understood "qualitative" judgements - the > tone of someone's voice may be instantly recognized as > threatening or friendly, even to a newborn infant. Many > animals clearly respond to "tone of voice" in certain > predictable ways. > That certainly makes a lot of sense and I would be interested in references that provide physiological confirmation of that hypothesis. Has anyone looked at P-300 in response to different tones? It would be interesting to repeat the experiment that we did on sound sequences of variable complexity to individual sounds of different degree of chaos. (see N. Birbaumer, W. Lutzenberger, H. Rau, G. Mayer-Kress, I. Choi, C. Braun, Perception of Music and Dimensional Complexity of Brain Activity, Tech.Rep. CCSR-94-28, http://www.ccsr.uiuc.edu/~gmk/Papers/ChaosMuTR2.ps) --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From gmk Sat Nov 26 13:44:59 1994 Date: Sat, 26 Nov 94 13:44:59 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Qualitative Perception of "Periodic, Chaotic, Random" Sequences Cc: Malcolm Slaney< malcolm@interval.com>, V.R. Algazi, Barbara Shinn-Cunningham Status: O Greetings, some time ago we had this discussion about qualitative vs quantitative perception and I tried to make the point that it is important for any meaningful rendering process to preserve the "qualitative features" of the system to be rendered. The categories that are among the most important ones in dynamical systems research are "Periodic, Chaotic, Random". I think these are also categories that are relevant for the perception of music. On my icad poster I had a paper N. Birbaumer, W. Lutzenberger, H. Rau, G. Mayer-Kress, I. Choi, C. Braun, Perception of Music and Dimensional Complexity of Brain Activity, Tech.Rep. CCSR-94-28 that I just made available on WWW/Mosaic: http://www.ccsr.uiuc.edu/~gmk/Papers/MusicEEG It contains all the different sound files (3 * 3) that have been used for the EEG recordings. Best, --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From gmk Tue Nov 29 19:50:23 1994 Return-Path: Received: by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA14178; Tue, 29 Nov 94 19:50:22 -0600 Date: Tue, 29 Nov 94 19:50:22 -0600 From: gmk (Gottfried Mayer-Kress) Message-Id: <9411300150.AA14178@ pegasos.ccsr.uiuc.edu > To: icad.pub@santafe.edu Subject: abstract for icad 94 Cc: gmk Status: RO Greetings! Could whoever is taking care of this please acknowledge receipt of this abstract? This is the fifth submission. Thanks, Gottfried Mayer-Kress Dynamical Resonances and Synchronization of Auditory Stimuli and Evoked Responses in Multi Channel EEG Gottfried Mayer-Kress CCSR, Beckman Inst, Physics Dept, UIUC and SFI 405 N-Mathews Urbana, Il 61801 gmk@santafe.edu Perceptions of complex structures or other cognitive events are associated with synchronous oscillation of neural cell assemblies which take place in the 40 Hz domain and last for about 100ms. We use EEG signals from 31 scalp electrodes to identify these short term synchronization event. The challenge for audification applications is to make these relatively rare events perceptible in the background of incoherent signals from all electrodes. In a first approach we use an "Orchestra Paradigm" by asigning a different standard musical instrument (Piano, Flute, Violin, Glocken) to four of these electrodes. We can clearly discriminate transitions between uncorrelated activities of these areas and the events of short term synchronization. We expect a general resonance principle to apply also in the inverse problem: The non-linear resonance hypothesis of music perception was tested in an experiment comparing a group of musically sophisticated and a group of less sophisticated subjects. The prediction that weakly chaotic music entrains less complex brain wave (EEG) oscillations at the prefrontal cortex was confirmed by using a correlational dimension algorithm. Strongly chaotic (stochastic) and periodic music both stimulated higher brain wave complexity. More sophisticated subjects who prefer classical music showed higher EEG dimensions while less sophisticated subjects responded with a drop in brain wave complexity to rhythmical weakly chaotic music. Subjects ratings of perceived complexity of the musical pieces followed mathematical (objective) structure of the music and did not reflect the changes in brain wave complexity. The results are interpreted in the context of an associated (Hebbian) network theory of non-linear brain dynamics. References: N. Birbaumer, W. Lutzenberger, H. Rau, G. Mayer-Kress, I. Choi, C. Braun, Perception of Music and Dimensional Complexity of Brain Activity,CCSR Tech-Rep-9 4-28 http://www.ccsr.uiuc.edu/~gmk/Papers/MusicEEG/ http://www.ccsr.uiuc.edu/People/gmk/Projects/EEGSound/ From gmk Tue Dec 6 22:28:23 1994 Date: Tue, 6 Dec 94 22:28:23 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Stephen Barrass paper Status: O I just downloaded Stephen Barrass' paper which I find quite impressive. I like the systematic outline and graphical representation of auditory display parameters. Stephen emphasizes "fidelity" of the representation but I didn't quite understand how he addressed the invariance-problem: The problem that certain _qualities_ of the displayed date need to be preserved while others can be displayed at low fidelity. I think it also would be nice to have some sonic examples on the ftp site. --- Gottfried Mayer-Kress http://www.santafe.edu/~gmk/ From kramer@listen.com Wed Dec 7 12:49:50 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA04668; Wed, 7 Dec 94 12:49:49 -0600 Received: from desiree.teleport.com by sfi.santafe.edu (4.1/SMI-4.1) id AA23452; Wed, 7 Dec 94 11:38:28 MST Received: from [204.119.60.148] (ip-pdx3-20.teleport.com [204.119.60.148]) by desiree.teleport.com (8.6.9/8.6.9) with SMTP id KAA13370 for ; Wed, 7 Dec 1994 10:39:49 -0800 Date: Wed, 7 Dec 1994 10:39:49 -0800 Message-Id: <199412071839.KAA13370@desiree.teleport.com> X-Sender: listen@mail.teleport.com Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: icad@santafe.edu From: kramer@listen.com (Gregory Kramer) Subject: Variations in fidelity Status: O Mayer-Kress mentions "The problem that certain _qualities_ of the displayed data need to be preserved while others can be displayed at low fidelity." While lacking the sytematic elegance of Barrass' proposed system, I have dealt with this issue by providing for scaling of sonification variables. In my Sonification Toolkit, data variables can be scaled after being normalized but before being routed to control auditory variables. Another way of looking at this, indeed the way that I have used the system more frequently, is to use this scaling to perceptually balance, or "equalize", the display when auditory variables of different "compellingness" are utilized. Of course, one could also rate data variables by "importance" and assign the more important data to the more compelling auditory variables. For further discussion see my paper, "Some Organizing Principles for Representing Data with Sound" in Auditory Display, the ICAD 92 Addison Wesley book. (pp. 199-202) Greg Gregory Kramer Clarity/Santa Fe Institute 1704 SW Spring Street Portland, OR 97201 503-223-9948 fax:503-223-8293 kramer@listen.com From mps@zappa.is.com Fri Dec 9 01:14:46 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA06740; Fri, 9 Dec 94 01:14:42 -0600 Received: from relay3.UU.NET by sfi.santafe.edu (4.1/SMI-4.1) id AA14430; Fri, 9 Dec 94 00:05:10 MST Received: from uucp2.UU.NET by relay3.UU.NET with SMTP id QQxtmq18252; Fri, 9 Dec 1994 02:06:40 -0500 Received: from mspboss.UUCP by uucp2.UU.NET with UUCP/RMAIL ; Fri, 9 Dec 1994 02:06:45 -0500 Received: by is.com with UUCP (NX5.67d/IS-Mailhost-1.0.1/Plexare-1.0.2) id AA05258; Fri, 9 Dec 94 00:53:02 -0600 Received: by zappa.is.com (NX5.67d/IS-Home-1.0.0/Plexare-1.0.2) id AA04351; Fri, 9 Dec 94 00:14:35 -0600 Date: Fri, 9 Dec 94 00:14:35 -0600 From: mps@zappa.is.com (Michael Pelz-Sherman) Message-Id: <9412090614.AA04351@zappa.is.com> Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: icad@santafe.edu Subject: Re: Sonic Examples / Variations in fidelity Status: O Gottfired Meyer-Kress and Stephen Barrass write: >> >> The problem that certain _qualities_ of the displayed >> >> data need to be preserved while others can be displayed at low fidelity. >> >> Do you have a simple example of what you mean here Gottfried ? > As we had discussed earlier on this list, there are certain > _qualitative_ features of the data that need to be > preserved in order to be meaningful in a scientific context. > A simple example might be the existence of a maximum: something > increases and then decreases. The maximum is a critical point, > one does have to hear that there is something that has the > quality of a maximum, otherwise the representation will be > useless. > The other example, that Insook Choi had presented at ICAD'94 is > the quality "periodic" as opposed to "chaotic". No matter how > you choose to represent the data, if something that is periodic > afterwards sounds as if it is chaotic, then something has gone > wrong. The relationship between "fidelity" and "preservation of quality" is not clear. For example, I recently heard a very early recording of Elvis Presley. You can clearly hear the unique "quality" of Elvis' voice, even though the fidelity of the recording is extremely low. Similarly, the successful recognition of digitized representations of human faces has been shown to require very few bits. (I forget exactly where I read about this, or what the number of bits was, but it was surprisingly low.) This leads me to believe that there is something more complex at work here than a simple linear relationship between "fidelity" and preservation of qualititative information. Familiarity with the subject matter plays a huge role, and certain stimuli will be more resistant to lower fidelity than others. Nonetheless, I would like to see some scientific studies done on the relationship between, say, digital audio bit depth and the ability to hear changes in amplitude over the course of a tone. This could, indeed, affect one's ability to perceive maxima if they are mapped to amplitude. Regarding the Choi examples that you refer to (and have referred to previously), I'm not sure I agree with your statements that the display we heard and saw at ICAD '94 did not preserve the qualities of periodicity and chaos. It seemed to me that when the cursor was placed at certain points along the mainfold, the sound would become noticably "pitched", whereas most of the time it sounded like various flavors of noise. Are you saying that there is no relationship between the actual behavior of the circuit and what we are seeing/hearing? If the aim of the system was to provide scientists with a way of seeking out periodic points in the parameter space of the Chua circuit, then I would say it performed its function quite admirably, as far as I could tell. Personally, I wouldn't care to listen to to the noise it made for more than a few minutes at a time, but then I don't particularly care to know how many periodic points there are in the paramater space of the Chua circuit! ;^) Mapping the qualities of "chaos" to noise and "periodicity" to pitch seems like a natural enough idea. Unfortunately, our ears do not seem to be particularly good at hearing different qualities of noise, so you lose a lot of potential information about the nature of the chaotic behavior that might prove interesting. That's why I find Choi's examples from ICAD '92 more interesting. Presenting the information at a slower rate allows us to hear "quasi-periodic" aspects (patterns) within the chaos. The trade off, of course, is that you have to listen for a long time to hear the behavior of different input parameters to the circuit. Michael Pelz-Sherman mpelzshe@ucsd.edu From gmk Fri Dec 9 07:09:37 1994 Date: Fri, 9 Dec 94 07:09:37 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Re: Sonic Examples / Variations in fidelity Status: O Michael Pelz-Sherman writes: > The relationship between "fidelity" and "preservation of > quality" is not clear. For example, I recently heard a very > early recording of Elvis Presley. You can clearly hear the > unique "quality" of Elvis' voice, even though the fidelity of > the recording is extremely low. Similarly, the successful > recognition of digitized representations of human faces has > been shown to require very few bits. (I forget exactly where I > read about this, or what the number of bits was, but it was > surprisingly low.) > I think that is an excellent example to make my point clearer: If you choose a "sonic representation" of that peace that "cleans up" the sound but you can't recognize the characteristic quality of Elvis' voice then you blew it. The result is something qualitatively different. > This leads me to believe that there is something more complex at > work here than a simple linear relationship between > "fidelity" and preservation of qualititative information. > I agree completely > Regarding the Choi examples that you refer to (and have > referred to previously), I'm not sure I agree with your > statements that the display we heard and saw at ICAD '94 did not > preserve the qualities of periodicity and chaos. It seemed to > me that when the cursor was placed at certain points along the > mainfold, the sound would become noticably "pitched", > whereas most of the time it sounded like various flavors of > noise. > I am referring to a different example that she played in the first part of her talk, when she played a sequence of five discrete sounds i.e. a periodic sequence. When she chose some interpolation scheme (that she didn't care to define) the resulting sound had clearly some aperiodic modulation involved that I called "artifacts" because it violated the conservation of that "periodic" quality. > That's why I find Choi's examples from ICAD '92 more > interesting. Presenting the information at a slower rate > allows us to hear "quasi-periodic" aspects (patterns) within > the chaos. > Are you talking about the sound sequences sequences that I created using the logistic map? In G. Mayer-Kress, R. Bargar, I. Choi , Musical Structures in Data >From Chaotic Attractors , Technical Report CCSR-92-14, Proceedings of the International Symposium on the Auditory Display (ICAD92), Santa Fe, NM Oct. 1992, Proceedings Volume XVIII Santa Fe Institute Series in the Sciences of Complexity, Addison Wesley, Reading, 1994, (http://www.ccsr.uiuc.edu/People/gmk/Papers/ICAD92/ICAD92.html) For example the file http://www.ccsr.uiuc.edu/People/gmk/Sounds/log52i22.snd plays a periodic sequence (period 5) on one stereo channel and a chaotic (intermittent) sequence on the other, the qualitative difference is quite striking and IMHO illustrates my point quite well. That is the only paper in ICAD'92 that I am aware of where Insook Choi is a co-author. From stephen@csis.dit.csiro.au Sun Dec 11 16:34:42 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA09645; Sun, 11 Dec 94 16:34:40 -0600 Received: from lynx.csis.dit.csiro.au by sfi.santafe.edu (4.1/SMI-4.1) id AA04887; Sun, 11 Dec 94 15:27:13 MST Received: from capella.csis.dit.csiro.au by lynx.csis.dit.csiro.au (8.6.8.1/1.06S) id JAA07667; Mon, 12 Dec 1994 09:28:43 +1100 From: stephen@csis.dit.csiro.au (Stephen Barrass) Received: by capella.csis.dit.csiro.au (8.6.8.1/1.07C) id JAA01470; Mon, 12 Dec 1994 09:28:41 +1100 Date: Mon, 12 Dec 1994 09:28:41 +1100 Message-Id: <199412112228.JAA01470@capella.csis.dit.csiro.au> To: icad@santafe.edu Subject: fidelity Status: O > The relationship between "fidelity" and "preservation of quality" is not clear> For example, I recently heard a very early recording of Elvis Presley. You can > clearly hear the unique "quality" of Elvis' voice, even though the fidelity of > the recording is extremely low. Similarly, the successful recognition of > digitized representations of human faces has been shown to require very few > bits. (I forget exactly where I read about this, or what the number of bits > was, but it was surprisingly low.) > > > This leads me to believe that there is something more complex at work here than > a simple linear relationship between "fidelity" and preservation of > qualititative information. Familiarity with the subject matter plays a huge > role, and certain stimuli will be more resistant to lower fidelity than others. Thanks for illustrating what fidelity means - I may have used it wrongly in my abstract since the intended meaning was "faithfulness" of the subjective perception to the characteristics of the data (i.e. categorical or scalar) not faithfulness to the precision of the physical presentation (i.e. bits, kHz, stereo etc.) which seems to have been implied. stephen From gmk Mon Dec 12 09:23:37 1994 Date: Mon, 12 Dec 94 09:23:37 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Re: fidelity / EEGs Status: O Stephen Barrass writes: > the intended meaning was "faithfulness" of the subjective > perception to the characteristics of the data (i.e. > categorical or scalar) not faithfulness to the precision of > the physical presentation (i.e. bits, kHz, stereo etc.) which > seems to have been implied. > I would agree that this "faithfulness" would be the most important property of acoustic rendering from a scientific point of view especially in the context of complex and/or chaotic systems. I wanted to mention one quality that I think very interesting important for many application, it is the quality of "synchrony" of signals. I tried to present a simple attempt in that direction in my EEG Sound demo (see also http://www.ccsr.uiuc.edu/People/gmk/Projects/EEGSound/) where one would like to identify events of many electrodes being in synchrony. BTW, yesterday I found by chance a URL on the NCSA site by Steven M. Cohen (http://www.ncsa.uiuc.edu/EVL/docs/SHOWCASE/steve/steve.html) on Mapping of Cognitive Function with Sudural Electrodes & Registration of Cerebral Evoked Potentials on 3-D MRI of the SIGGRAPH'92 showcase. (It is strange that the local sound experts at NCSA seemed not to be aware of that work.) I tried to contact the authors but I was wondering if anyone is familiar with this or related work. Thanks, --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From stephen@csis.dit.csiro.au Mon Dec 12 18:12:33 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA00248; Mon, 12 Dec 94 18:12:31 -0600 Received: from lynx.csis.dit.csiro.au by sfi.santafe.edu (4.1/SMI-4.1) id AA17969; Mon, 12 Dec 94 16:55:28 MST Received: from capella.csis.dit.csiro.au by lynx.csis.dit.csiro.au (8.6.8.1/1.06S) id KAA05998; Tue, 13 Dec 1994 10:56:42 +1100 From: stephen@csis.dit.csiro.au (Stephen Barrass) Received: by capella.csis.dit.csiro.au (8.6.8.1/1.07C) id KAA01706; Tue, 13 Dec 1994 10:56:33 +1100 Date: Tue, 13 Dec 1994 10:56:33 +1100 Message-Id: <199412122356.KAA01706@capella.csis.dit.csiro.au> To: icad@santafe.edu Subject: hearing synchrony Status: O Gottfried Meyer-Kress writes > I wanted to mention one quality that I think very interesting > important for many application, it is the quality of "synchrony" > of signals. have you considered the "gestalt" possibilities of streaming due to shared modulations (or common fate) ? (Bregman) Perhaps if each eeg probe is assigned a simple sinusoidal tone (of identical Hz) which is frequency or amplitude modulated in a perceptually linearised range by the data than those with similar modulation rates will group into separate perceptual streams. If each probe had a unique freq then perhaps the timbre and pitch of the stream would identify the component traces too. This idea comes from an approach I read in a (draft?) paper "application of scene analysis principles to sonification" by Jon Barker and Martin Cooke, Uni of Sheffield. Better listen to your example to hear what you are actually doiing .... Some Time Later... this example is similar in concept but you employ large modulations of discrete pitches in distinct timbres which causes strong stream segregation - relying rather on grouping by onset synchrony. these examples are very helpful for thinking through approaches and ideas... thanks stephen #@%#%^! Internet : stephen@csis.dit.csiro.au #Q!^&%#$@!^&$#& Phone : +61-6-2750938 #!@&^#@!^&&^%#&^@! Fax : +61-6-2571052 @!#@#!&(*00!@#&&@#* Address : CSIRO Division of Information Technology, 0#@!%^!@#^%&&^&& GPO Box 664, Canberra ACT 2601 @#!%$^@#!%^^ AUSTRALIA @#!*^&*^#@# #%#$ ^&^&%^ "How did you get so much custard out of such a small cat?"$@ $#5 Blackadder 4 & *% @ From gmk Mon Dec 12 18:25:55 1994 Date: Mon, 12 Dec 94 18:25:55 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Re: hearing synchrony Status: O Stephen Barrass writes: > have you considered the "gestalt" possibilities of streaming due to > shared modulations (or common fate) ? (Bregman) > ummm, I am not sure... >Perhaps if each eeg probe is assigned a simple sinusoidal tone (of identical Hz) > yes, some of them, that is what we want to detect. The frequency is around 40 Hz in the original signal and it lasts for about 100-200ms, i.e. 4-8 oscillations then it's gone, back to chaos. > which is frequency or amplitude modulated in a perceptually linearised range > by the data than those with similar modulation rates will group into separate > perceptual streams. > you mean? > this example is similar in concept but you employ large > modulations of discrete pitches in distinct timbres which > causes strong stream segregation - relying rather on grouping > by onset synchrony. > I am just a dumb physicist, sorry, it was the best I could think of. >these examples are very helpful for thinking through approaches and ideas... > I am looking forward to hear what you come up with. Thanks for the comments, Gottfried --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From stephen@csis.dit.csiro.au Mon Dec 12 19:14:49 1994 Return-Path: Received: from santafe.edu by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA00299; Mon, 12 Dec 94 19:14:48 -0600 Received: from lynx.csis.dit.csiro.au by sfi.santafe.edu (4.1/SMI-4.1) id AA18541; Mon, 12 Dec 94 18:02:41 MST Received: from capella.csis.dit.csiro.au by lynx.csis.dit.csiro.au (8.6.8.1/1.06S) id MAA07637; Tue, 13 Dec 1994 12:04:02 +1100 From: stephen@csis.dit.csiro.au (Stephen Barrass) Received: by capella.csis.dit.csiro.au (8.6.8.1/1.07C) id MAA01713; Tue, 13 Dec 1994 12:03:59 +1100 Date: Tue, 13 Dec 1994 12:03:59 +1100 Message-Id: <199412130103.MAA01713@capella.csis.dit.csiro.au> To: icad@santafe.edu Subject: hearing synchronicity again Status: O wooops, I think my previous analysis of what Gottfried was doing is wrong! He does suggest using onset timing in his WWW page but the sound example is of another technique where the pitch of distinct timbres is modulated over a large range in sychronous steps. The onset synchrony tends to cause causes grouping at each step which competes with pitch and timbre segregation - this allows the perception of pitch "outliers" which seperate from the overall orchestral gestalt. stephen From gmk Mon Dec 12 19:25:16 1994 Date: Mon, 12 Dec 94 19:25:16 GMT-0600 From: gmk (To: icad@santafe.edu) To: icad@santafe.edu Subject: Re: hearing synchronicity again Status: O Stephen, I don't understand why your analysis is wrong. Yes, I did suggest to use Staggered Onset but I have not done that yet. > The onset synchrony tends to cause causes grouping at each step > which competes with pitch and timbre segregation - this allows the perception > of pitch "outliers" which seperate from the overall orchestral gestalt. > do you have a suggestion of how to improve the resolution? Thanks, Gottfried From spf@hoqst1.att.com Mon Jan 2 23:34:28 1995 Return-Path: Received: from gw1.att.com by sfi.santafe.edu (4.1/SMI-4.1) id AA25425; Mon, 2 Jan 95 23:26:22 MST Received: from hoqst1.UUCP by ig1.att.att.com id AA04342; Tue, 3 Jan 95 01:26:00 EST Message-Id: <9501030626.AA04342@ig1.att.att.com> From: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596) Date: 3 Jan 95 06:31:00 GMT To: pegasos.ccsr.uiuc.edu!gmk@ig2.att.att.com.att.com (Gottfried Mayer-Kress) Original-From: hoqst1!spf (Steven P Frysinger +1 908 949 7596) Original-To: att!ig2.att.att.com!pegasos.ccsr.uiuc.edu!gmk (Gottfried Mayer-Kress) Cc: Computer Music Centre@pegasos.ccsr.uiuc.edu, , icad@santafe.edu, sound@ACM.ORG, sound@acm.org Subject: Re: Double mail Status: RO I don't think the WWW facility will serve the same purpose as a newsgroup. True, one can (and should be able to) find and read past threads via the Web. But to actively participate in an ongoing discussion implies a degree of interactivity which I would describe as a dialog (albeit a public one). Usenet groups work well for this because they incorporate the best of both worlds - they are interactive, yet can be browsed. Archives should, naturally, be maintained on the Web, for the benefit of lurkers or other folks not involved in the actual dialog. Another benefit, of course, is that Usenet is available to folks living in the 2400 baud world or without TCP/IP access or reliable phone lines. Given the relatively impoverished nature of our line of research, I would not want to assume that all of our number have access to the latest and greatest technology. Indeed, forced to a choice, I'd rather folks put their money into a lab and ran ADR experiments, in lieu of subscribing to a TCP/IP service if need be. I intend that the ICAD WWW presence be the best that it can be - but I also hope that we provide an interactive medium (be it the mailing list or a Usenet group) for the benefit of ALL workers in auditory display. While I'm sorely impressed with WWW/Mosaic and the possibilities it opens up to us, I'll have to paraphrase Mark Twain: "the rumors of the demise of Usenet are largely overstated." Steven P. Frysinger AT&T Bell Laboratories TEL: 908/949-7596 FAX: 201/952-4266 NET: spf@hoqst1.att.com SLO: 101 Crawfords Corner Road Holmdel, New Jersey 07733 From gmk@pegasos.ccsr.uiuc.edu Tue Jan 3 00:32:48 1995 Return-Path: Received: from complex.ccsr.uiuc.edu by sfi.santafe.edu (4.1/SMI-4.1) id AA25591; Tue, 3 Jan 95 00:24:19 MST Received: from pegasos.ccsr.uiuc.edu (pegasos.ccsr.uiuc.edu) by complex.ccsr.uiuc.edu with SMTP id AA10901 (5.67a/IDA-1.5 for ); Tue, 3 Jan 1995 01:25:24 -0600 Received: by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA01849; Tue, 3 Jan 95 01:22:22 -0600 Date: Tue, 3 Jan 95 01:22:22 -0600 From: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress) Message-Id: <9501030722.AA01849@ pegasos.ccsr.uiuc.edu > Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596) Subject: Re: Double mail Cc: dhj@watson.ibm.com, icad@santafe.edu, sound@ACM.ORG Status: RO Steven P Frysinger writes: > I don't think the WWW facility will serve the same purpose as a newsgroup. > True, one can (and should be able to) find and read past threads via the > Web. But to actively participate in an ongoing discussion implies a degree > of interactivity which I would describe as a dialog (albeit a public one). > Usenet groups work well for this because they incorporate the best of both > worlds - they are interactive, yet can be browsed. > well, there are several technical possibilities of the integration that I mentioned in my previous mail and which would satisfy all of the points that you addressed here: - you can include in a WWW page any news group you want interactively. As an example you can check my home page http://www.ccsr.uiuc.edu/People/gmk/Links/Start.html where I have a link to the newsgroup clari.world.top the corresponding entry in the html file is: News For a more professional set-up check the discussion groups of Time Magazine, very nicely organized with different threads and forms for changing the arrangements. For users who don't want to ore are unable to use forms, they have an e-mail option, that's what I meant by "integration". What I could imagine to be useful for this group is a page that contains a standard mailing list archive together with links to the relevant newsgroups and an interactive, forms based bbs. >Another benefit, of course, is that Usenet is available to folks living in the 2400 >baud world or without TCP/IP access or reliable phone lines. > that's indeed an important issue, especially if you consider participants in developing countries. We have discussed this problem at the WWW-II conference in Chicago in the fall and suggested to use the html documents in a stand-alone mode. The pages can then be sent by e-mail or even put on a floppy and sent by regular mail or sneaker-net. (For more details on this project for the World Hunger Program: http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/CorInfSys/gmk/Mayer-Kress/HungerConf.html ) > up to us, I'll have to paraphrase Mark Twain: "the rumors of the demise of > Usenet are largely overstated." > well, it is already fully integrated in the web... --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From gmk@pegasos.ccsr.uiuc.edu Tue Jan 3 08:22:04 1995 Return-Path: Received: from complex.ccsr.uiuc.edu by sfi.santafe.edu (4.1/SMI-4.1) id AA27026; Tue, 3 Jan 95 08:15:11 MST Received: from pegasos.ccsr.uiuc.edu (pegasos.ccsr.uiuc.edu) by complex.ccsr.uiuc.edu with SMTP id AA12137 (5.67a/IDA-1.5 for ); Tue, 3 Jan 1995 09:16:17 -0600 Received: by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA02166; Tue, 3 Jan 95 09:13:15 -0600 Date: Tue, 3 Jan 95 09:13:15 -0600 From: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress) Message-Id: <9501031513.AA02166@ pegasos.ccsr.uiuc.edu > Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596) Subject: Re: Double mail Cc: dhj@watson.ibm.com, icad@santafe.edu, sound@ACM.ORG Status: RO Steven P. Frysinger writes: > Agreed - if there's a newsgroup, it should be accessible via the Web. But > for non-Web users, I'd rather their fallback option be a newsgroup in their > .newsrc, rather than a mailing list. Just a preference on my part. Historically, > as subjects mature they evolve from mailing list to newsgroup, because of > the human factors benefits I previously cited. > Steve, I understand your preference for newsgroups, my point is that I believe that the access method (mail, netnews, WWW) should be an interface issue and not a content issue. What I suggest as a goal for the ICAD installation is to give the user the option how to access the information. Sending mail to a list, posting to a news group, and using a WWW forms interface should be left a a personal preference to the member but they all should have access to the same information. Thus I don't see it as an either/or problem but suggest to do leave the choice of interface to the user (we didn't mention gopher, I am sure some would prefer to monitor the discussion via gopher and then send some occasional mail.) (Incidentally, I just wrote a little essay on a closely related subject for the Knowbotic Research group in Cologne and would be interested in comments from this group. The URL is: http://www.ccsr.uiuc.edu/People/gmk/Papers/gmk_kr_cf/ ) Gottfried --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ Warranty: None of our calculations were done on a Pentium chip! From gmk@pegasos.ccsr.uiuc.edu Tue Jan 3 12:33:39 1995 Return-Path: Received: from complex.ccsr.uiuc.edu by sfi.santafe.edu (4.1/SMI-4.1) id AA01702; Tue, 3 Jan 95 12:10:07 MST Received: from pegasos.ccsr.uiuc.edu (pegasos.ccsr.uiuc.edu) by complex.ccsr.uiuc.edu with SMTP id AA13592 (5.67a/IDA-1.5 for ); Tue, 3 Jan 1995 13:11:14 -0600 Received: by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA02516; Tue, 3 Jan 95 13:08:12 -0600 Date: Tue, 3 Jan 95 13:08:12 -0600 From: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress) Message-Id: <9501031908.AA02516@ pegasos.ccsr.uiuc.edu > Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: icad@santafe.edu Subject: Re: Double mail Cc: SOUND@ACM.ORG Status: RO David Jaffe writes: > Unless you work for a company or university, it is quite expensive to have WWW or > Usenet access, whereas unlimited email (e.g via UUCP) is less than a dollar a day. > > I vote for keeping it as is. > the non-realtime aspects of mail are indeed a major cost advantage (especially for developing countries). For discussions like on newsgroups it really is not necessary to have an online connection. The organizational advantages of netnews can also be implemented locally. WWW/Mosaic in the local mode, that I described before, fulfills both goals: you receive the html files by regular mail and use the hypertext capabilities of WWW/Mosaic for the structuring and organization in a completely consistent manner. Important aspect especially for this group is also the multi-media capabilities of modern mail systems: if you want to discuss a certain sound file, it will be more difficult on netnews then having the file automatically sent by e.g. a mime server. --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From ballas@AIC.NRL.Navy.Mil Tue Jan 3 13:17:16 1995 Received: from Sun0.AIC.NRL.Navy.Mil by sfi.santafe.edu (4.1/SMI-4.1) id AA02403; Tue, 3 Jan 95 13:11:29 MST Received: from sun35.aic.nrl.navy.mil by Sun0.AIC.NRL.Navy.Mil (4.1/SMI-4.0) id AA24134; Tue, 3 Jan 95 15:13:19 EST Return-Path: Received: by sun35.aic.nrl.navy.mil; Tue, 3 Jan 95 15:13:18 EST Date: Tue, 3 Jan 95 15:13:18 EST From: ballas@AIC.NRL.Navy.Mil Message-Id: <9501032013.AA15306@sun35.aic.nrl.navy.mil> To: icad@santafe.edu Subject: WWW vs. newsgroup Status: RO I would prefer a newsgroup also. When I have had the option of using WWW in local mode with files distributed as GMK suggests, I filed most of the info, and had to delete it before I fully explored it. From gmk@pegasos.ccsr.uiuc.edu Tue Jan 3 13:44:12 1995 Return-Path: Received: from complex.ccsr.uiuc.edu by sfi.santafe.edu (4.1/SMI-4.1) id AA02565; Tue, 3 Jan 95 13:33:13 MST Received: from pegasos.ccsr.uiuc.edu (pegasos.ccsr.uiuc.edu) by complex.ccsr.uiuc.edu with SMTP id AA14069 (5.67a/IDA-1.5 for ); Tue, 3 Jan 1995 14:34:21 -0600 Received: by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA02664; Tue, 3 Jan 95 14:31:19 -0600 Date: Tue, 3 Jan 95 14:31:19 -0600 From: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress) Message-Id: <9501032031.AA02664@ pegasos.ccsr.uiuc.edu > Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: ballas@AIC.NRL.Navy.Mil Subject: Re: WWW vs. newsgroup Cc: icad@santafe.edu Status: RO Jim Ballas writes: > I would prefer a newsgroup also. When I have had the option of > using WWW in local mode with files distributed as GMK suggests, > I filed most of the info, and had to delete it before I fully explored it. > the nice thing about these new Internet toys is that they can be integrated and are no longer XOR --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ From spf@hoqst1.att.com Tue Jan 3 16:32:48 1995 Return-Path: Received: from gw1.att.com by sfi.santafe.edu (4.1/SMI-4.1) id AA05726; Tue, 3 Jan 95 16:27:48 MST Received: from hoqst1.UUCP by ig1.att.att.com id AA13772; Tue, 3 Jan 95 09:38:49 EST Message-Id: <9501031438.AA13772@ig1.att.att.com> From: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596) Date: 3 Jan 95 14:33:00 GMT To: pegasos.ccsr.uiuc.edu!gmk@ig2.att.att.com.att.com (Gottfried Mayer-Kress) Original-From: hoqst1!spf (Steven P Frysinger +1 908 949 7596) Original-To: att!ig2.att.att.com!pegasos.ccsr.uiuc.edu!gmk (Gottfried Mayer-Kress) Cc: dhj@watson.ibm.com, icad@santafe.edu, sound@ACM.ORG Subject: Re: Double mail Status: RO Agreed - if there's a newsgroup, it should be accessible via the Web. But for non-Web users, I'd rather their fallback option be a newsgroup in their .newsrc, rather than a mailing list. Just a preference on my part. Historically, as subjects mature they evolve from mailing list to newsgroup, because of the human factors benefits I previously cited. As you suggest, the dialog forum (whether newsgroup or mailing list) should be accessible via the Web for those who have the resources and desire. Steve Steven P. Frysinger AT&T Bell Laboratories TEL: 908/949-7596 FAX: 201/952-4266 NET: spf@hoqst1.att.com SLO: 101 Crawfords Corner Road Holmdel, New Jersey 07733 From spf@hoqst1.att.com Tue Jan 3 16:34:18 1995 Return-Path: Received: from gw1.att.com by sfi.santafe.edu (4.1/SMI-4.1) id AA05730; Tue, 3 Jan 95 16:27:54 MST Received: from hoqst1.UUCP by ig1.att.att.com id AA03666; Tue, 3 Jan 95 12:43:13 EST Message-Id: <9501031743.AA03666@ig1.att.att.com> From: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596) Date: 3 Jan 95 16:02:00 GMT To: pegasos.ccsr.uiuc.edu!gmk@ig2.att.att.com.att.com (Gottfried Mayer-Kress) Original-From: hoqst1!spf (Steven P Frysinger +1 908 949 7596) Original-To: att!ig2.att.att.com!pegasos.ccsr.uiuc.edu!gmk (Gottfried Mayer-Kress) Cc: dhj@watson.ibm.com, icad@santafe.edu, sound@ACM.ORG Subject: Re: Double mail Status: RO Gotcha, Gottfried - I'll check it out. Anybody else have thoughts on the issue of setting up an ICAD newsgroup on Usenet? Steven P. Frysinger AT&T Bell Laboratories TEL: 908/949-7596 FAX: 201/952-4266 NET: spf@hoqst1.att.com SLO: 101 Crawfords Corner Road Holmdel, New Jersey 07733 From powell@utopia.esd.sgi.com Wed Jan 4 12:01:55 1995 Return-Path: Received: from sgigate.sgi.com by sfi.santafe.edu (4.1/SMI-4.1) id AA12451; Wed, 4 Jan 95 11:50:35 MST Received: from sgihub.corp.sgi.com by sgigate.sgi.com via ESMTP (940627.SGI.8.6.9/911001.SGI) id KAA27754; Wed, 4 Jan 1995 10:51:08 -0800 Received: from utopia.esd.sgi.com by sgihub.corp.sgi.com via ESMTP (940519.SGI.8.6.9/911001.SGI) id KAA20111; Wed, 4 Jan 1995 10:51:05 -0800 Received: by utopia.esd.sgi.com (940816.SGI.8.6.9/911001.SGI) id KAA10911; Wed, 4 Jan 1995 10:50:38 -0800 From: "Roger A. Powell" Message-Id: <9501041050.ZM10909@utopia.esd.sgi.com> Date: Wed, 4 Jan 1995 10:50:22 -0800 In-Reply-To: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596) "Re: Double mail" (Jan 3, 6:31am) References: <9501030626.AA04342@ig1.att.att.com> X-Mailer: Z-Mail (3.2.0 26oct94 MediaMail) To: spf@hoqst1.att.com (Steven P Frysinger +1 908 949 7596), pegasos.ccsr.uiuc.edu!gmk@ig2.att.att.com.att.com (Gottfried Mayer-Kress) Subject: Re: Double mail Cc: Computer.Music.Centre@pegasos.ccsr.uiuc.edu, , icad@santafe.edu, sound@ACM.ORG Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Status: RO Alas, much too much for me...Please take me off this list... All the best in '95! Roger Powell From gmk@pegasos.ccsr.uiuc.edu Wed Jan 4 14:27:28 1995 Return-Path: Received: from complex.ccsr.uiuc.edu by sfi.santafe.edu (4.1/SMI-4.1) id AA14692; Wed, 4 Jan 95 14:12:43 MST Received: from pegasos.ccsr.uiuc.edu (pegasos.ccsr.uiuc.edu) by complex.ccsr.uiuc.edu with SMTP id AA19038 (5.67a/IDA-1.5 for ); Wed, 4 Jan 1995 15:13:50 -0600 Received: by pegasos.ccsr.uiuc.edu (NX5.67d/NeXT-2.0) id AA01685; Wed, 4 Jan 95 15:10:43 -0600 Date: Wed, 4 Jan 95 15:10:43 -0600 From: gmk@pegasos.ccsr.uiuc.edu (Gottfried Mayer-Kress) Message-Id: <9501042110.AA01685@ pegasos.ccsr.uiuc.edu > Received: by NeXT.Mailer (1.100) Received: by NeXT Mailer (1.100) To: chris@conceit.isem.smu.edu (Chris Hayward - local) Subject: Re: Personal Notes from ICAD '94 Cc: icad@santafe.edu, sound@ACM.ORG Status: RO Chris Hayward writes: > I've thought long and hard -- well I gave it a passing thought -- > about posting the following long set of notes to the list. I decided > to go ahead with this one posting rather than breaking it up (this > way there is just one message from me to delete), and post it now > rather than continuing to edit it (it's been two months already - > it would probably be a year or more before getting back to it). > > So, for all of those whose mailboxes I break, you have my apologies. > If we had a web server, I'd leave the document there instead. > I just created a temporary page http://www.santafe.edu/~gmk/ICAD/ which can be used for such and similar purposes until the official ICAD page comes online. It contains links to WWW pages, mail archives, and news groups. Thus some of the issues we discussed earlier on this list can be tested. Enjoy, --- Gottfried Mayer-Kress Center for Complex Systems Research, Beckman Institute, Physics Dept. UIUC 405 N Mathews, Urbana, Il 61801,gmk@pegasos.ccsr.uiuc.edu, gmk@igc.apc.org, gmk@santafe.edu, (217)-244-5877(voice/fax1),x8371(fax2), x1994 (msg) URLs: http://www.santafe.edu/~gmk/, ftp://ftp.santafe.edu/pub/gmk http://www.ccsr.uiuc.edu/~gmk/ Warranty: None of our calculations were done on Pentium chips! From jpff@maths.bath.ac.uk Thu Jan 5 04:15:08 1995 Return-Path: Received: from tamarin.bath.ac.uk by sfi.santafe.edu (4.1/SMI-4.1) id AA19811; Thu, 5 Jan 95 04:06:57 MST Message-Id: <9501051106.AA19811@sfi.santafe.edu> Received: from omphalos.maths.bath.ac.uk by tamarin.bath.ac.uk with SMTP (PP) id <18977-0@tamarin.bath.ac.uk>; Thu, 5 Jan 1995 11:08:23 +0000 Date: Thu, 5 Jan 95 11:10:09 GMT From: jpff@maths.bath.ac.uk Subject: Re: Double mail To: icad@santafe.edu, SOUND@acm.org Status: RO Message written at 4 Jan 1995 20:29:11 +0000 --- Copy of mail to gmk@edu.uiuc.ccsr.pegasos --- In-reply-to: <9501031908.AA02516@ pegasos.ccsr.uiuc.edu > (message from Gottfried Mayer-Kress on Tue, 03 Jan 1995 13:08:12 -0600) I would like to add my own feelings on this. I _much_ prefer e-mail for important newsgroups (as distinct from marginal things like uk.politics). That way I am forced to consider the information. The News server is not always up; it refuses to read certain groups if there is a large file in it (even if I have already read it). WWW is a righ-royal pain. I cannot use it at home (where i usually deal with all my mail -- like now) it is slow and cumbersome, and I am not convinced by the powers of hypertext. I prefer a well argued case to unstructured wandering. So like David Jaffe I like email, despite working (mainly) at a university. ==John ff From stephen@csis.dit.csiro.au Thu Jan 5 22:25:22 1995 Return-Path: Received: from lynx.csis.dit.csiro.au by sfi.santafe.edu (4.1/SMI-4.1) id AA28412; Thu, 5 Jan 95 22:20:50 MST Received: from capella.csis.dit.csiro.au by lynx.csis.dit.csiro.au (8.6.8.1/1.06S) id QAA17446; Fri, 6 Jan 1995 16:22:41 +1100 From: stephen@csis.dit.csiro.au (Stephen Barrass) Received: by capella.csis.dit.csiro.au (8.6.8.1/1.07C) id QAA13264; Fri, 6 Jan 1995 16:22:40 +1100 Date: Fri, 6 Jan 1995 16:22:40 +1100 Message-Id: <199501060522.QAA13264@capella.csis.dit.csiro.au> To: icad@santafe.edu Subject: zipi Status: RO hi icad, I just read about a thing called zipi it is a proposal for a music parameter description language which addresses many of the shortcomings of midi for music, such as lack of articulation, event addressing based on pitch, low bandwidth serial protocol, keyboard orientation, etc. etc. etc. although it is for music it may also be helpful for sonification and auditory displays because it is much more flexible than midi, if you are interested there is a description at ftp://mitpress.mit.edu/pub/Computer-Music-Journal/Texts/ZIPI (BTW: how do you know if you are making "noise" or "signal" ?, for instance maybe everyone else knows about this....) stephen