Home Syllabus Week-by-week Homework FAQ Archives Projects

ISE575
CSCI575
EE675

 
Engineering Approaches to
Music Perception and Cognition
Spring 2005   Daniel J. Epstein Department of Industrial and Systems Engineering
University of Southern California Andrew and Erna Viterbi School of Engineering
Homework FAQ

SUN, May 8, 2005

All pictures are also online now ... http://www-scf.usc.edu/~ise575/pictures. Enjoy!

SAT, May 7, 2005

I have posted your projects online. Please visit http://www-scf.usc.edu/~ise575/projects and check that you are satisfied with the way your project appears/is presented.

THU, April 28, 2005

Your project presentation will take place TODAY. You should do whatever you think necessary to communicate and demonstrate to the class what you did (be it a powerpoint show or a walk-through of your project code and website, a demo is usually quite useful). Yes, you are supposed to document your project the way the students did last year, and yes that is due today ... several students have already completed and submitted this portion of their work. I will take revised submissions up through noon on Monday, May 2nd.

According to the syllabus, the final project is worth 25% of your grade. The final project grade is split into 35% presentation, 35% content and 30% documentation. I expect to receive a zip/tar file of all your project (including web) documents. In my experience, personal websites come and go; so, I am planning to keep a mirror of your project website on the course webpage for posterity even if you currently have the same documentation on your own website. (Please raise any objections now.)

See you all this evening. I am looking forward to your presentations.

MON, April 11, 2005

For those of you who were not at last week's class, this week's homework is to choose two of the music toolkits mentioned in your syllabus (see the online one as the list has increased from 3 to 4), play with each and make something that you will bring to class to show-and-tell. You are responsible for giving a score for each of the projects shown by your classmates; for each person, I will eliminate the highest and lowest score (like in the ice-skating competitions) and take the average as the grade.

Q: I have been trying to work on this week's hw. I remember from the class that we are supposed to work with and compose with two of the systems. But every one I checked was a toolbox/library for programming and analysis of midi/music data. Did I miss something here? Are we supposed to write a program in which we can compose?

I am modifying the homework description to: either compose or analyze a piece with the tools. For a composition, you will likely have to write a short program or script to generate (algorithmically or by other means) some piece. For an analysis (say, using the midi toolbox), you will have to find a piece of music to analyze.

FRI, April 1, 2005

Please send in a paragraph-long project proposal. I have only received one so far.

The paper that I can't seem to download is the neural networks paper by Bresin.

The earlier Bresin paper is not available digitally online. JNMR only has digital collections from Vol 28 onwards, and the assigned reading is in Vol 27. See me about getting a copy.

WED, March 22, 2005

Q: In the homework, we need to analyze a midi file. But the question is: where can we find the midi file for the first choice, Bach's G major Minuet?

Q: Quick question, are we supposed to transcribe the G Major minuet in the appendix to a notefile ourselves? Or is there a MIDI file posted of that one. I googled around and found a couple different midi files but they seem to subtly differ from the appendix, at least by the produced notefile results. I don't have sibleius or finale handy to visually check the differences. I was just curious if an official MIDI of the appendix segment was posted.

A: No, you do not need to transcribe the Minuet yourself to get a MIDI file or notefile. You should be able to find a MIDI file of Bach's Minuet in G easily on the web (through your favorite search engine). The version you find will likely have a few more notes than the score (ornaments, extra notes, which are typical of the style will likely be included). The main criteria you should worry about is that the file should be expressionless ... i.e. the beats extremely regular.

TUE, March 21, 2005

From David Temperley: Anja just wrote to me to say she had a problem compiling mftext, the program that converts midifiles to notefiles. This problem occurs on certain machines, and maybe you had it also. You can fix it by going into mftext.c and replacing

(void) fprintf(stderr,"*** ERROR *** Cannot open '%s'!\n",name); if ( errno <= sys_nerr ) errmess = sys_errlist[errno]; else errmess = "Unknown error!"; (void) fprintf(stderr,"************* Reason: %s\n",errmess);
with
fprintf(stderr,"*** ERROR in mftext *** Cannot open '%s'!\n",name); perror("Reason");
...Or you can just redownload mftext.c from the ftp site (I just changed it there). (The students will need this to do the homework!)

THU, March 10, 2005

For uniformity and documentation of your progress, I would like to request that the filenames of the reports you turn in (from now on) conform to the following format: YourLastname-WeekNumber.PaperNumberAsListedInOnlineSyllabus.pdf . For example, my report on the 3rd paper of week 7 would be labeled chew-7.3.pdf .

THU, February 10, 2005

Although reading the papers and specially trying to summarize them into a one-page report prepares us for the class presentations, and helps us to understand them better, it is not always easy to understand everything in the papers and sometimes I don't seem to understand a part! What am I supposed to do in that case? My guess is that it should be brought up in the next class, during/after the presentation; but in that case, I am not able to write a good report on the paper. Just as a concern, how important our reports are and what are the important factors in grading them?
I don't expect you to understand everything in the papers. Apart from the fact that you may not be familiar with the terminology and techniques, not all research papers are written completely and with all clarity and details. I do expect you to make your best effort to understand the ideas, and to re-present and critically examine them as best you can in a report. The weekly paper reviews are 40% of your final grade, which makes each report less than 2% of your your final grade. I have posted some of the better reports under "Week-by-week Guide" as examples.

THU, February 3, 2005

Sample HW1's have been posted. Please visit these homeworks to familiarize yourself with your colleagues' work and to larn from them. Anja Volk will be leading the class next week (I will also be present). Please contact me if you need a copy of next week's reading. As Anja has pointed out, the authors have taken their copies offline and CMJ has yet to post the 29:1 issue.

THU, January 27, 2005

[Musedata] I know the piece has a key signature of Eb, Bb. The format only has a place for a Key parameter that is an integer number -40 to 40. There is no explanation in spec for how to translate a key signature into this key. The only thing that describes a -40 to 40 integer is the transpose parameter, but the explanation does not seem to apply to key. Is there any advice you can give on this subject? Currently I am just using text parameters.
A major key with Eb and Bb in the key signature is in the key of Bb major. From the examples, it appears that the key is notated according to the number of sharps/flats. So for example, C major (no sharps or flats) is K:0; A major (with 3 sharps) is K:3. Bb major (with two flats) should be K:-2.

My other question was if there is an implied rest on the 1st and 5th measure? They seem to only have 10 counts instead of 12.
Yes. You can either assume a rest of 1/8 note after the quarter note, or assume that the quarter note is missing a dot, which would make it a dotted quarter note having value 3/8.

WED, January 26, 2005

While reading the documentation, I noticed that there are many sections in addition to "pitch" and "duration", e.g. ORNAMENTS, UP & DOWN BOWS, N-TUPLETS. Are we supposed to ignore those and write the program only to translate pitches and durations?
The intent of the homework is for the encoding of the score to be done with anything in the documentation needed to represent the score. The "translation" part of the assignment is to extract only the pitch and onset information from the file.

MON, January 24, 2005

Thank you, all who submitted your homeworks and MIDI files on time. I have posted all compositions I have received at http://www-scf.usc.edu/~ise575/students/hw1.html. Kindly let me know if you have not signed up for the class or have dropped it so that this mailing list stays current.

Q1. For the symbols on top of the first staff, like Bb, F/Bb, are they corresponding to the notes on the bottom staff? or other meanings?
These are guitar chords, or chords to guide other instrumentalists to improvise on top of the score.

Q2. Should the first note on the middle staff to be a dotted half note instead of a quarter note?
Yes, a dotted quarter note would be more consistent with the rest of the score. Do state this assumption in your solution.

FRI, January 21, 2005

Please pick up your latest assignment ... [PDF].

MON, January 17, 2005

I understand that we have to compose something in Impromptu. Do we have to select any of the files in the catalog or just the Ambrosian which was used by the students in the paper? After doing that do we have to write about our experience in detail like they did?
You may select any file in the catalogue, you are not limited to Ambrosian. Yes, you should document your experience like they did.

I'm still not entirely sure what's expected for HW1. From my understanding, I should:
* Create a piece with Impromptu and record a log of my changes.
* Write a review of the experience
* Do the same thing with Hyperscore
Is that all? I wanted to make sure I'm not missing something.

You should also read the two papers listed with the software. Your writeup of your experience should also be a report on these papers and how they compare with your experience.

Updated 8 May 2005.

The University of Southern California does not screen or control the content on this website and thus does not guarantee the accuracy, integrity, or quality of such content. All content on this website is provided by and is the sole responsibility of the person from which such content originated, and such content does not necessarily reflect the opinions of the University administration or the Board of Trustees